Measurement Error

This article reassesses affective polarization measurement and reports that much partisan animus reflects misperceptions: partisans picture out-party members as ideologically extreme and highly engaged, but typical out-partisans are far less extreme. When respondents learn about the actual distribution of out-party attitudes and engagement, reported animus declines markedly. The paper highlights measurement artifacts and suggests remedies for more accurate assessments of interparty dislike.

Many political scientists measure dispositions—like populism, conspiracism, racism, and sexism—using agree/disagree statements, but that format can produce acquiescence bias that inflates endorsement rates. This article asks how acquiescence affects criterion validity and evaluates common remedies. Drawing on 160 tests and an internal meta-analysis, the authors examine how bias changes estimated relationships between constructs and assess balanced scales and measurement models as potential solutions. They end with design advice and recommendations for redesigning several popular measures of political and psychological dispositions.

This note examines whether low-quality online survey panels can produce misleading inferences about acquiescence bias in measures of conspiratorial beliefs and political misinformation. The authors reanalyze and replicate an earlier survey experiment, identify a coding error that led to the surprising result that respondents with postgraduate education appeared most prone to acquiescence, and then run two preregistered replications to probe that pattern.

This paper develops a dynamic discrete choice framework for multiple attitudes to identify which attitudes are most influential within an attitude system and to distinguish sources of stability and instability. The model accommodates ordinal attitudinal measures, time-varying gaps between interviews, and a multiple-factor structure to capture time-invariant individual traits. The authors apply the approach to five core political attitudes and to issue attitudes about government spending and immigration, preserving qualifications about the roles of permanent heterogeneity and transitory shocks.

Taxpayers often hide true income because admitting evasion is socially undesirable, so identifying noncompliance is difficult. This study uses a list experiment to reduce social desirability bias and uncover who is most likely to underreport income on tax returns.

This study asks whether commonly observed gaps—women scoring lower on political knowledge, expressing less interest, and being less likely to give valid answers—are driven by how surveys define "politics." The author shifts the focus from questions about political institutions to a broader set of issues that can be considered political, and uses an experimental design to test whether measurement choices systematically disadvantage women. The contribution is a qualified challenge to the conventional interpretation of political knowledge and interest: the gender gaps appear to depend on the traditional, institution-centered content of survey items rather than on an overall deficit in women’s political information or engagement.

The Facebook URLs Dataset contains over 40 trillion cell values, making it one of the largest social science research datasets ever assembled. The release applies a version of differential privacy that adds specially calibrated random noise, providing mathematical guarantees for the privacy of individual research subjects while aiming to preserve aggregate patterns useful to social scientists.

This paper examines whether mixing different orders of integration on the left- and right-hand sides of a balanced time-series equation prevents researchers from identifying a true relationship. Building on prior work about equation balance, the authors focus on consequences for type II error when using general error correction/autoregressive distributed lag (GECM/ADL) models and consider both the asymptotic case and finite-sample behavior. Their analysis combines theoretical results and simulations to show how model selection and standard assumption checks affect practical inference.

Experiments should be designed so that experimental measurement error can be detected. Implementing identical protocols across diverse modes makes it possible to see whether apparent treatment effect heterogeneity is driven by substantive differences or by measurement problems.

The spatial weights matrix W sits at the heart of spatial regression models, yet there are few techniques for validating a chosen specification of W. Reframing misspecification of W as a measurement error reveals a distinctive form of endogeneity that matters for spatial inference.

Measurement error undermines survey validity, particularly for sensitive questions. List experiments reduce deliberate misreporting but remain vulnerable to nonstrategic measurement error from poor implementation and respondent inattention. Such error violates the assumptions of the standard maximum likelihood regression (MLreg) estimator and can produce misleading inferences—especially when the sensitive trait is rare.

Event-level data, typically derived from reports of social conflict, often masks crucial variations by aggregating multiple sources. This analysis argues that standard aggregation methods limit researchers' ability to evaluate report-specific measurement issues and discard valuable information about these differences. We advocate for using report-level data instead.

Monte Carlo experiments vary list-experiment design choices and introduce small amounts of non-strategic respondent error to evaluate two estimators: the item count technique maximum likelihood estimator (ICT-MLE) and the simple difference-in-means. The simulations focus on conditions created by common best practices that reduce strategic misrepresentation and thus produce very few respondents at the extrema (choosing no items or all items).

How do we measure partisan balance in state governments?