<?xml version="1.0" encoding="UTF-8"?>
 <rdf:RDF xmlns="http://purl.org/rss/1.0/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:cc="http://web.resource.org/cc/" xmlns:syn="http://purl.org/rss/1.0/modules/syndication/" xmlns:admin="http://webns.net/mvcb/">
  <channel rdf:about="http://pinboard.in">
    <title>Pinboard (cshalizi)</title>
    <link>https://pinboard.in/u:cshalizi/public/</link>
    <description>recent bookmarks from cshalizi</description>
    <items>
      <rdf:Seq>	<rdf:li rdf:resource="https://faculty.washington.edu/yenchic/short_note/note_MoM.pdf"/>
	<rdf:li rdf:resource="https://theoryandpractice.org/2024/10/Yes,%20we%20did%20discover%20the%20Higgs!/"/>
	<rdf:li rdf:resource="https://www.cambridge.org/core/journals/philosophy-of-science/article/abs/carnapian-inductive-logic-for-exponential-smoothing/84DD31142459DFD0289CCE8915E79952?WT.mc_id=New%2520Cambridge%2520Alert%2520-%2520Articles"/>
	<rdf:li rdf:resource="https://www.aeaweb.org/articles?id=10.1257/aer.20240246"/>
	<rdf:li rdf:resource="https://www.cambridge.org/core/journals/philosophy-of-science/article/measured-inference-scales-statistics-and-scientific-inference/527F6793A1C954F01A321E72F780E931?WT.mc_id=New%2520Cambridge%2520Alert%2520-%2520Issues"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2512.24999"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2512.00175"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2510.16174"/>
	<rdf:li rdf:resource="https://www.cambridge.org/core/journals/philosophy-of-science/article/measured-inference-scales-statistics-and-scientific-inference/527F6793A1C954F01A321E72F780E931?WT.mc_id=New%2520Cambridge%2520Alert%2520-%2520Articles"/>
	<rdf:li rdf:resource="https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5152665"/>
	<rdf:li rdf:resource="https://www.cambridge.org/core/journals/philosophy-of-science/article/logit-model-measurement-problem/4270AECB0E1E0CB6DEBE82A67A3ABD9F?WT.mc_id=New%2520Cambridge%2520Alert%2520-%2520Issues"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2505.13432"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2412.20355"/>
	<rdf:li rdf:resource="https://www.cambridge.org/us/universitypress/subjects/physics/mathematical-methods/probability-theory-quantitative-scientists?format=HB&amp;WT.mc_id=NGV_IOC_BC_BK%253b_PHYS%253b_Probability%2BTheory%2Bfor%2BQuantitative%2BScientists_Jul25"/>
	<rdf:li rdf:resource="https://imstat.org/2017/04/01/obituary-ulf-grenander-1923-2016/"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2504.03503"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2503.15850"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2401.15800"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2501.07772"/>
	<rdf:li rdf:resource="https://elevanth.org/blog/2023/07/17/none-of-the-above/"/>
	<rdf:li rdf:resource="https://hdsr.mitpress.mit.edu/pub/g9mau4m0/release/2"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2502.20755"/>
	<rdf:li rdf:resource="https://www.degruyter.com/document/isbn/9781503642102/html"/>
	<rdf:li rdf:resource="https://notstatschat.rbind.io/2025/02/26/ordinal-data-taking-transformation-invariance-seriously/"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2308.03296"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2502.17814"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2308.07037"/>
	<rdf:li rdf:resource="https://www.nber.org/papers/w32906"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2411.00247"/>
	<rdf:li rdf:resource="https://royalsocietypublishing.org/doi/10.1098/rspa.2021.0549"/>
	<rdf:li rdf:resource="https://psycnet.apa.org/record/2016-22467-001"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2306.16591"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2305.04116"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2409.18842"/>
	<rdf:li rdf:resource="https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2694998"/>
	<rdf:li rdf:resource="https://dspace.mit.edu/handle/1721.1/155358"/>
	<rdf:li rdf:resource="https://www.pnas.org/doi/10.1073/pnas.2409264121"/>
	<rdf:li rdf:resource="https://link.springer.com/article/10.3758/s13423-013-0572-3"/>
	<rdf:li rdf:resource="https://en.m.wikipedia.org/wiki/Spiders_Georg"/>
	<rdf:li rdf:resource="https://www.cambridge.org/core/journals/philosophy-of-science/article/bamboozled-by-bonferroni/0F570CAC41364F1F4D808A419A881277?WT.mc_id=New%2520Cambridge%2520Alert%2520-%2520Articles"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2404.11464"/>
	<rdf:li rdf:resource="https://link.springer.com/article/10.1007/s11186-023-09529-0"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2402.15213"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2309.10140"/>
	<rdf:li rdf:resource="http://approximateinference.org/accepted/"/>
	<rdf:li rdf:resource="https://www.nature.com/articles/s41586-023-06079-4"/>
	<rdf:li rdf:resource="https://press.princeton.edu/books/paperback/9780691222752/data-science-for-neuroimaging"/>
	<rdf:li rdf:resource="https://projecteuclid.org/journals/annals-of-statistics/volume-30/issue-1/The-screening-effect-in-Kriging/10.1214/aos/1015362194.full"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2206.06421"/>
	<rdf:li rdf:resource="https://today.yougov.com/topics/politics/articles-reports/2022/03/15/americans-misestimate-small-subgroups-population"/>
	<rdf:li rdf:resource="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5001799/"/>
	<rdf:li rdf:resource="https://doi.org/10.1080/01621459.2023.2197686"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2305.08429"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2301.13724"/>
	<rdf:li rdf:resource="https://osf.io/7vy2f/"/>
	<rdf:li rdf:resource="https://academic.oup.com/ej/article-abstract/127/605/F236/5069452?login=false"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/physics/0001019"/>
	<rdf:li rdf:resource="https://joss.theoj.org/papers/10.21105/joss.02232"/>
	<rdf:li rdf:resource="https://www.jstor.org/stable/2336834"/>
	<rdf:li rdf:resource="https://www.tandfonline.com/doi/abs/10.1080/01621459.2021.1941053"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2109.03582"/>
	<rdf:li rdf:resource="https://www.aeaweb.org/articles?id=10.1257/app.20190297"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2210.07491"/>
	<rdf:li rdf:resource="https://www.pnas.org/doi/10.1073/pnas.2203150119"/>
	<rdf:li rdf:resource="https://nap.nationalacademies.org/catalog/5499/health-effects-of-exposure-to-radon-beir-vi"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/2112.03626"/>
	<rdf:li rdf:resource="https://jmlr.org/papers/v23/20-644.html"/>
	<rdf:li rdf:resource="https://arxiv.org/abs/1811.11603"/>
	<rdf:li rdf:resource="https://www.jstor.org/stable/144855"/>
	<rdf:li rdf:resource="https://www.cambridge.org/core/books/mathematical-pictures-at-a-data-science-exhibition/0AD8668B4657A24C5135DA7DA6641B41?pageNum=2&amp;searchWithinIds=0AD8668B4657A24C5135DA7DA6641B41&amp;productType=BOOK_PART&amp;searchWithinIds=0AD8668B4657A24C5135DA7DA6641B41&amp;productType=BOOK_PART&amp;sort=mtdMetadata.bookPartMeta._mtdPositionSortable%3Aasc&amp;pageSize=30&amp;template=cambridge-core%2Fbook%2Fcontents%2Flistings&amp;ignoreExclusions=true#fndtn-information"/>
      </rdf:Seq>
    </items>
  </channel><item rdf:about="https://faculty.washington.edu/yenchic/short_note/note_MoM.pdf">
    <title>A short note on the median-of-means estimator (Yen-Chi Chen, 2020)</title>
    <dc:date>2026-04-23T16:43:42+00:00</dc:date>
    <link>https://faculty.washington.edu/yenchic/short_note/note_MoM.pdf</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA[Very nice.]]></description>
<dc:subject>to:NB have_read statistics heavy_tails estimation empirical_processes</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:ee5135168ce3/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:have_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:heavy_tails"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:estimation"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:empirical_processes"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://theoryandpractice.org/2024/10/Yes,%20we%20did%20discover%20the%20Higgs!/">
    <title>Yes, we did discover the Higgs! - Theory And Practice</title>
    <dc:date>2026-04-16T17:30:22+00:00</dc:date>
    <link>https://theoryandpractice.org/2024/10/Yes,%20we%20did%20discover%20the%20Higgs!/</link>
    <dc:creator>cshalizi</dc:creator><dc:subject>cranmer.kyle particle_physics hypothesis_testing statistics philosophy_of_science via:? sociology_of_science science_as_a_social_process have_read to:blog</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:86fa85118401/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:cranmer.kyle"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:particle_physics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:hypothesis_testing"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:philosophy_of_science"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:via:?"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:sociology_of_science"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:science_as_a_social_process"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:have_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:blog"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://www.cambridge.org/core/journals/philosophy-of-science/article/abs/carnapian-inductive-logic-for-exponential-smoothing/84DD31142459DFD0289CCE8915E79952?WT.mc_id=New%2520Cambridge%2520Alert%2520-%2520Articles">
    <title>Carnapian Inductive Logic for Exponential Smoothing | Philosophy of Science | Cambridge Core</title>
    <dc:date>2026-04-15T13:16:02+00:00</dc:date>
    <link>https://www.cambridge.org/core/journals/philosophy-of-science/article/abs/carnapian-inductive-logic-for-exponential-smoothing/84DD31142459DFD0289CCE8915E79952?WT.mc_id=New%2520Cambridge%2520Alert%2520-%2520Articles</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["This paper explores the inductive logic associated with exponential smoothing, the most widely used predictive rule that manifests the idea that more recent observations have a stronger influence on predictive probabilities than more remote ones. The main result shows that exponential smoothing can be derived from a set of plausible qualitative invariance assumptions about conditional probabilities. I discuss various aspects of the resulting inductive logic, including its connections to exchangeable processes, to Bayesian predictive inference and kernel methods in machine learning, as well as the philosophy of probabilistic invariance conditions and symmetries."]]></description>
<dc:subject>to:NB prediction statistics non-stationarity induction to_read</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:0b80819ed32c/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:prediction"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:non-stationarity"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:induction"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_read"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://www.aeaweb.org/articles?id=10.1257/aer.20240246">
    <title>Robust Misspecified Models - American Economic Association</title>
    <dc:date>2026-04-09T13:11:47+00:00</dc:date>
    <link>https://www.aeaweb.org/articles?id=10.1257/aer.20240246</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["This paper studies which misspecified models are likely to persist when decision-makers compare them with competing models. The main result characterizes such models based on two features that can be derived from primitives: The model's asymptotic accuracy in predicting the equilibrium distribution of observed outcomes and the "tightness" of the prior around such equilibria. Misspecified models can be robust, persisting against any arbitrary competing model—including the true model—despite decision-makers observing an infinite amount of data. Moreover, simple misspecified models equipped with entrenched priors can be more robust than complex correctly specified models."]]></description>
<dc:subject>decision_theory misspecification re:bayes_as_evol statistics in_NB</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:f59185400440/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:decision_theory"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:misspecification"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:re:bayes_as_evol"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:in_NB"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://www.cambridge.org/core/journals/philosophy-of-science/article/measured-inference-scales-statistics-and-scientific-inference/527F6793A1C954F01A321E72F780E931?WT.mc_id=New%2520Cambridge%2520Alert%2520-%2520Issues">
    <title>Measured Inference: Scales, Statistics, and Scientific Inference | Philosophy of Science | Cambridge Core</title>
    <dc:date>2026-04-08T16:56:58+00:00</dc:date>
    <link>https://www.cambridge.org/core/journals/philosophy-of-science/article/measured-inference-scales-statistics-and-scientific-inference/527F6793A1C954F01A321E72F780E931?WT.mc_id=New%2520Cambridge%2520Alert%2520-%2520Issues</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Despite the recent “epistemic turn” in the philosophy of measurement, philosophers have ignored a nearly 80-year controversy about the relationship between statistical inference and measurement theory. Some scholars maintain that measurement theory places no constraints on statistics, whereas others argue that the measurement scale (e.g., ordinal or interval) of one’s data determines which statistical methods are “permissible.” I defend an intermediate position: Even if existing measurement theory were irrelevant to statistical inference, it would be critical for scientific inference, which requires connecting statistical hypotheses to broader research hypotheses."]]></description>
<dc:subject>to:NB measurement philosophy_of_science statistics mayo-wilson.conor to_read</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:cb7083aa4fd5/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:measurement"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:philosophy_of_science"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:mayo-wilson.conor"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_read"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2512.24999">
    <title>[2512.24999] Basic Inequalities for First-Order Optimization with Applications to Statistical Risk Analysis</title>
    <dc:date>2026-01-07T15:13:39+00:00</dc:date>
    <link>https://arxiv.org/abs/2512.24999</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["We introduce \textit{basic inequalities} for first-order iterative optimization algorithms, forming a simple and versatile framework that connects implicit and explicit regularization. While related inequalities appear in the literature, we isolate and highlight a specific form and develop it as a well-rounded tool for statistical analysis. Let f denote the objective function to be optimized. Given a first-order iterative algorithm initialized at θ0 with current iterate θT, the basic inequality upper bounds f(θT)−f(z) for any reference point z in terms of the accumulated step sizes and the distances between θ0, θT, and z. The bound translates the number of iterations into an effective regularization coefficient in the loss function. We demonstrate this framework through analyses of training dynamics and prediction risk bounds. In addition to revisiting and refining known results on gradient descent, we provide new results for mirror descent with Bregman divergence projection, for generalized linear models trained by gradient descent and exponentiated gradient descent, and for randomized predictors. We illustrate and supplement these theoretical findings with experiments on generalized linear models."]]></description>
<dc:subject>to:NB to_read optimization statistics re:HEAS tibshirani.ryan telgarsky.matus via:mraginsky</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:00ba72cd5eb8/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:optimization"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:re:HEAS"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:tibshirani.ryan"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:telgarsky.matus"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:via:mraginsky"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2512.00175">
    <title>[2512.00175] Comparing Two Proxy Methods for Causal Identification</title>
    <dc:date>2025-12-07T15:10:34+00:00</dc:date>
    <link>https://arxiv.org/abs/2512.00175</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Identifying causal effects in the presence of unmeasured variables is a fundamental challenge in causal inference, for which proxy variable methods have emerged as a powerful solution. We contrast two major approaches in this framework: (1) bridge equation methods, which leverage solutions to integral equations to recover causal targets, and (2) array decomposition methods, which recover latent factors composing counterfactual quantities by exploiting unique determination of eigenspaces. We compare the model restrictions underlying these two approaches and provide insight into implications of the underlying assumptions, clarifying the scope of applicability for each method."]]></description>
<dc:subject>to:NB causal_inference statistics ogburn.elizabeth to_read</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:17f84a7565d6/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:causal_inference"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:ogburn.elizabeth"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_read"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2510.16174">
    <title>[2510.16174] COWs and their Hybrids: A Statistical View of Custom Orthogonal Weights</title>
    <dc:date>2025-10-24T19:41:22+00:00</dc:date>
    <link>https://arxiv.org/abs/2510.16174</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["A recurring challenge in high energy physics is inference of the signal component from a distribution for which observations are assumed to be a mixture of signal and background events. A standard assumption is that there exists information encoded in a discriminant variable that is effective at separating signal and background. This can be used to assign a signal weight to each event, with these weights used in subsequent analyses of one or more control variables of interest. The custom orthogonal weights (COWs) approach of Dembinski, et al.(2022), a generalization of the sPlot approach of Barlow (1987) and Pivk and Le Diberder (2005), is tailored to address this objective. The problem, and this method, present interesting and novel statistical issues. Here we formalize the assumptions needed and the statistical properties, while also considering extensions and alternative approaches."]]></description>
<dc:subject>to:NB classifiers hypothesis_testing statistics particle_physics kith_and_kin wasserman.larry</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:d79263d603f5/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:classifiers"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:hypothesis_testing"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:particle_physics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:kith_and_kin"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:wasserman.larry"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://www.cambridge.org/core/journals/philosophy-of-science/article/measured-inference-scales-statistics-and-scientific-inference/527F6793A1C954F01A321E72F780E931?WT.mc_id=New%2520Cambridge%2520Alert%2520-%2520Articles">
    <title>Measured Inference: Scales, Statistics, and Scientific Inference | Philosophy of Science | Cambridge Core</title>
    <dc:date>2025-09-06T15:11:03+00:00</dc:date>
    <link>https://www.cambridge.org/core/journals/philosophy-of-science/article/measured-inference-scales-statistics-and-scientific-inference/527F6793A1C954F01A321E72F780E931?WT.mc_id=New%2520Cambridge%2520Alert%2520-%2520Articles</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Despite the recent “epistemic turn” in the philosophy of measurement, philosophers have ignored a nearly 80 year controversy about the relationship between statistical inference and measurement theory. Some scholars maintain that measurement theory places no constraints on statistics, whereas others argue that the measurement scale (e.g., ordinal or interval) of one’s data determines which statistical methods are “permissible.” I defend an intermediate position: even if existing measurement theory were irrelevant to statistical inference, it would be critical for scientific inference, which requires connecting statistical hypotheses to broader research hypotheses."

]]></description>
<dc:subject>to:NB philosophy_of_science measurement statistics mayo-wilson.conor</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:4117bf85534f/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:philosophy_of_science"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:measurement"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:mayo-wilson.conor"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5152665">
    <title>The Sources of Researcher Variation in Economics by Nick Huntington-Klein, Claus C. Pörtner, Yubraj Acharya, Matus Adamkovic, Joop Adema, Lameck Ondieki Agasa, Imtiaz Ahmad, Mevlude Akbulut-Yuksel, Martin Eckhoff Andresen, David Angenendt, José-Ignacio </title>
    <dc:date>2025-09-05T16:16:09+00:00</dc:date>
    <link>https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5152665</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["We use a rigorous three-stage many-analysts design to assess how different researcher decisions—specifically data cleaning, research design, and the interpretation of a policy question—affect the variation in estimated treatment effects. A total of 146 research teams each completed the same causal inference task three times each: first with few constraints, then using a shared research design, and finally with pre-cleaned data in addition to a specified design. We find that even when analyzing the same data, teams reach different conclusions. In the first stage, the interquartile range (IQR) of the reported policy effect was 3.1 percentage points, with substantial outliers. Surprisingly, the second stage, which restricted research design choices, exhibited slightly higher IQR (4.0 percentage points), largely attributable to imperfect adherence to the prescribed protocol. By contrast, the final stage, featuring standardized data cleaning, narrowed variation in estimated effects, achieving an IQR of 2.4 percentage points. Reported sample sizes also displayed significant convergence under more restrictive conditions, with the IQR dropping from 295,187 in the first stage to 29,144 in the second, and effectively zero by the third. Our findings underscore the critical importance of data cleaning in shaping applied microeconomic results and highlight avenues for future replication efforts."]]></description>
<dc:subject>to:NB social_science_methodology statistics to_teach:undergrad-ADA to_teach:undergrad-research</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:23ac152addad/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:social_science_methodology"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach:undergrad-ADA"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach:undergrad-research"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://www.cambridge.org/core/journals/philosophy-of-science/article/logit-model-measurement-problem/4270AECB0E1E0CB6DEBE82A67A3ABD9F?WT.mc_id=New%2520Cambridge%2520Alert%2520-%2520Issues">
    <title>The Logit Model Measurement Problem | Philosophy of Science | Cambridge Core</title>
    <dc:date>2025-09-05T16:12:09+00:00</dc:date>
    <link>https://www.cambridge.org/core/journals/philosophy-of-science/article/logit-model-measurement-problem/4270AECB0E1E0CB6DEBE82A67A3ABD9F?WT.mc_id=New%2520Cambridge%2520Alert%2520-%2520Issues</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Traditional wisdom dictates that statistical model outputs are estimates, not measurements. Despite this, statistical models are employed as measurement instruments in the social sciences. In this article, I scrutinize the use of a specific model—the logit model—for psychological measurement. Given the adoption of a criterion for measurement that I call comparability, I show that the logit model fails to yield measurements due to properties that follow from its fixed residual variance."]]></description>
<dc:subject>to:NB social_science_methodology measurement logistic_regression statistics</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:814432fc462e/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:social_science_methodology"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:measurement"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:logistic_regression"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2505.13432">
    <title>[2505.13432] Synthetic-Powered Predictive Inference</title>
    <dc:date>2025-09-04T13:48:51+00:00</dc:date>
    <link>https://arxiv.org/abs/2505.13432</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Conformal prediction is a framework for predictive inference with a distribution-free, finite-sample guarantee. However, it tends to provide uninformative prediction sets when calibration data are scarce. This paper introduces Synthetic-powered predictive inference (SPI), a novel framework that incorporates synthetic data -- e.g., from a generative model -- to improve sample efficiency. At the core of our method is a score transporter: an empirical quantile mapping that aligns nonconformity scores from trusted, real data with those from synthetic data. By carefully integrating the score transporter into the calibration process, SPI provably achieves finite-sample coverage guarantees without making any assumptions about the real and synthetic data distributions. When the score distributions are well aligned, SPI yields substantially tighter and more informative prediction sets than standard conformal prediction. Experiments on image classification -- augmenting data with synthetic diffusion-model generated images -- and on tabular regression demonstrate notable improvements in predictive efficiency in data-scarce settings."]]></description>
<dc:subject>to:NB prediction conformal_prediction statistics</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:b33a182cf177/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:prediction"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:conformal_prediction"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2412.20355">
    <title>[2412.20355] Confidence Interval Construction and Conditional Variance Estimation with Dense ReLU Networks</title>
    <dc:date>2025-09-04T13:48:03+00:00</dc:date>
    <link>https://arxiv.org/abs/2412.20355</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["This paper addresses the problems of conditional variance estimation and confidence interval construction in nonparametric regression using dense networks with the Rectified Linear Unit (ReLU) activation function. We present a residual-based framework for conditional variance estimation, deriving nonasymptotic bounds for variance estimation under both heteroscedastic and homoscedastic settings. We relax the sub-Gaussian noise assumption, allowing the proposed bounds to accommodate sub-Exponential noise and beyond. Building on this, for a ReLU neural network estimator, we derive non-asymptotic bounds for both its conditional mean and variance estimation, representing the first result for variance estimation using ReLU networks. Furthermore, we develop a ReLU network based robust bootstrap procedure (Efron, 1992) for constructing confidence intervals for the true mean that comes with a theoretical guarantee on the coverage, providing a significant advancement in uncertainty quantification and the construction of reliable confidence intervals in deep learning settings."]]></description>
<dc:subject>to:NB statistics confidence_sets neural_networks regression uncertainty_for_neural_networks</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:edb0cef1f9c4/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:confidence_sets"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:neural_networks"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:regression"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:uncertainty_for_neural_networks"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://www.cambridge.org/us/universitypress/subjects/physics/mathematical-methods/probability-theory-quantitative-scientists?format=HB&amp;WT.mc_id=NGV_IOC_BC_BK%253b_PHYS%253b_Probability%2BTheory%2Bfor%2BQuantitative%2BScientists_Jul25">
    <title>Probability Theory for Quantitative Scientists | Cambridge University Press &amp; Assessment</title>
    <dc:date>2025-08-16T23:02:29+00:00</dc:date>
    <link>https://www.cambridge.org/us/universitypress/subjects/physics/mathematical-methods/probability-theory-quantitative-scientists?format=HB&amp;WT.mc_id=NGV_IOC_BC_BK%253b_PHYS%253b_Probability%2BTheory%2Bfor%2BQuantitative%2BScientists_Jul25</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Based on the long-running Probability Theory course at the Sapienza University of Rome, this book offers a fresh and in-depth approach to probability and statistics, while remaining intuitive and accessible in style. The fundamentals of probability theory are elegantly presented, supported by numerous examples and illustrations, and modern applications are later introduced giving readers an appreciation of current research topics. The text covers distribution functions, statistical inference and data analysis, and more advanced methods including Markov chains and Poisson processes, widely used in dynamical systems and data science research. The concluding section, 'Entropy, Probability and Statistical Mechanics' unites key concepts from the text with the authors' impressive research experience, to provide a clear illustration of these powerful statistical tools in action. Ideal for students and researchers in the quantitative sciences this book provides an authoritative account of probability theory, written by leading researchers in the field."

--- If Diaconis calls your probability textbook "mind-blowing", the rest of us should at least check it out...]]></description>
<dc:subject>to:NB books:noted probability statistics stochastic_processes statistical_mechanics of_course_its_really_a_spin_glass</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:62d2299db5f4/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:books:noted"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:probability"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:stochastic_processes"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistical_mechanics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:of_course_its_really_a_spin_glass"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://imstat.org/2017/04/01/obituary-ulf-grenander-1923-2016/">
    <title>Institute of Mathematical Statistics | Obituary: Ulf Grenander, 1923–2016</title>
    <dc:date>2025-07-09T18:14:55+00:00</dc:date>
    <link>https://imstat.org/2017/04/01/obituary-ulf-grenander-1923-2016/</link>
    <dc:creator>cshalizi</dc:creator><dc:subject>have_read to:NB lives_of_the_scientists grenander.ulf statistics</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:3cc0fdc69628/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:have_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:lives_of_the_scientists"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:grenander.ulf"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2504.03503">
    <title>[2504.03503] Operator Learning: A Statistical Perspective</title>
    <dc:date>2025-04-28T01:23:12+00:00</dc:date>
    <link>https://arxiv.org/abs/2504.03503</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Operator learning has emerged as a powerful tool in scientific computing for approximating mappings between infinite-dimensional function spaces. A primary application of operator learning is the development of surrogate models for the solution operators of partial differential equations (PDEs). These methods can also be used to develop black-box simulators to model system behavior from experimental data, even without a known mathematical model. In this article, we begin by formalizing operator learning as a function-to-function regression problem and review some recent developments in the field. We also discuss PDE-specific operator learning, outlining strategies for incorporating physical and mathematical constraints into architecture design and training processes. Finally, we end by highlighting key future directions such as active data collection and the development of rigorous uncertainty quantification frameworks."]]></description>
<dc:subject>to:NB approximation tewari.ambuj statistics equations_of_motion_from_a_time_series</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:ad7e303203c2/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:approximation"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:tewari.ambuj"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:equations_of_motion_from_a_time_series"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2503.15850">
    <title>[2503.15850] Uncertainty Quantification and Confidence Calibration in Large Language Models: A Survey</title>
    <dc:date>2025-04-23T14:56:40+00:00</dc:date>
    <link>https://arxiv.org/abs/2503.15850</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Large Language Models (LLMs) excel in text generation, reasoning, and decision-making, enabling their adoption in high-stakes domains such as healthcare, law, and transportation. However, their reliability is a major concern, as they often produce plausible but incorrect responses. Uncertainty quantification (UQ) enhances trustworthiness by estimating confidence in outputs, enabling risk mitigation and selective prediction. However, traditional UQ methods struggle with LLMs due to computational constraints and decoding inconsistencies. Moreover, LLMs introduce unique uncertainty sources, such as input ambiguity, reasoning path divergence, and decoding stochasticity, that extend beyond classical aleatoric and epistemic uncertainty. To address this, we introduce a new taxonomy that categorizes UQ methods based on computational efficiency and uncertainty dimensions (input, reasoning, parameter, and prediction uncertainty). We evaluate existing techniques, assess their real-world applicability, and identify open challenges, emphasizing the need for scalable, interpretable, and robust UQ approaches to enhance LLM reliability."]]></description>
<dc:subject>in_NB statistics large_language_models_(so_called) uncertainty_for_neural_networks to_teach:statistics_and_generative_ai</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:7758c0625db0/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:in_NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:large_language_models_(so_called)"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:uncertainty_for_neural_networks"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach:statistics_and_generative_ai"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2401.15800">
    <title>[2401.15800] Statistical Significance of Feature Importance Rankings</title>
    <dc:date>2025-03-31T23:37:55+00:00</dc:date>
    <link>https://arxiv.org/abs/2401.15800</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Feature importance scores are ubiquitous tools for understanding the predictions of machine learning models. However, many popular attribution methods suffer from high instability due to random sampling. Leveraging novel ideas from hypothesis testing, we devise techniques that ensure the most important features are correct with high-probability guarantees. These assess the set of K top-ranked features, as well as the order of its elements. Given a set of local or global importance scores, we demonstrate how to retrospectively verify the stability of the highest ranks. We then introduce two efficient sampling algorithms that identify the K most important features, perhaps in order, with probability exceeding 1−α. The theoretical justification for these procedures is validated empirically on SHAP and LIME."]]></description>
<dc:subject>to:NB statistics variable_selection hooker.giles</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:a0e052646ae3/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:variable_selection"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:hooker.giles"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2501.07772">
    <title>[2501.07772] Bridging Root-$n$ and Non-standard Asymptotics: Dimension-agnostic Adaptive Inference in M-Estimation</title>
    <dc:date>2025-03-24T00:12:30+00:00</dc:date>
    <link>https://arxiv.org/abs/2501.07772</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["This manuscript studies a general approach to construct confidence sets for the solution of population-level optimization, commonly referred to as M-estimation. Statistical inference for M-estimation poses significant challenges due to the non-standard limiting behaviors of the corresponding estimator, which arise in settings with increasing dimension of parameters, non-smooth objectives, or constraints. We propose a simple and unified method that guarantees validity in both regular and irregular cases. Moreover, we provide a comprehensive width analysis of the proposed confidence set, showing that the convergence rate of the diameter is adaptive to the unknown degree of instance-specific regularity. We apply the proposed method to several high-dimensional and irregular statistical problems."]]></description>
<dc:subject>to:NB statistics estimation kuchibhotla.arun_kmar re:HEAS via:lal.apoorva</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:493622f19ddd/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:estimation"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:kuchibhotla.arun_kmar"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:re:HEAS"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:via:lal.apoorva"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://elevanth.org/blog/2023/07/17/none-of-the-above/">
    <title>None of the Above | Elements of Evolutionary Anthropology</title>
    <dc:date>2025-03-23T17:17:45+00:00</dc:date>
    <link>https://elevanth.org/blog/2023/07/17/none-of-the-above/</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA[--- If I am honest with myself, incorporating something like this (or even my own paper with Gelman!) into undergrad ADA would require a big re-design of the course, because it's currently "here is an array of sometimes-useful statistical methods", not "here is how you turn scientific questions into data-analytic problems, and statistical solutions back into scientific answers".  Knowing a lot of methods is _helpful_ to that undertaking, but it's different.  Maybe that's too much to ask of an undergrad class with >200 students/year...]]></description>
<dc:subject>statistics data_analysis have_read mcelreath.richard closing_old_tabs re:phil-of-bayes_paper to_teach:undergrad-ADA</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:1079612a0b01/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:data_analysis"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:have_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:mcelreath.richard"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:closing_old_tabs"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:re:phil-of-bayes_paper"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach:undergrad-ADA"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://hdsr.mitpress.mit.edu/pub/g9mau4m0/release/2">
    <title>Data Science at the Singularity · Issue 6.1, Winter 2024</title>
    <dc:date>2025-03-17T00:44:57+00:00</dc:date>
    <link>https://hdsr.mitpress.mit.edu/pub/g9mau4m0/release/2</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Something fundamental to computation-based research has really changed in the last 10 years. In certain fields, progress is simply dramatically faster than ever. Researchers in affected fields are living through a period of profound transformation as the fields undergo a transition to frictionless reproducibility (FR). This transition markedly changes the rate at which ideas and practices spread, affects scientific mindsets and the goals of science, and erases memories of much that came before.
"The emergence of FR flows from three data science principles that matured together after decades of work by many technologists and numerous research communities. The mature principles involve data sharing, code sharing, and competitive challenges, however implemented in the particularly strong form of frictionless open services.
"Empirical machine learning is today’s leading adherent field; its hidden superpower is adherence to frictionless reproducibility practices; these practices are responsible for the striking and surprising progress in AI that we see everywhere; and these practices can be learned and adhered to by researchers in any research field, automatically increasing the rate of progress in each adherent field."]]></description>
<dc:subject>to:NB statistics machine_learning methodological_advice donoho.david</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:cc9f3e206846/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:machine_learning"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:methodological_advice"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:donoho.david"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2502.20755">
    <title>[2502.20755] Minimax Optimal Kernel Two-Sample Tests with Random Features</title>
    <dc:date>2025-03-16T19:31:58+00:00</dc:date>
    <link>https://arxiv.org/abs/2502.20755</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Reproducing Kernel Hilbert Space (RKHS) embedding of probability distributions has proved to be an effective approach, via MMD (maximum mean discrepancy) for nonparametric hypothesis testing problems involving distributions defined over general (non-Euclidean) domains. While a substantial amount of work has been done on this topic, only recently, minimax optimal two-sample tests have been constructed that incorporate, unlike MMD, both the mean element and a regularized version of the covariance operator. However, as with most kernel algorithms, the computational complexity of the optimal test scales cubically in the sample size, limiting its applicability. In this paper, we propose a spectral regularized two-sample test based on random Fourier feature (RFF) approximation and investigate the trade-offs between statistical optimality and computational efficiency. We show the proposed test to be minimax optimal if the approximation order of RFF (which depends on the smoothness of the likelihood ratio and the decay rate of the eigenvalues of the integral operator) is sufficiently large. We develop a practically implementable permutation-based version of the proposed test with a data-adaptive strategy for selecting the regularization parameter and the kernel. Finally, through numerical experiments on simulated and benchmark datasets, we demonstrate that the proposed RFF-based test is computationally efficient and performs almost similar (with a small drop in power) to the exact test."]]></description>
<dc:subject>to:NB hilbert_space statistics two-sample_tests random_features</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:e765167ec668/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:hilbert_space"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:two-sample_tests"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:random_features"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://www.degruyter.com/document/isbn/9781503642102/html">
    <title>Thinking Through Data: How Outliers, Aggregates, and Patterns Shape Perception</title>
    <dc:date>2025-03-14T17:49:16+00:00</dc:date>
    <link>https://www.degruyter.com/document/isbn/9781503642102/html</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["We encounter digital data processing on a range of platforms and in a multitude of contexts today: in the predictive algorithms of the financial sector, in drones, insurance, and risk management, in smart cities, biometrics, medicine, and more. This fascinating book explores the historical context of the current data-driven paradigm and explains how elusive yet crucial statistical concepts such as outliers, aggregates, and patterns form how we sense and make sense of data. From the sixteenth century's embodied measurements of the foot, through the blurred facial features of L'Homme Moyen, to the image aggregates of today's security systems, the examples collected in this book illustrate the central role of aesthetics throughout the history of statistical knowledge production. Taking its point of departure in analyses and discussions of contemporary artistic experiments by Rossella Biscotti, Stéphanie Solinas, and Adam Broomberg and Oliver Chanarin, the book broadens our understanding of the structures of knowledge and methods in statistical computation beyond optimistic narratives of calculative power. Venturing out into the tails of the distributions—to the systemically overlooked and excluded—this book challenges us to embrace an alternative view of modern data processing."

--- Sounds like "baby stats. for the Theory-brained, via pictures", but I actually think that would be amazing so I hope that's what this is and it's good. 
]]></description>
<dc:subject>to:NB books:noted statistics aesthetics history_of_statistics visual_display_of_quantitative_information downloaded</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:adb95a135c62/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:books:noted"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:aesthetics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:history_of_statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:visual_display_of_quantitative_information"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:downloaded"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://notstatschat.rbind.io/2025/02/26/ordinal-data-taking-transformation-invariance-seriously/">
    <title>Ordinal data: taking transformation invariance seriously - Biased and Inefficient</title>
    <dc:date>2025-03-11T15:44:29+00:00</dc:date>
    <link>https://notstatschat.rbind.io/2025/02/26/ordinal-data-taking-transformation-invariance-seriously/</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Rank tests (for data without ties, at least) are distribution-free in that you use the same numerical scores (the ranks) regardless of the actual data values. They aren’t ordinal, because they use numerical scores (the ranks) and do arithmetic on these scores. There’s nothing magical about ranks that lets them get around the invariance requirement; they just don’t satisfy it.
"If you really want to take the transformation-invariance of ordinal data seriously, you need either to assume (and check) a data generating model that guarantees the true CDFs will never cross, or get used to most two-sample comparisons being undefined."]]></description>
<dc:subject>statistics measurement lumley.thomas</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:191b567df688/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:measurement"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:lumley.thomas"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2308.03296">
    <title>[2308.03296] Studying Large Language Model Generalization with Influence Functions</title>
    <dc:date>2025-03-10T15:14:03+00:00</dc:date>
    <link>https://arxiv.org/abs/2308.03296</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["When trying to gain better visibility into a machine learning model in order to understand and mitigate the associated risks, a potentially valuable source of evidence is: which training examples most contribute to a given behavior? Influence functions aim to answer a counterfactual: how would the model's parameters (and hence its outputs) change if a given sequence were added to the training set? While influence functions have produced insights for small models, they are difficult to scale to large language models (LLMs) due to the difficulty of computing an inverse-Hessian-vector product (IHVP). We use the Eigenvalue-corrected Kronecker-Factored Approximate Curvature (EK-FAC) approximation to scale influence functions up to LLMs with up to 52 billion parameters. In our experiments, EK-FAC achieves similar accuracy to traditional influence function estimators despite the IHVP computation being orders of magnitude faster. We investigate two algorithmic techniques to reduce the cost of computing gradients of candidate training sequences: TF-IDF filtering and query batching. We use influence functions to investigate the generalization patterns of LLMs, including the sparsity of the influence patterns, increasing abstraction with scale, math and programming abilities, cross-lingual generalization, and role-playing behavior. Despite many apparently sophisticated forms of generalization, we identify a surprising limitation: influences decay to near-zero when the order of key phrases is flipped. Overall, influence functions give us a powerful new tool for studying the generalization properties of LLMs."]]></description>
<dc:subject>in_NB have_read optimization statistics computational_statistics neural_networks large_language_models_(so_called) estimation re:large_language_models_in_statistical_perspective feral_library_catalogs re:gopnikism to_teach:statistics_and_generative_ai</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:3b2888a6a8c1/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:in_NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:have_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:optimization"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:computational_statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:neural_networks"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:large_language_models_(so_called)"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:estimation"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:re:large_language_models_in_statistical_perspective"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:feral_library_catalogs"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:re:gopnikism"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach:statistics_and_generative_ai"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2502.17814">
    <title>[2502.17814] An Overview of Large Language Models for Statisticians</title>
    <dc:date>2025-03-10T15:03:46+00:00</dc:date>
    <link>https://arxiv.org/abs/2502.17814</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Large Language Models (LLMs) have emerged as transformative tools in artificial intelligence (AI), exhibiting remarkable capabilities across diverse tasks such as text generation, reasoning, and decision-making. While their success has primarily been driven by advances in computational power and deep learning architectures, emerging problems -- in areas such as uncertainty quantification, decision-making, causal inference, and distribution shift -- require a deeper engagement with the field of statistics. This paper explores potential areas where statisticians can make important contributions to the development of LLMs, particularly those that aim to engender trustworthiness and transparency for human users. Thus, we focus on issues such as uncertainty quantification, interpretability, fairness, privacy, watermarking and model adaptation. We also consider possible roles for LLMs in statistical analysis. By bridging AI and statistics, we aim to foster a deeper collaboration that advances both the theoretical foundations and practical applications of LLMs, ultimately shaping their role in addressing complex societal challenges."

--- From an initial scan, I find this not very interesting, and so (given some of the names on the masthead) rather disappointing.  They're parameteric high-order Markov chains based on kernel smoothing!  We should be able to say something more interesting than "you can find examples on HuggingFace" and "sometimes they memorize personally identifiable information, which is naughty".]]></description>
<dc:subject>to:NB to_read have_skimmed large_language_models_(so_called) statistics re:large_language_models_in_statistical_perspective</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:0e6beffb3457/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:have_skimmed"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:large_language_models_(so_called)"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:re:large_language_models_in_statistical_perspective"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2308.07037">
    <title>[2308.07037] Bayesian Flow Networks</title>
    <dc:date>2025-03-10T13:55:12+00:00</dc:date>
    <link>https://arxiv.org/abs/2308.07037</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["This paper introduces Bayesian Flow Networks (BFNs), a new class of generative model in which the parameters of a set of independent distributions are modified with Bayesian inference in the light of noisy data samples, then passed as input to a neural network that outputs a second, interdependent distribution. Starting from a simple prior and iteratively updating the two distributions yields a generative procedure similar to the reverse process of diffusion models; however it is conceptually simpler in that no forward process is required. Discrete and continuous-time loss functions are derived for continuous, discretised and discrete data, along with sample generation procedures. Notably, the network inputs for discrete data lie on the probability simplex, and are therefore natively differentiable, paving the way for gradient-based sample guidance and few-step generation in discrete domains such as language modelling. The loss function directly optimises data compression and places no restrictions on the network architecture. In our experiments BFNs achieve competitive log-likelihoods for image modelling on dynamically binarized MNIST and CIFAR-10, and outperform all known discrete diffusion models on the text8 character-level language modelling task."]]></description>
<dc:subject>to:NB graphical_models neural_networks simulation statistics computational_statistics to_read</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:4f89537c0697/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:graphical_models"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:neural_networks"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:simulation"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:computational_statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_read"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://www.nber.org/papers/w32906">
    <title>Adapting to Misspecification | NBER</title>
    <dc:date>2025-03-08T21:25:44+00:00</dc:date>
    <link>https://www.nber.org/papers/w32906</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Empirical research typically involves a robustness-efficiency tradeoff. A researcher seeking to estimate a scalar parameter can invoke strong assumptions to motivate a restricted estimator that is precise but may be heavily biased, or they can relax some of these assumptions to motivate a more robust, but variable, unrestricted estimator. When a bound on the bias of the restricted estimator is available, it is optimal to shrink the unrestricted estimator towards the restricted estimator. For settings where a bound on the bias of the restricted estimator is unknown, we propose adaptive estimators that minimize the percentage increase in worst case risk relative to an oracle that knows the bound. We show that adaptive estimators solve a weighted convex minimax problem and provide lookup tables facilitating their rapid computation. Revisiting some well known empirical studies where questions of model specification arise, we examine the advantages of adapting to—rather than testing for—misspecification."

--- TODO: See if they mention Hjort & Claesken's "focused information criterion", which is at least similar in inspiration; re-read H&C on this point.]]></description>
<dc:subject>to:NB estimation misspecification statistics</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:87581585e77d/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:estimation"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:misspecification"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2411.00247">
    <title>[2411.00247] Deep Learning Through A Telescoping Lens: A Simple Model Provides Empirical Insights On Grokking, Gradient Boosting &amp; Beyond</title>
    <dc:date>2025-01-22T15:36:23+00:00</dc:date>
    <link>https://arxiv.org/abs/2411.00247</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Deep learning sometimes appears to work in unexpected ways. In pursuit of a deeper understanding of its surprising behaviors, we investigate the utility of a simple yet accurate model of a trained neural network consisting of a sequence of first-order approximations telescoping out into a single empirically operational tool for practical analysis. Across three case studies, we illustrate how it can be applied to derive new empirical insights on a diverse range of prominent phenomena in the literature -- including double descent, grokking, linear mode connectivity, and the challenges of applying deep learning on tabular data -- highlighting that this model allows us to construct and extract metrics that help predict and understand the a priori unexpected performance of neural networks. We also demonstrate that this model presents a pedagogical formalism allowing us to isolate components of the training process even in complex contemporary settings, providing a lens to reason about the effects of design choices such as architecture & optimization strategy, and reveals surprising parallels between neural network learning and gradient boosting."]]></description>
<dc:subject>to:NB neural_networks learning_theory statistics to_read via:mraginsky to_teach:statistics_and_generative_ai</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:b500158ab0cc/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:neural_networks"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:learning_theory"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:via:mraginsky"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach:statistics_and_generative_ai"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://royalsocietypublishing.org/doi/10.1098/rspa.2021.0549">
    <title>USP: an independence test that improves on Pearson’s chi-squared and the G-test | Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences</title>
    <dc:date>2025-01-10T15:06:42+00:00</dc:date>
    <link>https://royalsocietypublishing.org/doi/10.1098/rspa.2021.0549</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["We present the $U$-statistic permutation (USP) test of independence in the context of discrete data displayed in a contingency table. Either Pearson’s $\chi^2$-test of independence, or the $G$-test, are typically used for this task, but we argue that these tests have serious deficiencies, both in terms of their inability to control the size of the test, and their power properties. By contrast, the USP test is guaranteed to control the size of the test at the nominal level for all sample sizes, has no issues with small (or zero) cell counts, and is able to detect distributions that violate independence in only a minimal way. The test statistic is derived from a $U$-statistic estimator of a natural population measure of dependence, and we prove that this is the unique minimum variance unbiased estimator of this population quantity. The practical utility of the USP test is demonstrated on both simulated data, where its power can be dramatically greater than those of Pearson’s test, the $G$-test and Fisher’s exact test, and on real data. The USP test is implemented in the R package USP."]]></description>
<dc:subject>to:NB dependence_measures hypothesis_testing independence_testing statistics samworth.richard_j.</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:0861a549ae01/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:dependence_measures"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:hypothesis_testing"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:independence_testing"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:samworth.richard_j."/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://psycnet.apa.org/record/2016-22467-001">
    <title>On the unnecessary ubiquity of hierarchical linear modeling.</title>
    <dc:date>2025-01-08T14:30:08+00:00</dc:date>
    <link>https://psycnet.apa.org/record/2016-22467-001</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["In psychology and the behavioral sciences generally, the use of the hierarchical linear model (HLM) and its extensions for discrete outcomes are popular methods for modeling clustered data. HLM and its discrete outcome extensions, however, are certainly not the only methods available to model clustered data. Although other methods exist and are widely implemented in other disciplines, it seems that psychologists have yet to consider these methods in substantive studies. This article compares and contrasts HLM with alternative methods including generalized estimating equations and cluster-robust standard errors. These alternative methods do not model random effects and thus make a smaller number of assumptions and are interpreted identically to single-level methods with the benefit that estimates are adjusted to reflect clustering of observations. Situations where these alternative methods may be advantageous are discussed including research questions where random effects are and are not required, when random effects can change the interpretation of regression coefficients, challenges of modeling with random effects with discrete outcomes, and examples of published psychology articles that use HLM that may have benefitted from using alternative methods. Illustrative examples are provided and discussed to demonstrate the advantages of the alternative methods and also when HLM would be the preferred method. "]]></description>
<dc:subject>to:NB regression to_teach:linear_models psychology statistics</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:0afe5c9a51c9/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:regression"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach:linear_models"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:psychology"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2306.16591">
    <title>[2306.16591] Nonparametric Causal Decomposition of Group Disparities</title>
    <dc:date>2025-01-06T14:48:19+00:00</dc:date>
    <link>https://arxiv.org/abs/2306.16591</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["We introduce a new nonparametric causal decomposition approach that identifies the mechanisms by which a treatment variable contributes to a group-based outcome disparity. Our approach distinguishes three mechanisms: group differences in 1) treatment prevalence, 2) average treatment effects, and 3) selection into treatment based on individual-level treatment effects. Our approach reformulates classic Kitagawa-Blinder-Oaxaca decompositions in causal and nonparametric terms, complements causal mediation analysis by explaining group disparities instead of group effects, and isolates conceptually distinct mechanisms conflated in recent random equalization decompositions. In contrast to all prior approaches, our framework uniquely identifies differential selection into treatment as a novel disparity-generating mechanism. Our approach can be used for both the retrospective causal explanation of disparities and the prospective planning of interventions to change disparities. We present both an unconditional and a conditional decomposition, where the latter quantifies the contributions of the treatment within levels of certain covariates. We develop nonparametric estimators that are n‾√-consistent, asymptotically normal, semiparametrically efficient, and multiply robust. We apply our approach to analyze the mechanisms by which college graduation causally contributes to intergenerational income persistence (the disparity in adult income between the children of high- vs low-income parents). Empirically, we demonstrate a previously undiscovered role played by the new selection component in intergenerational income persistence."]]></description>
<dc:subject>to:NB to_read have_skimmed causal_inference statistics elwert.felix to_teach:statistics_of_inequality_and_discrimination sds_icsd_search</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:3123a651d148/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:have_skimmed"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:causal_inference"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:elwert.felix"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach:statistics_of_inequality_and_discrimination"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:sds_icsd_search"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2305.04116">
    <title>[2305.04116] The Fundamental Limits of Structure-Agnostic Functional Estimation</title>
    <dc:date>2024-12-11T19:37:38+00:00</dc:date>
    <link>https://arxiv.org/abs/2305.04116</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Many recent developments in causal inference, and functional estimation problems more generally, have been motivated by the fact that classical one-step (first-order) debiasing methods, or their more recent sample-split double machine-learning avatars, can outperform plugin estimators under surprisingly weak conditions. These first-order corrections improve on plugin estimators in a black-box fashion, and consequently are often used in conjunction with powerful off-the-shelf estimation methods. These first-order methods are however provably suboptimal in a minimax sense for functional estimation when the nuisance functions live in Holder-type function spaces. This suboptimality of first-order debiasing has motivated the development of "higher-order" debiasing methods. The resulting estimators are, in some cases, provably optimal over Holder-type spaces, but both the estimators which are minimax-optimal and their analyses are crucially tied to properties of the underlying function space.
"In this paper we investigate the fundamental limits of structure-agnostic functional estimation, where relatively weak conditions are placed on the underlying nuisance functions. We show that there is a strong sense in which existing first-order methods are optimal. We achieve this goal by providing a formalization of the problem of functional estimation with black-box nuisance function estimates, and deriving minimax lower bounds for this problem. Our results highlight some clear tradeoffs in functional estimation -- if we wish to remain agnostic to the underlying nuisance function spaces, impose only high-level rate conditions, and maintain compatibility with black-box nuisance estimators then first-order methods are optimal. When we have an understanding of the structure of the underlying nuisance functions then carefully constructed higher-order estimators can outperform first-order estimators."]]></description>
<dc:subject>to:NB to_read statistics nonparametrics entropy_estimation kith_and_kin kennedy.edward_h. wasserman.larry balakrishnan.sivaraman causal_inference</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:0ee8eb04a835/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:nonparametrics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:entropy_estimation"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:kith_and_kin"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:kennedy.edward_h."/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:wasserman.larry"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:balakrishnan.sivaraman"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:causal_inference"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2409.18842">
    <title>[2409.18842] Classical Statistical (In-Sample) Intuitions Don't Generalize Well: A Note on Bias-Variance Tradeoffs, Overfitting and Moving from Fixed to Random Designs</title>
    <dc:date>2024-12-11T16:38:39+00:00</dc:date>
    <link>https://arxiv.org/abs/2409.18842</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["The sudden appearance of modern machine learning (ML) phenomena like double descent and benign overfitting may leave many classically trained statisticians feeling uneasy -- these phenomena appear to go against the very core of statistical intuitions conveyed in any introductory class on learning from data. The historical lack of earlier observation of such phenomena is usually attributed to today's reliance on more complex ML methods, overparameterization, interpolation and/or higher data dimensionality. In this note, we show that there is another reason why we observe behaviors today that appear at odds with intuitions taught in classical statistics textbooks, which is much simpler to understand yet rarely discussed explicitly. In particular, many intuitions originate in fixed design settings, in which in-sample prediction error (under resampling of noisy outcomes) is of interest, while modern ML evaluates its predictions in terms of generalization error, i.e. out-of-sample prediction error in random designs. Here, we highlight that this simple move from fixed to random designs has (perhaps surprisingly) far-reaching consequences on textbook intuitions relating to the bias-variance tradeoff, and comment on the resulting (im)possibility of observing double descent and benign overfitting in fixed versus random designs."]]></description>
<dc:subject>regression statistics interpolation_aka_memorizing_the_training_data in_NB</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:bee6248ae087/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:regression"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:interpolation_aka_memorizing_the_training_data"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:in_NB"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2694998">
    <title>Specification Curve: Descriptive and Inferential Statistics on All Reasonable Specifications by Uri Simonsohn, Joseph P. Simmons, Leif D. Nelson :: SSRN</title>
    <dc:date>2024-12-09T21:26:14+00:00</dc:date>
    <link>https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2694998</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Empirical results often hinge on data analytic decisions that are simultaneously defensible, arbitrary, and motivated. To mitigate this problem we introduce Specification-Curve Analysis, which consists of three steps: (i) identifying the set of theoretically justified, statistically valid, and non-redundant analytic specifications, (ii) displaying alternative results graphically, allowing the identification of decisions producing different results, and (iii) conducting statistical tests to determine whether as a whole results are inconsistent with the null hypothesis. We illustrate its use by applying it to three published findings. One proves robust, one weak, one not robust at all."

--- Item (1) seems like a tall order!]]></description>
<dc:subject>to:NB data_analysis statistics model_checking color_me_skeptical more_exactly_the_impulse_is_sound_but_i_doubt_they_can_really_do_it</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:888da2dd46e4/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:data_analysis"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:model_checking"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:color_me_skeptical"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:more_exactly_the_impulse_is_sound_but_i_doubt_they_can_really_do_it"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://dspace.mit.edu/handle/1721.1/155358">
    <title>Likelihood-Free Hypothesis Testing and Applications of the Energy Distance</title>
    <dc:date>2024-12-06T14:04:25+00:00</dc:date>
    <link>https://dspace.mit.edu/handle/1721.1/155358</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["This thesis studies questions in nonparametric testing and estimation that are inspired by machine learning. One of the main problems of our interest is likelihood-free hypothesis testing: given three samples X, Y and Z with sample sizes n, n and m respectively, one must decide whether the distribution of Z is closer to that of X or that of Y . We fully characterize the problem’s sample complexity for multiple distribution classes and with high probability. We uncover connections to two-sample, goodness-of-fit and robust testing, and show the existence of a trade-off of the form mn ≍ k/ε^4, where k is an appropriate notion of complexity and ε is the total variation separation between the distributions of X and Y . We generalize our problem to allow Z to come from a mixture of the distributions of X and Y , and propose a kernel-based test for its solution, and also verify the existence of a trade-off between m and n on experimental data from particle physics. In addition, we demonstrate that the family of “classifier accuracy” tests are not only popular in practice but also provably near-optimal, recovering and simplifying a multitude of classical and recent results. Finally, we study affine classifiers as a tool for estimation and testing, with the key technical tool being a connection to the energy distance. In particular, we propose a density estimation routine based on minimizing the generalized energy distance, targeting smooth densities and Gaussian mixtures. We interpret our results in terms of half-space separability over these classes, and derive analogous results for discrete distributions. As a consequence we deduce that any two discrete distributions are well-separated by a half-space, provided their support is embedded as a packing of a high-dimensional unit ball. We also scrutinize two recent applications of the energy distance in the two-sample testing literature."
]]></description>
<dc:subject>to:NB to_read hypothesis_testing two-sample_tests statistics via:_onionesque kernel_methods goodness-of-fit</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:d56e7c266c6a/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:hypothesis_testing"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:two-sample_tests"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:via:_onionesque"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:kernel_methods"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:goodness-of-fit"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://www.pnas.org/doi/10.1073/pnas.2409264121">
    <title>Physician–patient racial concordance and newborn mortality | PNAS</title>
    <dc:date>2024-11-28T00:40:46+00:00</dc:date>
    <link>https://www.pnas.org/doi/10.1073/pnas.2409264121</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["The racial gap in infant mortality is a pressing public-health concern, and [B. N. Greenwood et al., Proc. Natl. Acad. Sci. U.S.A. 117, 21194–21200 (2020), https://doi.org/10.1073/pnas.1913405117] suggest that Black newborns are more likely to survive if cared for by Black physicians after birth, even in models that control for numerous variables, including hospital and physician fixed effects, and the 65 most common comorbidities affecting newborns (as described by International Classification of Disease codes). We acquired the data used in the study, covering Florida hospital discharges from 1992 through the third quarter of 2015, to replicate and extend the analysis. We find that the magnitude of the concordance effect is substantially reduced after controlling for diagnoses indicating very low birth weight (<1,500 g), which are a strong predictor of neonatal mortality but not among the 65 most common comorbidities. In fact, the estimated effect is near zero and statistically insignificant in the expanded specifications that control for very low birth weight and include hospital and physician fixed effects."

--- This is extremely persuasive.  The one escape hatch I can see is if black mothers getting white doctors somehow leads to low birth weight.  This seems very implausible, because this "assignment" of doctors to mothers seems to happen right at going to the hospital for birth, or at best very shortly before.  (I.e., it's not the mother's regular obstetrician, assuming she has one.)

(My recollection is that Borjas has some unfortunate political connections, but that's a secondary issue.  Just look at figures 1 and 2!)
]]></description>
<dc:subject>statistics medicine to_teach:statistics_of_inequality_and_discrimination have_read</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:4cef516c110a/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:medicine"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach:statistics_of_inequality_and_discrimination"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:have_read"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://link.springer.com/article/10.3758/s13423-013-0572-3">
    <title>Robust misinterpretation of confidence intervals | Psychonomic Bulletin &amp; Review</title>
    <dc:date>2024-07-11T14:03:21+00:00</dc:date>
    <link>https://link.springer.com/article/10.3758/s13423-013-0572-3</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Null hypothesis significance testing (NHST) is undoubtedly the most common inferential technique used to justify claims in the social sciences. However, even staunch defenders of NHST agree that its outcomes are often misinterpreted. Confidence intervals (CIs) have frequently been proposed as a more useful alternative to NHST, and their use is strongly encouraged in the APA Manual. Nevertheless, little is known about how researchers interpret CIs. In this study, 120 researchers and 442 students—all in the field of psychology—were asked to assess the truth value of six particular statements involving different interpretations of a CI. Although all six statements were false, both researchers and students endorsed, on average, more than three statements, indicating a gross misunderstanding of CIs. Self-declared experience with statistics was not related to researchers’ performance, and, even more surprisingly, researchers hardly outperformed the students, even though the students had not received any education on statistical inference whatsoever. Our findings suggest that many researchers do not know the correct interpretation of a CI. The misunderstandings surrounding p-values and CIs are particularly unfortunate because they constitute the main tools by which psychologists draw conclusions from data."

--- After skimming: On the one hand, I like a good "psychologists really do not understand anything about statistics" story as much as the next statistician, and I dare say more than most of my colleagues.

OTOH, I have my doubts about these survey items. Number 3 ("the hypothesis that the true mean is zero is likely to be incorrect") and number 5 ("we can be 95% confident that the true mean is between 0.1 and 0.4") are, IMHO, at least arguably true!  For #3: A parameter value falls outside a level-alpha confidence set iff the corresponding test rejects that parameter value with a p-value of at most 1-alpha, so 0 is rejected by this test, whatever it is, at the conventional 5% level.  Glossing "rejected by a reliable test" as "unlikely to be true" seems pragmatically fine.  (Frequentists do, after all, assign _likelihoods_ to hypotheses.)  As for #5, absent some clarification of what "confident" means, this is ambiguous.  As I explain in teaching [http://bactra.org/notebooks/confidence-sets.html], a confidence set offers the reader a dilemma: _either_ the true parameter is in the set, _or_ we got data that what really unlikely and unrepresentative under any parameter value.
That said, I am unable to come up with face-saving interpretations of the other four items.]]></description>
<dc:subject>to_read confidence_sets statistics teaching via:? have_skimmed in_NB</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:1a4dda6ad6cb/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:confidence_sets"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:teaching"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:via:?"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:have_skimmed"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:in_NB"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://en.m.wikipedia.org/wiki/Spiders_Georg">
    <title>Spiders Georg - Wikipedia</title>
    <dc:date>2024-06-10T13:11:29+00:00</dc:date>
    <link>https://en.m.wikipedia.org/wiki/Spiders_Georg</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA[--- Bookmarking to have a stable reference for when I use this in teaching.]]></description>
<dc:subject>statistics funny:pointed to_teach:statistics_of_inequality_and_discrimination heavy_tails to_teach</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:af1852c94ecf/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:funny:pointed"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach:statistics_of_inequality_and_discrimination"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:heavy_tails"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://www.cambridge.org/core/journals/philosophy-of-science/article/bamboozled-by-bonferroni/0F570CAC41364F1F4D808A419A881277?WT.mc_id=New%2520Cambridge%2520Alert%2520-%2520Articles">
    <title>Bamboozled by Bonferroni | Philosophy of Science | Cambridge Core</title>
    <dc:date>2024-04-24T18:13:43+00:00</dc:date>
    <link>https://www.cambridge.org/core/journals/philosophy-of-science/article/bamboozled-by-bonferroni/0F570CAC41364F1F4D808A419A881277?WT.mc_id=New%2520Cambridge%2520Alert%2520-%2520Articles</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["When many statistical hypotheses are evaluated simultaneously, statisticians often recommend adjusting (or correcting) standard hypothesis tests. In this paper, I (1) distinguish two senses of adjustment, (2) investigate the prudential and epistemic goals that adjustment might achieve, and (3) identify conditions under which a researcher should not adjust for multiplicity in the two senses I identify. I tentatively conclude that the goals of scientists and the public may be misaligned with the decision criteria used to evaluate multiple testing regimes."]]></description>
<dc:subject>to:NB to_read multiple_testing statistics philosophy_of_science mayo-wilson.conor</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:8a22642ee27e/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:multiple_testing"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:philosophy_of_science"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:mayo-wilson.conor"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2404.11464">
    <title>[2404.11464] Rates of convergence and normal approximations for estimators of local dependence random graph models</title>
    <dc:date>2024-04-24T14:12:23+00:00</dc:date>
    <link>https://arxiv.org/abs/2404.11464</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Local dependence random graph models are a class of block models for network data which allow for dependence among edges under a local dependence assumption defined around the block structure of the network. Since being introduced by Schweinberger and Handcock (2015), research in the statistical network analysis and network science literatures have demonstrated the potential and utility of this class of models. In this work, we provide the first statistical disclaimers which provide conditions under which estimation and inference procedures can be expected to provide accurate and valid inferences. This is accomplished by deriving convergence rates of inference procedures for local dependence random graph models based on a single observation of the graph, allowing both the number of model parameters and the sizes of blocks to tend to infinity. First, we derive the first non-asymptotic bounds on the ℓ2-error of maximum likelihood estimators, along with convergence rates. Second, and more importantly, we derive the first non-asymptotic bounds on the error of the multivariate normal approximation. In so doing, we introduce the first principled approach to providing statistical disclaimers through quantifying the uncertainty about statistical conclusions based on data."

--- I kind of like the phrase "statistical disclaimer", but I'm pretty sure it's just good old fashioned probably-approximately-correct, a.k.a. consistency.]]></description>
<dc:subject>statistics network_data_analysis exponential_family_random_graphs stochastic_block_models in_NB</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:42f4a77dbcd5/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:network_data_analysis"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:exponential_family_random_graphs"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:stochastic_block_models"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:in_NB"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://link.springer.com/article/10.1007/s11186-023-09529-0">
    <title>The gates to the profession are open: the alternative institutionalization of data science | Theory and Society</title>
    <dc:date>2024-04-01T03:55:53+00:00</dc:date>
    <link>https://link.springer.com/article/10.1007/s11186-023-09529-0</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["In this study, I examine the institutional model of data science as a nascent profession undergoing an occupational founding phase. Drawing on interviews with sixty data scientists, senior managers, and professors from Israel as well as observations at the local professional community’s events, I argue that data scientists endorse an open institutional model, upholding largely internet-based institutions focusing on knowledge sharing, networking, and collaboration. This model grants data scientists expertise, autonomy, and authority vis-à-vis clients, employers, and states; provides them with continued credentialing independent of employing organizations; encourages the wide entry of new members; and helps them deal with the accelerated temporality of their field. This open model enables an omnivorous spreading of data science expertise and is used to challenge professionalization as an occupational-institutional model in other professions. Still, this model faces many challenges."]]></description>
<dc:subject>to:NB sociology system_of_professions statistics data_science_so_called</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:304810d738e8/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:sociology"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:system_of_professions"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:data_science_so_called"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2402.15213">
    <title>[2402.15213] Statistical Agnostic Regression: a machine learning method to validate regression models</title>
    <dc:date>2024-03-05T16:37:22+00:00</dc:date>
    <link>https://arxiv.org/abs/2402.15213</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Regression analysis is a central topic in statistical modeling, aiming to estimate the relationships between a dependent variable, commonly referred to as the response variable, and one or more independent variables, i.e., explanatory variables. Linear regression is by far the most popular method for performing this task in several fields of research, such as prediction, forecasting, or causal inference. Beyond various classical methods to solve linear regression problems, such as Ordinary Least Squares, Ridge, or Lasso regressions - which are often the foundation for more advanced machine learning (ML) techniques - the latter have been successfully applied in this scenario without a formal definition of statistical significance. At most, permutation or classical analyses based on empirical measures (e.g., residuals or accuracy) have been conducted to reflect the greater ability of ML estimations for detection. In this paper, we introduce a method, named Statistical Agnostic Regression (SAR), for evaluating the statistical significance of an ML-based linear regression based on concentration inequalities of the actual risk using the analysis of the worst case. To achieve this goal, similar to the classification problem, we define a threshold to establish that there is sufficient evidence with a probability of at least 1-eta to conclude that there is a linear relationship in the population between the explanatory (feature) and the response (label) variables. Simulations in only two dimensions demonstrate the ability of the proposed agnostic test to provide a similar analysis of variance given by the classical F test for the slope parameter."

--- I should read this, but the last tag applies with force.  The "classical F test for the slope parameter" in no way tests/validates the existence of a _linear_ relationship, even if all the classical assumptions hold.]]></description>
<dc:subject>statistics linear_regression hypothesis_testing color_me_skeptical in_NB</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:3c164bcb0d95/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:linear_regression"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:hypothesis_testing"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:color_me_skeptical"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:in_NB"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2309.10140">
    <title>[2309.10140] A Geometric Framework for Neural Feature Learning</title>
    <dc:date>2023-12-08T14:18:57+00:00</dc:date>
    <link>https://arxiv.org/abs/2309.10140</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["We present a novel framework for learning system design based on neural feature extractors by exploiting geometric structures in feature spaces. First, we introduce the feature geometry, which unifies statistical dependence and features in the same functional space with geometric structures. By applying the feature geometry, we formulate each learning problem as solving the optimal feature approximation of the dependence component specified by the learning setting. We propose a nesting technique for designing learning algorithms to learn the optimal features from data samples, which can be applied to off-the-shelf network architectures and optimizers. To demonstrate the application of the nesting technique, we further discuss multivariate learning problems, including conditioned inference and multimodal learning, where we present the optimal features and reveal their connections to classical approaches."]]></description>
<dc:subject>to:NB information_geometry variable_selection neural_networks statistics</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:47874c9b2e08/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:information_geometry"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:variable_selection"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:neural_networks"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="http://approximateinference.org/accepted/">
    <title>5th Symposium on Approximate Bayesian Inference (2023)</title>
    <dc:date>2023-12-08T14:10:44+00:00</dc:date>
    <link>http://approximateinference.org/accepted/</link>
    <dc:creator>cshalizi</dc:creator><dc:subject>statistics computational_statistics bayesianism to_download</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:01cc79977083/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:computational_statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:bayesianism"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_download"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://www.nature.com/articles/s41586-023-06079-4">
    <title>Polygenic scoring accuracy varies across the genetic ancestry continuum | Nature</title>
    <dc:date>2023-12-05T15:19:53+00:00</dc:date>
    <link>https://www.nature.com/articles/s41586-023-06079-4</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Polygenic scores (PGSs) have limited portability across different groupings of individuals (for example, by genetic ancestries and/or social determinants of health), preventing their equitable use1,2,3. PGS portability has typically been assessed using a single aggregate population-level statistic (for example, R2)4, ignoring inter-individual variation within the population. Here, using a large and diverse Los Angeles biobank5 (ATLAS, n = 36,778) along with the UK Biobank6 (UKBB, n = 487,409), we show that PGS accuracy decreases individual-to-individual along the continuum of genetic ancestries7 in all considered populations, even within traditionally labelled ‘homogeneous’ genetic ancestries. The decreasing trend is well captured by a continuous measure of genetic distance (GD) from the PGS training data: Pearson correlation of −0.95 between GD and PGS accuracy averaged across 84 traits. When applying PGS models trained on individuals labelled as white British in the UKBB to individuals with European ancestries in ATLAS, individuals in the furthest GD decile have 14% lower accuracy relative to the closest decile; notably, the closest GD decile of individuals with Hispanic Latino American ancestries show similar PGS performance to the furthest GD decile of individuals with European ancestries. GD is significantly correlated with PGS estimates themselves for 82 of 84 traits, further emphasizing the importance of incorporating the continuum of genetic ancestries in PGS interpretation. Our results highlight the need to move away from discrete genetic ancestry clusters towards the continuum of genetic ancestries when considering PGSs."]]></description>
<dc:subject>to:NB statistics human_genetics</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:de2fea5810ff/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:human_genetics"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://press.princeton.edu/books/paperback/9780691222752/data-science-for-neuroimaging">
    <title>Data Science for Neuroimaging | Princeton University Press</title>
    <dc:date>2023-11-16T18:14:01+00:00</dc:date>
    <link>https://press.princeton.edu/books/paperback/9780691222752/data-science-for-neuroimaging</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["As neuroimaging turns toward data-intensive discovery, researchers in the field must learn to access, manage, and analyze datasets at unprecedented scales. Concerns about reproducibility and increased rigor in reporting of scientific results also demand higher standards of computational practice. This book offers neuroimaging researchers an introduction to data science, presenting methods, tools, and approaches that facilitate automated, reproducible, and scalable analysis and understanding of data. Through guided, hands-on explorations of openly available neuroimaging datasets, the book explains such elements of data science as programming, data management, visualization, and machine learning, and describes their application to neuroimaging. Readers will come away with broadly relevant data science skills that they can easily translate to their own questions."]]></description>
<dc:subject>to:NB books:noted fmri neural_data_analysis yarkoni.tal statistics self-recommending books:suggest_to_library</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:575e7d937ce9/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:books:noted"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:fmri"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:neural_data_analysis"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:yarkoni.tal"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:self-recommending"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:books:suggest_to_library"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://projecteuclid.org/journals/annals-of-statistics/volume-30/issue-1/The-screening-effect-in-Kriging/10.1214/aos/1015362194.full">
    <title>The screening effect in Kriging</title>
    <dc:date>2023-10-05T01:22:50+00:00</dc:date>
    <link>https://projecteuclid.org/journals/annals-of-statistics/volume-30/issue-1/The-screening-effect-in-Kriging/10.1214/aos/1015362194.full</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["When predicting the value of a stationary random field at a location x in some region in which one has a large number of observations, it may be difficult to compute the optimal predictor. One simple way to reduce the computational burden is to base the predictor only on those observations nearest to x. As long as the number of observations used in the predictor is sufficiently large, one might generally expect the best predictor based on these observations to be nearly optimal relative to the best predictor using all observations. Indeed, this phenomenon has been empirically observed in numerous circumstances and is known as the screening effect in the geostatistical literature. For linear predictors, when observations are on a regular grid, this work proves that there generally is a screening effect as the grid becomes increasingly dense. This result requires that, at high frequencies, the spectral density of the random field not decay faster than algebraically and not vary too quickly. Examples demonstrate that there may be no screening effect if these conditions on the spectral density are violated."]]></description>
<dc:subject>have_skimmed spatial_statistics random_fields statistics to_teach:data_over_space_and_time in_NB</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:eb89cb46ceea/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:have_skimmed"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:spatial_statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:random_fields"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach:data_over_space_and_time"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:in_NB"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2206.06421">
    <title>[2206.06421] Repro Samples Method for Finite- and Large-Sample Inferences</title>
    <dc:date>2023-10-04T17:51:27+00:00</dc:date>
    <link>https://arxiv.org/abs/2206.06421</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["This article presents a novel, general, and effective simulation-inspired approach, called {\it repro samples method}, to conduct statistical inference. The approach studies the performance of artificial samples, referred to as {\it repro samples}, obtained by mimicking the true observed sample to achieve uncertainty quantification and construct confidence sets for parameters of interest with guaranteed coverage rates. Both exact and asymptotic inferences are developed. An attractive feature of the general framework developed is that it does not rely on the large sample central limit theorem and is likelihood-free. As such, it is thus effective for complicated inference problems which we can not solve using the large sample central limit theorem. The proposed method is applicable to a wide range of problems, including many open questions where solutions were previously unavailable, for example, those involving discrete or non-numerical parameters. To reduce the large computational cost of such inference problems, we develop a unique matching scheme to obtain a data-driven candidate set. Moreover, we show the advantages of the proposed framework over the classical Neyman-Pearson framework. We demonstrate the effectiveness of the proposed approach on various models throughout the paper and provide a case study that addresses an open inference question on how to quantify the uncertainty for the unknown number of components in a normal mixture model. To evaluate the empirical performance of our repro samples method, we conduct simulations and study real data examples with comparisons to existing approaches. Although the development pertains to the settings where the large sample central limit theorem does not apply, it also has direct extensions to the cases where the central limit theorem does hold."

--- Based on the talk on Monday, I don't see how this _isn't_ just the Neyman inversion method, with a very clever idea about how to do the testing that I need to wrap my head around.  But it seems very cool, and to be, potentially, very useful to me.  So this needs careful attention.

--- ETA after reading carefully: It's Neyman inversion.  Also, they're not actually getting valid confidence intervals for the number of mixture components, because there's no way to give an upper confidence limit for the number of mixture components.  (For any distribution which really does have k components, there are others with arbitrarily many more clusters, arbitrarily close in distribution.)  They _think_ they can do this because they arbitrarily limit how many clusters they consider.
Now, in the talk Xie gave a rather more convincing example of a confidence set for a discrete parameter, viz., which node on a network some process started spreading from.  The difference, I think, is that in this 2nd case, we can't switch the value of the discrete parameter while making an _arbitrarily small_, and hence undetectably small, change to the distribution.]]></description>
<dc:subject>heard_the_talk confidence_sets simulation-based_inference statistics re:codename:catherine_wheel in_NB have_read</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:72263b788043/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:heard_the_talk"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:confidence_sets"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:simulation-based_inference"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:re:codename:catherine_wheel"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:in_NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:have_read"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://today.yougov.com/topics/politics/articles-reports/2022/03/15/americans-misestimate-small-subgroups-population">
    <title>​​From millionaires to Muslims, small subgroups of the population seem much larger to many Americans | YouGov</title>
    <dc:date>2023-06-15T19:01:21+00:00</dc:date>
    <link>https://today.yougov.com/topics/politics/articles-reports/2022/03/15/americans-misestimate-small-subgroups-population</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA[--- Thought I'd bookmarked this already?
--- Incidentally, this bit:
"Black Americans estimate that, on average, Black people make up 52% of the U.S. adult population; non-Black Americans estimate the proportion is roughly 39%, closer to the real figure of 12%. First-generation immigrants we surveyed estimate that first-generation immigrants account for 40% of U.S. adults, while non-immigrants guess it is around 31%, closer to the actual figure of 14%."
seems like it explains a lot of our recent politics.]]></description>
<dc:subject>have_read statistics cognitive_science demography to_teach:statistics_of_inequality_and_discrimination</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:11dc79e7b78b/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:have_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:cognitive_science"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:demography"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach:statistics_of_inequality_and_discrimination"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5001799/">
    <title>Is the United States Maternal Mortality Rate Increasing? Disentangling trends from measurement issues Short title: U.S. Maternal Mortality Trends - PMC</title>
    <dc:date>2023-06-15T19:00:21+00:00</dc:date>
    <link>https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5001799/</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Background: A pregnancy question was added to the U.S. standard death certificate in 2003 to improve ascertainment of maternal deaths. The delayed adoption of this question among states led to data incompatibilities, and impeded accurate trend analysis. Our objectives were to develop methods for trend analysis, and to provide an overview of U.S. maternal mortality trends from 2000–2014.
"Methods: This observational study analyzed vital statistics maternal mortality data from all U.S. states in relation to the format and year-of-adoption of the pregnancy question. Correction factors were developed to adjust data from before the standard pregnancy question was adopted, to promote accurate trend analysis. Joinpoint regression was used to analyze trends for groups of states with similar pregnancy questions.
"Results: The estimated maternal mortality rate (per 100,000 live births) for 48 states and Washington D.C. (excluding California and Texas, analyzed separately) increased by 26.6%, from 18.8 in 2000 to 23.8 in 2014. California showed a declining trend, while Texas had a sudden increase in 2011–2012. Analysis of the measurement change suggests that U.S. rates in the early 2000s were higher than previously reported.
"Discussion: Despite the United Nations Millennium Development Goal for a 75% reduction in maternal mortality by 2015, the estimated maternal mortality rate for 48 states and Washington D.C. increased from 2000–2014, while the international trend was in the opposite direction. There is a need to redouble efforts to prevent maternal deaths and improve maternity care for the 4 million U.S. women giving birth each year."

--- To teach to The Kids, with the moral being that even something that seems straightforward is really complicated, and usually a mess.

--- WTH is up with Texas?  Doubling over two years cries out "measurement issues" but they say they looked into it carefully...]]></description>
<dc:subject>have_read social_measurement demography statistics to_teach in_NB</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:b3ff5c1d8dde/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:have_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:social_measurement"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:demography"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:in_NB"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://doi.org/10.1080/01621459.2023.2197686">
    <title>Cross-Validation: What Does It Estimate and How Well Does It Do It?</title>
    <dc:date>2023-06-08T15:37:53+00:00</dc:date>
    <link>https://doi.org/10.1080/01621459.2023.2197686</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Cross-validation is a widely used technique to estimate prediction error, but its behavior is complex and not fully understood. Ideally, one would like to think that cross-validation estimates the prediction error for the model at hand, fit to the training data. We prove that this is not the case for the linear model fit by ordinary least squares; rather it estimates the average prediction error of models fit on other unseen training sets drawn from the same population. We further show that this phenomenon occurs for most popular estimates of prediction error, including data splitting, bootstrapping, and Mallow’s $C_p$. Next, the standard confidence intervals for prediction error derived from cross-validation may have coverage far below the desired level. Because each data point is used for both training and testing, there are correlations among the measured accuracies for each fold, and so the usual estimate of variance is too small. We introduce a nested cross-validation scheme to estimate this variance more accurately, and show empirically that this modification leads to intervals with approximately correct coverage in many examples where traditional cross-validation intervals fail. Lastly, our analysis also shows that when producing confidence intervals for prediction accuracy with simple data splitting, one should not refit the model on the combined data, since this invalidates the confidence intervals."

--- (Last part is obvious, no?)]]></description>
<dc:subject>cross-validation statistics tibshirani.robert hastie.trevor to_read to_teach in_NB</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:92e07da3433c/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:cross-validation"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:tibshirani.robert"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:hastie.trevor"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:in_NB"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2305.08429">
    <title>[2305.08429] Bayesian inference for misspecified generative models</title>
    <dc:date>2023-06-08T15:32:21+00:00</dc:date>
    <link>https://arxiv.org/abs/2305.08429</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Bayesian inference is a powerful tool for combining information in complex settings, a task of increasing importance in modern applications. However, Bayesian inference with a flawed model can produce unreliable conclusions. This review discusses approaches to performing Bayesian inference when the model is misspecified, where by misspecified we mean that the analyst is unwilling to act as if the model is correct. Much has been written about this topic, and in most cases we do not believe that a conventional Bayesian analysis is meaningful when there is serious model misspecification. Nevertheless, in some cases it is possible to use a well-specified model to give meaning to a Bayesian analysis of a misspecified model and we will focus on such cases. Three main classes of methods are discussed - restricted likelihood methods, which use a model based on a non-sufficient summary of the original data; modular inference methods which use a model constructed from coupled submodels and some of the submodels are correctly specified; and the use of a reference model to construct a projected posterior or predictive distribution for a simplified model considered to be useful for prediction or interpretation."]]></description>
<dc:subject>to:NB bayesianism misspecification statistics re:phil-of-bayes_paper to_read</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:683fd1cb974b/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:bayesianism"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:misspecification"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:re:phil-of-bayes_paper"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_read"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2301.13724">
    <title>[2301.13724] The passive symmetries of machine learning</title>
    <dc:date>2023-05-02T21:11:45+00:00</dc:date>
    <link>https://arxiv.org/abs/2301.13724</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Any representation of data involves arbitrary investigator choices. Because those choices are external to the data-generating process, each choice leads to an exact symmetry, corresponding to the group of transformations that takes one possible representation to another. These are the passive symmetries; they include coordinate freedom, gauge symmetry and units covariance, all of which have led to important results in physics. Our goal is to understand the implications of passive symmetries for machine learning: Which passive symmetries play a role (e.g., permutation symmetry in graph neural networks)? What are dos and don'ts in machine learning practice? We assay conditions under which passive symmetries can be implemented as group equivariances. We also discuss links to causal modeling, and argue that the implementation of passive symmetries is particularly valuable when the goal of the learning problem is to generalize out of sample. While this paper is purely conceptual, we believe that it can have a significant impact on helping machine learning make the transition that took place for modern physics in the first half of the Twentieth century."]]></description>
<dc:subject>symmetry machine_learning statistics via:rvenkat modeling scholkopf.bernhard in_NB</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:83362f63b370/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:symmetry"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:machine_learning"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:via:rvenkat"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:modeling"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:scholkopf.bernhard"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:in_NB"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://osf.io/7vy2f/">
    <title>OSF Preprints | Quantitative Political Science Research is Greatly Underpowered</title>
    <dc:date>2023-05-02T20:12:17+00:00</dc:date>
    <link>https://osf.io/7vy2f/</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["We analyze the statistical power of political science research by collating over 16,000 hypothesis tests from about 2,000 articles. Even with generous assumptions, the median analysis has about 10% power, and only about 1 in 10 tests have at least 80% power to detect the consensus effects reported in the literature. There is also substantial heterogeneity in tests across research areas, with some being characterized by high-power but most having very low power. To contextualize our findings, we survey political methodologists to assess their expectations about power levels. Most methodologists greatly overestimate the statistical power of political science research."]]></description>
<dc:subject>to:NB to_read political_science social_science_methodology statistics hypothesis_testing estimation re:neutral_model_of_inquiry</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:c31528a09f2b/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:political_science"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:social_science_methodology"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:hypothesis_testing"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:estimation"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:re:neutral_model_of_inquiry"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://academic.oup.com/ej/article-abstract/127/605/F236/5069452?login=false">
    <title>Power of Bias in Economics Research | The Economic Journal | Oxford Academic</title>
    <dc:date>2023-05-02T20:10:08+00:00</dc:date>
    <link>https://academic.oup.com/ej/article-abstract/127/605/F236/5069452?login=false</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["We investigate two critical dimensions of the credibility of empirical economics research: statistical power and bias. We survey 159 empirical economics literatures that draw upon 64,076 estimates of economic parameters reported in more than 6,700 empirical studies. Half of the research areas have nearly 90% of their results under‐powered. The median statistical power is 18%, or less. A simple weighted average of those reported results that are adequately powered (power ≥ 80%) reveals that nearly 80% of the reported effects in these empirical economics literatures are exaggerated; typically, by a factor of two and with one‐third inflated by a factor of four or more."

--- Power's really a function, not a number, so where's "18%" come from?  Is that the power to detect an effect of the magnitude estimated (a little weirdly recursive...), or some standard-size magnitude?
--- ETA after reading: Yes, for each area of economics they do a supposedly-robust meta-estimate of the effect size, and try to work out the power to detect an effect that big.]]></description>
<dc:subject>to:NB economics econometrics statistics hypothesis_testing re:neutral_model_of_inquiry estimation have_read</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:2c2f32247eac/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:economics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:econometrics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:hypothesis_testing"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:re:neutral_model_of_inquiry"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:estimation"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:have_read"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/physics/0001019">
    <title>[physics/0001019] Approaching the parameter estimation quality of maximum likelihood via generalized moments</title>
    <dc:date>2023-05-01T20:12:15+00:00</dc:date>
    <link>https://arxiv.org/abs/physics/0001019</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["A simple criterion is presented for a practical construction of generalized moments that allow one to approach the theoretical Rao-Cramer limit for parameter estimation while avoiding the complexity of the maximum likelihood method in the cases of complicated probability distributions and/or very large event samples."

--- To summarize: (1) maximimizing the MLE is equivalent to finding the parameter value where the mean of the gradient of the log-likelihood is zero.  (This is what Anglophone statistics calls the "score function", which is a horribly opaque name.)  (2) If we replace the gradient of the log-likelihood by a function that's close to it, but more tractable, setting _its_ mean to zero gives us estimates that are almost as efficient as the MLE.  (FVT works out the Taylor series.)  (3) The functions only have to be close in regions of high (true) probability.  I don't think this is as radical as the author did (it's just another Z estimator and I don't see how it helps when the true pdf is really intractable), but it's clever and illuminating and potentially helpful.]]></description>
<dc:subject>in_NB have_read likelihood estimation statistics cleaning_out_the_filing_cabinet_for_the_first_time_since_2005</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:418dde8076ff/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:in_NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:have_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:likelihood"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:estimation"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:cleaning_out_the_filing_cabinet_for_the_first_time_since_2005"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://joss.theoj.org/papers/10.21105/joss.02232">
    <title>Journal of Open Source Software: policytree: Policy learning via doubly robust empirical welfare maximization over trees</title>
    <dc:date>2023-04-27T14:40:08+00:00</dc:date>
    <link>https://joss.theoj.org/papers/10.21105/joss.02232</link>
    <dc:creator>cshalizi</dc:creator><dc:subject>to:NB decision_trees athey.susan wager.stefan statistics</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:364aa068e29b/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:decision_trees"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:athey.susan"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:wager.stefan"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://www.jstor.org/stable/2336834">
    <title>Parameter-Based Asymptotics on JSTOR</title>
    <dc:date>2023-04-24T21:32:07+00:00</dc:date>
    <link>https://www.jstor.org/stable/2336834</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["In certain cases statistical methods based on standard maximum likelihood asymptotics become valid as the true parameter value approaches a boundary of the parameter space. Examples are given which motivate a general parameter-based asymptotic theory, and a result is obtained which covers such situations. Of particular interest are applications to stochastic process models."]]></description>
<dc:subject>to_reread estimation statistics cleaning_out_the_filing_cabinet_for_the_first_time_since_2005 likelihood statistical_inference_for_stochastic_processes re:HEAS in_NB</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:5c278c6fcc6d/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_reread"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:estimation"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:cleaning_out_the_filing_cabinet_for_the_first_time_since_2005"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:likelihood"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistical_inference_for_stochastic_processes"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:re:HEAS"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:in_NB"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://www.tandfonline.com/doi/abs/10.1080/01621459.2021.1941053">
    <title>The Generalized Oaxaca-Blinder Estimator: Journal of the American Statistical Association: Vol 0, No 0</title>
    <dc:date>2023-03-07T15:40:40+00:00</dc:date>
    <link>https://www.tandfonline.com/doi/abs/10.1080/01621459.2021.1941053</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["After performing a randomized experiment, researchers often use ordinary least-square (OLS) regression to adjust for baseline covariates when estimating the average treatment effect. It is widely known that the resulting confidence interval is valid even if the linear model is misspecified. In this article, we generalize that conclusion to covariate adjustment with nonlinear models. We introduce an intuitive way to use any “simple” nonlinear model to construct a covariate-adjusted confidence interval for the average treatment effect. The confidence interval derives its validity from randomization alone, and when nonlinear models fit the data better than linear models, it is narrower than the usual interval from OLS adjustment."]]></description>
<dc:subject>to:NB causal_inference experimental_design statistics</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:c6eec1e68747/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:causal_inference"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:experimental_design"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2109.03582">
    <title>[2109.03582] Higher Order Kernel Mean Embeddings to Capture Filtrations of Stochastic Processes</title>
    <dc:date>2023-02-15T19:57:50+00:00</dc:date>
    <link>https://arxiv.org/abs/2109.03582</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Stochastic processes are random variables with values in some space of paths. However, reducing a stochastic process to a path-valued random variable ignores its filtration, i.e. the flow of information carried by the process through time. By conditioning the process on its filtration, we introduce a family of higher order kernel mean embeddings (KMEs) that generalizes the notion of KME and captures additional information related to the filtration. We derive empirical estimators for the associated higher order maximum mean discrepancies (MMDs) and prove consistency. We then construct a filtration-sensitive kernel two-sample test able to pick up information that gets missed by the standard MMD test. In addition, leveraging our higher order MMDs we construct a family of universal kernels on stochastic processes that allows to solve real-world calibration and optimal stopping problems in quantitative finance (such as the pricing of American options) via classical kernel-based regression methods. Finally, adapting existing tests for conditional independence to the case of stochastic processes, we design a causal-discovery algorithm to recover the causal graph of structural dependencies among interacting bodies solely from observations of their multidimensional trajectories."

--- ETA after reading: This feels like a strange paper.  I'm not sure I truly understand what theiur "signature statistics" do, nor do I quite get the claimed advantage of higher-order process kernels over "first-order" kernels.  (Proofs are referred to other, older papers.)  And the notion of "causality" between processes seems very weird, since I don't see how it accounts for the flow of time, and of influence, within or across processes, they're being treated like big but indecomposable objects.  Probably should track down references and see if this makes more sense when I put those together.]]></description>
<dc:subject>to:NB stochastic_processes kernel_methods causal_discovery time_series statistical_inference_for_stochastic_processes hilbert_space re:codename:catherine_wheel two-sample_tests statistics have_read path_signatures</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:105636bb7bad/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:stochastic_processes"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:kernel_methods"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:causal_discovery"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:time_series"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistical_inference_for_stochastic_processes"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:hilbert_space"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:re:codename:catherine_wheel"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:two-sample_tests"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:have_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:path_signatures"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://www.aeaweb.org/articles?id=10.1257/app.20190297">
    <title>Mortality Change among Less Educated Americans - American Economic Association</title>
    <dc:date>2023-01-01T04:26:29+00:00</dc:date>
    <link>https://www.aeaweb.org/articles?id=10.1257/app.20190297</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Measurements of mortality change among less educated Americans can be biased because the least educated groups (e.g., dropouts) become smaller and more negatively selected over time. We show that mortality changes at constant education percentiles can be bounded with minimal assumptions. Middle-age mortality increases among non-Hispanic Whites from 1992 to 2018 are driven almost entirely by the bottom 10 percent of the education distribution. Drivers of mortality change differ substantially across groups. Deaths of despair explain most of the mortality change among young non-Hispanic Whites, but less among older Whites and non-Hispanic Blacks. Our bounds are applicable in many other contexts."

--- Very curious about how they get these percentiles.]]></description>
<dc:subject>to:NB demography statistics inequality to_teach:statistics_of_inequality_and_discrimination</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:f915cce2a34c/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:demography"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:inequality"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach:statistics_of_inequality_and_discrimination"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2210.07491">
    <title>[2210.07491] Latent process models for functional network data</title>
    <dc:date>2022-12-09T20:01:58+00:00</dc:date>
    <link>https://arxiv.org/abs/2210.07491</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Network data are often sampled with auxiliary information or collected through the observation of a complex system over time, leading to multiple network snapshots indexed by a continuous variable. Many methods in statistical network analysis are traditionally designed for a single network, and can be applied to an aggregated network in this setting, but that approach can miss important functional structure. Here we develop an approach to estimating the expected network explicitly as a function of a continuous index, be it time or another indexing variable. We parameterize the network expectation through low dimensional latent processes, whose components we represent with a fixed, finite-dimensional functional basis. We derive a gradient descent estimation algorithm, establish theoretical guarantees for recovery of the low-dimensional structure, compare our method to competitors, and apply it to a dataset of international political interactions over time, showing our proposed method to adapt well to data, outperform competitors, and provide interpretable and meaningful results."]]></description>
<dc:subject>to:NB network_data_analysis statistics inference_to_latent_objects functional_data_analysis levina.liza to_read</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:df335b19ec43/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:network_data_analysis"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:inference_to_latent_objects"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:functional_data_analysis"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:levina.liza"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_read"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://www.pnas.org/doi/10.1073/pnas.2203150119">
    <title>Observing many researchers using the same data and hypothesis reveals a hidden universe of uncertainty | PNAS</title>
    <dc:date>2022-11-19T20:36:42+00:00</dc:date>
    <link>https://www.pnas.org/doi/10.1073/pnas.2203150119</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["This study explores how researchers’ analytical choices affect the reliability of scientific findings. Most discussions of reliability problems in science focus on systematic biases. We broaden the lens to emphasize the idiosyncrasy of conscious and unconscious decisions that researchers make during data analysis. We coordinated 161 researchers in 73 research teams and observed their research decisions as they used the same data to independently test the same prominent social science hypothesis: that greater immigration reduces support for social policies among the public. In this typical case of social science research, research teams reported both widely diverging numerical findings and substantive conclusions despite identical start conditions. Researchers’ expertise, prior beliefs, and expectations barely predict the wide variation in research outcomes. More than 95% of the total variance in numerical results remains unexplained even after qualitative coding of all identifiable decisions in each team’s workflow. This reveals a universe of uncertainty that remains hidden when considering a single study in isolation. The idiosyncratic nature of how researchers’ results and conclusions varied is a previously underappreciated explanation for why many scientific hypotheses remain contested. These results call for greater epistemic humility and clarity in reporting scientific findings."]]></description>
<dc:subject>to:NB to_read data_analysis statistics social_science_methodology</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:1d84ac94f53f/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:data_analysis"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:social_science_methodology"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://nap.nationalacademies.org/catalog/5499/health-effects-of-exposure-to-radon-beir-vi">
    <title>Health Effects of Exposure to Radon: BEIR VI |The National Academies Press</title>
    <dc:date>2022-10-24T02:02:36+00:00</dc:date>
    <link>https://nap.nationalacademies.org/catalog/5499/health-effects-of-exposure-to-radon-beir-vi</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Radon progeny—the decay products of radon gas—are a well-recognized cause of lung cancer in miners working underground. When radon was found to be a ubiquitous indoor air pollutant, however, it raised a more widespread alarm for public health.
"To develop appropriate public policy for indoor radon, decisionmakers need a characterization of the risk of radon exposure across the range of exposures people actually receive. In response, the BEIR VI committee has developed a mathematical model for the lung cancer risk associated with radon, incorporating the latest information from epidemiology and scientific studies.
"In this book the committee provides a fresh assessment of exposure-dose relationships. The volume discusses key issues—such as the weight of biological evidence and extrapolation from radon-exposed miners to the larger population—in estimating the risk posed by indoor radon. It also addresses such uncertainties as the combined effects of smoking and radon and the impact of the rate of exposure.
"The committee considered the entire body of evidence on radon and lung cancer, integrating findings from epidemiological studies with evidence from animal experiments and other lines of laboratory investigation. The conclusions will be important to policymakers and environmental advocates, while the technical findings will be of interest to environmental scientists and engineers."]]></description>
<dc:subject>to_read radon epidemiology statistics</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:450d60515b71/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:radon"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:epidemiology"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/2112.03626">
    <title>[2112.03626] Phase transitions in nonparametric regressions: a curse of exploiting higher degree smoothness assumptions in finite samples</title>
    <dc:date>2022-07-25T17:15:55+00:00</dc:date>
    <link>https://arxiv.org/abs/2112.03626</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["When the regression function belongs to a smooth class consisting of univariate functions with derivatives up to the (γ+1)th order bounded in absolute values for a finite γ, it is generally viewed that exploiting higher degree smoothness assumptions helps reduce the estimation error. This paper shows that the minimax optimal mean integrated squared error (MISE) increases in γ when the sample size n is small relative to the order of (γ+1)2γ+3, and decreases in γ when n is large relative to the order of (γ+1)2γ+3. In particular, this phase transition property is shown to be achieved by common nonparametric procedures. Consider γ1 and γ2 such that γ1<γ2, where the (γ2+1)th degree smoothness class is a subset of the (γ1+1)th degree class. What is surprising about our results is that they imply, if n is small relative to the order of (γ1+1)2γ1+3, the optimal rate achieved by the estimator constrained to be in the smoother class (to exploit the (γ2+1)th degree smoothness) is slower. In data sets with fewer than hundreds-of-thousands observations, our results suggest that one should not exploit beyond the third or fourth degree of smoothness. To some extent, our results provide a theoretical basis for the widely adopted practical recommendations given by Gelman and Imbens (2019).
"The building blocks of our minimax optimality results are a set of metric entropy bounds we develop in this paper for smooth function classes. Some of our bounds are original, and some of them improve and/or generalize the ones in the literature."

--- This is really surprising to me, so I ought to see what makes it work.  (On the plus side, if right, it makes me feel better about not teaching The Kids about higher-order smoothness assumptions!)]]></description>
<dc:subject>in_NB regression nonparametrics minimax learning_theory statistics</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:e1ba3a57ef77/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:in_NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:regression"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:nonparametrics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:minimax"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:learning_theory"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://jmlr.org/papers/v23/20-644.html">
    <title>Data-Derived Weak Universal Consistency</title>
    <dc:date>2022-07-19T13:56:11+00:00</dc:date>
    <link>https://jmlr.org/papers/v23/20-644.html</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Many current applications in data science need rich model classes to adequately represent the statistics that may be driving the observations. Such rich model classes may be too complex to admit uniformly consistent estimators. In such cases, it is conventional to settle for estimators with guarantees on convergence rate where the performance can be bounded in a model-dependent way, i.e. pointwise consistent estimators. But this viewpoint has the practical drawback that estimator performance is a function of the unknown model within the model class that is being estimated. Even if an estimator is consistent, how well it is doing at any given time may not be clear, no matter what the sample size of the observations. In these cases, a line of analysis favors sample dependent guarantees. We explore this framework by studying rich model classes that may only admit pointwise consistency guarantees, yet enough information about the unknown model driving the observations needed to gauge estimator accuracy can be inferred from the sample at hand. In this paper we obtain a novel characterization of lossless compression problems over a countable alphabet in the data-derived framework in terms of what we term deceptive distributions. We also show that the ability to estimate the redundancy of compressing memoryless sources is equivalent to learning the underlying single-letter marginal in a data-derived fashion. We expect that the methodology underlying such characterizations in a data-derived estimation framework will be broadly applicable to a wide range of estimation problems, enabling a more systematic approach to data-derived guarantees."

--- Last tag is contingent on reading it and liking it, of course.  
]]></description>
<dc:subject>in_NB learning_theory statistics information_theory to_read to_teach:childs_garden_of_statistical_learning_theory</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:7ba350217b53/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:in_NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:learning_theory"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:information_theory"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach:childs_garden_of_statistical_learning_theory"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://arxiv.org/abs/1811.11603">
    <title>[1811.11603] Distribution Regression with Sample Selection, with an Application to Wage Decompositions in the UK</title>
    <dc:date>2022-07-09T18:57:08+00:00</dc:date>
    <link>https://arxiv.org/abs/1811.11603</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["We develop a distribution regression model under endogenous sample selection. This model is a semiparametric generalization of the Heckman selection model that accommodates much richer patterns of heterogeneity in the selection process and effect of the covariates. The model applies to continuous, discrete and mixed outcomes. We study the identification of the model, and develop a computationally attractive two-step method to estimate the model parameters, where the first step is a probit regression for the selection equation and the second step consists of multiple distribution regressions with selection corrections for the outcome equation. We construct estimators of functionals of interest such as actual and counterfactual distributions of latent and observed outcomes via plug-in rule. We derive functional central limit theorems for all the estimators and show the validity of multiplier bootstrap to carry out functional inference. We apply the methods to wage decompositions in the UK using new data. Here we decompose the difference between the male and female wage distributions into four effects: composition, wage structure, selection structure and selection sorting. After controlling for endogenous employment selection, we still find substantial gender wage gap -- ranging from 21% to 40% throughout the (latent) offered wage distribution that is not explained by observable labor market characteristics. We also uncover positive sorting for single men and negative sorting for married women that accounts for a substantive fraction of the gender wage gap at the top of the distribution. These findings can be interpreted as evidence of assortative matching in the marriage market and glass-ceiling in the labor market."

--- Last tag is "I should know this stuff", not "I should teach it to sophomores and juniors".]]></description>
<dc:subject>to:NB statistics density_estimation regression causal_inference inequality via:donsker_class to_teach:statistics_of_inequality_and_discrimination</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:a84c4133e283/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:density_estimation"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:regression"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:causal_inference"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:inequality"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:via:donsker_class"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach:statistics_of_inequality_and_discrimination"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://www.jstor.org/stable/144855">
    <title>Wage Discrimination: Reduced Form and Structural Estimates on JSTOR (Blinder, 1973)</title>
    <dc:date>2022-07-07T20:56:37+00:00</dc:date>
    <link>https://www.jstor.org/stable/144855</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["Regressions explaining the wage rates of white males, black males, and white females are used to analyze the white-black wage differential among men and the male-female wage differential among whites. A distinction is drawn between reduced form and structural wage equations, and both are estimated. They are shown to have very different implications for analyzing the white-black and male-female wage differentials. When the two sets of estimates are synthesized, they jointly imply that 70 percent of the overall race differential and 100 percent of the overall sex differential are ultimately attributable to discrimination of various sorts."

--- Cites Oaxaca; does not cite Kitagawa; of course does not even begin to consider checking whether any of the supposedly-linear relationships in the supposedly-structural model are in fact linear.]]></description>
<dc:subject>economics inequality statistics discrimination have_read to_teach:statistics_of_inequality_and_discrimination regression have_taught</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:11e7673afead/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:economics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:inequality"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:discrimination"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:have_read"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to_teach:statistics_of_inequality_and_discrimination"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:regression"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:have_taught"/>
</rdf:Bag></taxo:topics>
</item>
<item rdf:about="https://www.cambridge.org/core/books/mathematical-pictures-at-a-data-science-exhibition/0AD8668B4657A24C5135DA7DA6641B41?pageNum=2&amp;searchWithinIds=0AD8668B4657A24C5135DA7DA6641B41&amp;productType=BOOK_PART&amp;searchWithinIds=0AD8668B4657A24C5135DA7DA6641B41&amp;productType=BOOK_PART&amp;sort=mtdMetadata.bookPartMeta._mtdPositionSortable%3Aasc&amp;pageSize=30&amp;template=cambridge-core%2Fbook%2Fcontents%2Flistings&amp;ignoreExclusions=true#fndtn-information">
    <title>Mathematical Pictures at a Data Science Exhibition</title>
    <dc:date>2022-07-02T13:39:17+00:00</dc:date>
    <link>https://www.cambridge.org/core/books/mathematical-pictures-at-a-data-science-exhibition/0AD8668B4657A24C5135DA7DA6641B41?pageNum=2&amp;searchWithinIds=0AD8668B4657A24C5135DA7DA6641B41&amp;productType=BOOK_PART&amp;searchWithinIds=0AD8668B4657A24C5135DA7DA6641B41&amp;productType=BOOK_PART&amp;sort=mtdMetadata.bookPartMeta._mtdPositionSortable%3Aasc&amp;pageSize=30&amp;template=cambridge-core%2Fbook%2Fcontents%2Flistings&amp;ignoreExclusions=true#fndtn-information</link>
    <dc:creator>cshalizi</dc:creator><description><![CDATA["This text provides deep and comprehensive coverage of the mathematical background for data science, including machine learning, optimal recovery, compressed sensing, optimization, and neural networks. In the past few decades, heuristic methods adopted by big tech companies have complemented existing scientific disciplines to form the new field of Data Science. This text embarks the readers on an engaging itinerary through the theory supporting the field. Altogether, twenty-seven lecture-length chapters with exercises provide all the details necessary for a solid understanding of key topics in data science. While the book covers standard material on machine learning and optimization, it also includes distinctive presentations of topics such as reproducing kernel Hilbert spaces, spectral clustering, optimal recovery, compressed sensing, group testing, and applications of semidefinite programming. Students and data scientists with less mathematical background will appreciate the appendices that provide more background on some of the more abstract concepts."]]></description>
<dc:subject>to:NB books:noted statistics optimization compressed_sensing neural_networks downloaded</dc:subject>
<dc:source>https://pinboard.in/</dc:source>
<dc:identifier>https://pinboard.in/u:cshalizi/b:bdf98b946f53/</dc:identifier>
<taxo:topics><rdf:Bag>	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:to:NB"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:books:noted"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:statistics"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:optimization"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:compressed_sensing"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:neural_networks"/>
	<rdf:li rdf:resource="https://pinboard.in/u:cshalizi/t:downloaded"/>
</rdf:Bag></taxo:topics>
</item>
</rdf:RDF>