Cited by 179Psychometric Methodologies and TestingSocial and Intergroup PsychologyPsychological Testing and Assessment

The DOI matches its Crossref record; 45 of the 49 references it cites were found in a registry or an index.

Understanding and Using the Brief Implicit Association Test: Recommended Scoring Procedures

Open access
Nosek, Brian A., Bar-Anan, Yoav, Sriram, N., Axt, Jordan, Greenwald, Anthony G.·2014·PLoS ONE

Published version of a preprint

Abstract

A brief version of the Implicit Association Test (BIAT) has been introduced.The present research identified analytical best practices for overall psychometric performance of the BIAT.In 7 studies and multiple replications, we investigated analytic practices with several evaluation criteria: sensitivity to detecting known effects and group differences, internal consistency, relations with implicit measures of the same topic, relations with explicit measures of the same topic and other criterion variables, and resistance to an extraneous influence of average response time.The data transformation algorithms D outperformed other approaches.This replicates and extends the strong prior performance of D compared to conventional analytic techniques.We conclude with recommended analytic practices for standard use of the BIAT. Understanding and Using the Brief Implicit Association Test: I. Recommended Scoring ProceduresEven the most brilliant research ideas can flounder if data collection procedures and data analytic strategies applied in the pursuit of these ideas are suboptimal.Research efficiency, the knowledge gained in proportion to resources invested, can be improved by maximizing the quality of procedural and analytical methods.Numerous paradigms in mental chronometry, such as Stroop and lexical-decision tasks, define constructs derived from contrasting response latencies between performance conditions [1].The Implicit Association Test (IAT) [2] is a

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1371/journal.pone.0110938

Cites

45 of 45 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Schematics of the same single response trial of one block of the IAT on the left, and the BIAT on the right. In the IAT, the correct response is the left key because Awful belongs to the Category Bad. In the BIAT, the correct response is the left key because Awful does not belong to the categories Democrats or Good.
Figure 1 Schematics of the same single response trial of one block of the IAT on the left, and the BIAT on the right. In the IAT, the correct response is the left key because Awful belongs to the Category Bad. In the BIAT, the correct response is the left key because Awful does not belong to the categories Democrats or Good.CC BY · PubMed Central
Figure 2 Latency operating characteristics, showing variation in mean standardized values of the five candidate BIAT scoring algorithms across deciles of the sample's distribution of average speed of responding for the political BIAT. For this plot, the algorithms were computed after deleting 4 warm-up trials from each response block and also deleting latencies greater than 10,000 ms. There were 202 or 203 respondents in each decile. Most noticeable in the graph is the inferior performance (smaller effect sizes) for the reciprocal measure, and strongest performance for the D measure. Also noticeable is that the D measure was smallest for the slowest subjects, whereas the log and latency measures were largest for the slowest subjects.
Figure 2 Latency operating characteristics, showing variation in mean standardized values of the five candidate BIAT scoring algorithms across deciles of the sample's distribution of average speed of responding for the political BIAT. For this plot, the algorithms were computed after deleting 4 warm-up trials from each response block and also deleting latencies greater than 10,000 ms. There were 202 or 203 respondents in each decile. Most noticeable in the graph is the inferior performance (smaller effect sizes) for the reciprocal measure, and strongest performance for the D measure. Also noticeable is that the D measure was smallest for the slowest subjects, whereas the log and latency measures were largest for the slowest subjects.CC BY · PubMed Central
Figure 3 Latency operating characteristics, showing variation in mean standardized values of the five candidate BIAT scoring algorithms across deciles of the sample's distribution of average speed of responding for the political BIAT. Pretreatment of the data involved removing 4 warm-up trials per block, latencies slower than 10s, and latencies faster than 400 ms. There were 202 or 203 respondents in each decile. The most noticeable effects visible in the graph are improvement in performance of the reciprocal measure relative to its poor showing in Fig. 2, and the contrast between the relative stability across speed variations for four of the measures and the increasing magnitude of the (untransformed) latency-difference measure as responding went from fast (left of graph) to slow.
Figure 3 Latency operating characteristics, showing variation in mean standardized values of the five candidate BIAT scoring algorithms across deciles of the sample's distribution of average speed of responding for the political BIAT. Pretreatment of the data involved removing 4 warm-up trials per block, latencies slower than 10s, and latencies faster than 400 ms. There were 202 or 203 respondents in each decile. The most noticeable effects visible in the graph are improvement in performance of the reciprocal measure relative to its poor showing in Fig. 2, and the contrast between the relative stability across speed variations for four of the measures and the increasing magnitude of the (untransformed) latency-difference measure as responding went from fast (left of graph) to slow.CC BY · PubMed Central
Figure 4 Effects of seven criteria for excluding respondents as a function of their proportion of fast responses (latency <300 ms) on correlations with self-reported preference between Democrats and Republicans for five BIAT data transformations (Study 5). Higher correlations indicate better performance. The furthest left datapoint indicates no exclusion of participants; the furthest right datapoint indicates exclusion of all participants that had even a single fast response. Sample size (n) on the x-axis indicates the number of participants retained with that exclusion criterion.
Figure 4 Effects of seven criteria for excluding respondents as a function of their proportion of fast responses (latency <300 ms) on correlations with self-reported preference between Democrats and Republicans for five BIAT data transformations (Study 5). Higher correlations indicate better performance. The furthest left datapoint indicates no exclusion of participants; the furthest right datapoint indicates exclusion of all participants that had even a single fast response. Sample size (n) on the x-axis indicates the number of participants retained with that exclusion criterion.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Research works printed but not linked: 1

  1. Greenwald AG, Nosek BA (2008) Attitudinal dissociation: What does it mean? In: Petty RE, Fazio RH, Brinol Peditors. Attitudes: Insights from the new implicit measures. Hillsdale, NJ: Erlbaum. pp.65–82.

Cited by

Sorted by citations

See all 179

Metadata

Authors
Nosek, B. A., Bar-Anan, Y., Sriram, N., Axt, J., & Greenwald, A. G.
Affiliations
  • Center for Open Science (US)
  • University of Virginia (US)
  • University of Washington (US)
Corresponding author
Nosek, Brian A.
Venue
PLoS ONE
Published
2014-12-08
Volume
9
Issue
12
Pages
e110938
ISSN
1932-6203 (ISSN-L)
OpenAlex
W3124598427
Cited by
179
References
49
Citations by year
2026: 10 · 2025: 15 · 2024: 13 · 2023: 13 · 2022: 19 · 2021: 15 · 2020: 20 · 2019: 9 · 2018: 20 · 2017: 17 · 2016: 13 · 2015: 5 · 2014: 4
FWCI
16.42
Type
journal article
Language
en
Open access
Locations
  • PLoS ONE (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
  • Figshare (submittedVersion, cc-by)
Topics
  • Social Sciences › Decision Sciences › Management Science and Operations Research › Psychometric Methodologies and Testing
  • Social Sciences › Sociology and Political Science › Social and Intergroup Psychology
  • Social Sciences › Psychology › Applied Psychology › Psychological Testing and Assessment
Keywords
Brief Implicit Association Test, Implicit Association Test, implicit measures, scoring methods, reaction time (psychology), internal consistency, explicit measures, psychometric properties, response latency, group differences
MeSH terms
Attitude, Female, Humans, Male, Psychological Tests, Comprehension

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review