Fear and Loathing across Party Lines: New Evidence on Group Polarization
American Journal of Political Science, 2014 · paper · Political science
Measures affective polarization between American partisans using implicit, explicit and behavioral indicators. Finds hostility to the opposing party is automatic, as strong as racial divides, and produces discrimination against opposing partisans exceeding that based on race.
Checked: PDF from web.stanford.edu, fetched 5 Oct 2026. Page numbers are those of that PDF.
11 findings
Substance Independents' net thermometer score is printed as "slightly slanted toward Democrats (mean = 3.52, s.e. = 0.84, n = 697), t(696) = 4.19", with a positive sign, although Democratic preference is negative everywhere else in the paragraph (–30.64, –48.84) and Figure 5 plots the Independent mean just left of zero.
- p. 7 Partisan Affect among Independents and Leaners
ward Democrats (mean = 3.52, s.e. = 0.84, n = 697)
- p. 7 same paragraph, Democratic leaners and Democrats
preference for Democrats (mean = –30.64, s.e. = 1.60
- p. 8 Figure 5, Independent row
Independent
- p. 11 Study 2 results, Independents
cratic skew in Independents’ implicit and explicit partisan
Judge's note: Page 7 image shows '3.52' with no minus sign while the neighbouring Democratic values carry one; Republican leaners and Republicans are positive (28.18, 42.58). Figure 5 on page 8 puts the Independent dot slightly below zero on the same axis, and page 11 refers to a 'small Democratic skew' among Independents. The magnitude matches the t statistic; only the sign is missing (mean and t should be negative on the paper's scale). The words give the right direction, the number does not.
Second check: another model agreed
Report a problem with this finding
Substance "the probability of a Republican selecting the Democrat was only .15 (95% confidence interval [.16, .35])": the estimate lies below the lower bound of its own interval, and Figure 6 shows an interval of roughly [.06, .34].
- p. 10 Study 2 Results, right column, first lines
only .15 (95% confidence interval [.16, .35])
- p. 10 Figure 6, panel 'Democrat More Qualified', Republican row
Democrat More Qualified
- p. 10 Results, Study 2, right column
was only .15 (95% confidence interval [.16, .35])
- p. 10 Figure 6, Democrat More Qualified panel, Republican row
Republican
Only the authors can say what is right here.
Judge's note: Text read on the page 10 crop: '.15' and '[.16, .35]' are as printed. The point estimate lies outside the printed bounds. In Figure 6, 'Democrat More Qualified' panel, Republican row, the dot is at about .85 for selecting the Republican with a bar from about .66 to .94, which gives about .15 with roughly [.06, .34] for selecting the Democrat. The other intervals in the passage contain their estimates. The lower bound looks misprinted (possibly .06); the authors should confirm the bounds. Figure readings are approximate to about .02.
Second check: another model agreed
Report a problem with this finding
Substance The overall means "a mean of $4.17 (95% confidence interval [3.91, 4.43]) in the trust game, and a mean of $2.88 (95% confidence interval [2.66, 3.10]) in the dictator game" are lower than Table 4 allows ($3.82, $3.14, $2.68, $3.36 dictator; $4.71, $4.08, $4.21, $4.45 trust) and coincide with the Opposing Partisan / Different Race points in Figure 8, so the "bonus" of 10% and 24% "over the average allocation" is relative to that cell, not to the average.
- p. 13 Study 3 Results
a mean of $2.88 (95% confidence interval [2.66, 3.10])
- p. 13 Study 3 Results
a mean of $4.17 (95%
- p. 14 Table 4, Dictator Game, Amount column
$3.82
- p. 13 Study 3 Results
representing a “bonus” of some 10% over the aver-
Only the authors can say what is right here.
Judge's note: Read pages 13, 14 and Figure 8 on page 11. The design is within-subjects with every participant playing two copartisan and two opposing-partisan rounds, so the overall mean is a mix of the per-party averages of Table 4: dictator 3.48 (Democrats) and 3.02 (Republicans), trust 4.395 and 4.33. Any weighting of the two parties gives at least 3.02 and 4.33, above the printed 2.88 and 4.17. In Figure 8 the Opposing Partisan / Different Race dots sit at about $2.88 (dictator) and $4.17 (trust), and the copartisan different-race trust dot at about $4.58 with the interval the text prints, so the text appears to report model cells (baseline and copartisan effect) as overall means. 4.58/4.17 gives the stated 10%. I found no statement in the text defining these as anything other than means of the endowment allocated. Only the authors can say what the intended figures are.
Second check: another model agreed
Report a problem with this finding
Substance Footnote 14: "were slightly more generous in the trust game (by a margin of 23 cents)"; Table 4 gives $4.21 (Republican to Democrat) against $4.08 (Democrat to Republican), a margin of 13 cents.
- p. 13 Footnote 14
margin of 23 cents
- p. 14 Table 4, Trust Game
$4.21
- p. 14 Table 4, Trust Game
$4.08
- p. 13 footnote 14
were slightly more generous in the trust game (by a margin of 23 cents)
Judge's note: Read footnote 14 on the page 13 image and Table 4 on page 14. The footnote's dictator figure (46 cents) is the Table 4 difference 3.14 − 2.68, so the footnote is built from the table; the trust difference is 0.13. The table's $4.21 and $4.08 are corroborated by the body text ($0.24 = 4.45 − 4.21, $0.63 = 4.71 − 4.08), and the body's 'about 20 cents less ... to opposing partisans' fits 13 cents (average 16.5) better than 23 (average 11.5). The footnote is the odd one out.
Second check: another model agreed
Report a problem with this finding
Substance Footnote 12, attached to "players chose not to profit maximize", says "Player 1’s dominant strategy in both games is to share with or entrust to Player 2 some amount of money", which states the opposite of the sentence it annotates.
- p. 13 Footnote 12
Player 1’s dominant strategy in both games is to share with or entrust to Player 2 some amount of money.
- p. 13 Study 3 Results
players chose not to profit maximize,
- p. 13 Footnote 12
dominant strategy in both games is to share with or
- p. 13 Study 3 Results
players chose not
Only the authors can say what is right here.
Judge's note: Read page 13 image. The body says players did not profit-maximise 'but rather' allocated nontrivial amounts, and page 12 says Player 1 allocates nontrivial amounts 'contrary to the axioms of rationality' and that in the dictator game Player 2 cannot return anything. By the paper's own account the profit-maximising strategy is therefore to give nothing; the footnote as printed says it is to give something. A negation ('no amount') appears to be missing; the authors should confirm the intended wording.
Second check: another model agreed
Report a problem with this finding
Reference or label The text cites "Nosek, Banaji, and Greenwald 1980"; the only reference-list entry by these authors is dated 2002 ("Harvesting Implicit Group Attitudes and Beliefs from a Demonstration Web Site").
- p. 4 Study 1 Design
naji, and Greenwald 1980)
- p. 18 References
wald. 2002. “Harvesting Implicit Group Attitudes and Beliefs
- p. 18 References, Nosek et al.
wald.2002.
- p. 4 Design
Nosek, Ba-naji, and Greenwald 1980
Judge's note: Page 4 image shows 'see Nosek, Banaji, and Greenwald 1980'. The page 18 crop shows the single Nosek, Banaji and Greenwald entry, dated 2002. No 1980 entry by these authors is in the list (the only 1980 entry is Crosby, Bromley and Saxe).
Second check: another model agreed
Report a problem with this finding
Reference or label The Study 2 passage on leaners' candidate selection points to "(top of Figure 8)"; Figure 8 is "Effects of Race and Partisanship on Allocations to Player 2" (Study 3 dictator and trust games) and has no leaner rows. The leaners are shown in Figure 6.
- p. 10 Study 2 Results, right column, paragraph on leaners
numbers of leaners in the sample (top of Figure 8)
- p. 11 Figure 8 caption
Effects of Race and Partisanship on Allocations to Player 2
- p. 10 Study 2 Results, right column
(topofFigure8)
- p. 10 Study 2 results, leaners paragraph
smaller numbers of leaners in the sample (top of Figure 8)
Judge's note: Page 10 image has the sentence 'due to smaller numbers of leaners in the sample (top of Figure 8)'. Page 11 image shows Figure 8 with only 'Opposing Partisan' and 'Copartisan' rows and dollar allocations. Figure 6 on page 10 has the 'Lean Republican' and 'Lean Democrat' rows with their wide intervals, so Figure 6 is the evident target. The run-together words in one quote are an extraction artefact; the figure number is on the page.
Second check: another model agreed
Report a problem with this finding
Reference or label The text cites a quotation to "(Grimmer and King 2011, 2649)", but the reference entry gives the pages as "Proceedings of the National Academy of Sciences 108(7): 264–350."; page 2649 is outside 264–350.
- p. 16 Discussion, right column
mer and King 2011, 2649)
- p. 17 References, Grimmer and King
108(7): 264–350.
- p. 17 References, Grimmer and King
108(7): 264–350
- p. 16 Discussion
Grimmer and King 2011, 2649
Only the authors can say what is right here.
Judge's note: Read the citation on the page 16 image (right column) and the reference entry on the page 17 image. Both are printed as quoted. The cited page 2649 does not fall in the printed range 264–350, so at least one of the two is wrong; this needs no outside knowledge. The reviewers' proposed true range (2643–2650) is not in the files and cannot be verified here, so the authors should say which is right.
Second check: another model agreed
Report a problem with this finding
Wording The D-score is "calculated by subtracting the mean response times for the round pairing targets from Category A (Democrat/Black) with positive terms from the mean response times for the round pairing targets from Category B (Republican/White) with positive terms", yet "Positive D-scores indicate that participants respond faster to Republican-good than to Democrat-good pairings"; the two sentences give opposite signs.
- p. 4 Study 1 Design, D-score paragraph
subtracting the mean response times for the round pair-
- p. 4 Study 1 Design, D-score paragraph
tive D-scores indicate that participants respond faster to
- p. 4 Design, D-score description
subtracting the mean response times for the round pairing targets from Category A (Democrat/Black) with positive terms from the mean response times for the round pairing targets from Category B (Republican/White)
- p. 4 Design
Positive D-scores indicate that participants respond faster to Republican-good than to Democrat-good pairings
Only the authors can say what is right here.
Judge's note: Read on the page 4 image. As described, D is proportional to RT(Republican+good) minus RT(Democrat+good), which is positive when Republican-good responses are slower, not faster. The stated interpretation and the results (Republicans positive, page 7) need the reverse subtraction. The results follow the interpretation, so the description of the subtraction appears to be the reversed part; the reported values are not affected.
Second check: another model agreed
Report a problem with this finding
Wording Discussion: "Our comparisons across indicators of implicit and explicit party affect revealed no differences in the degree of affective polarization"; Study 1 results say the variation "reveals differences in the extent of affective polarization across the implicit and explicit indicators" and report "Cohen's d is .95 for the implicit measure ... but it is even higher (1.72) in the case of the net thermometer score".
- p. 15 Discussion, right column
explicit party affect revealed no differences in the degree
- p. 6 Study 1 Results
even higher (1.72) in the case of the net thermometer
- p. 5 Study 1 Results
in the extent of affective polarization across the implicit
- p. 15 Discussion
revealed no differences in the degree of affective polarization
Only the authors can say what is right here.
Judge's note: Read the Discussion sentence on the page 15 image and the Cohen's d passage on the page 6 image; the page 5 sentence was read in the extracted text only. The results section says the two indicators differ in the extent of polarization (d .95 against 1.72, and the D-score explaining only 17.5% of thermometer variance), and the Discussion says they showed no differences in degree. The Discussion's own hedge ("if anything, conventional survey measures may be somewhat inflated") softens the sentence but does not remove the contradiction with the results wording. No statement elsewhere reconciles the two. Graded S3 because the numbers are not in question, only the Discussion's description of them.
Second check: another model agreed
Report a problem with this finding
Wording Study 4 is dated "in the winter of 2012" in its own section but footnote 11 says "Study 4, which was fielded in the winter of 2013" and the Discussion refers to "winter 2013 and spring 2013"; no study in the paper is described as fielded in spring 2013.
- p. 14 Study 4, first paragraph
in the winter of 2012
- p. 12 Footnote 11
in Study 4, which was fielded in the winter of 2013
- p. 16 Discussion, limitations paragraph
winter 2013 and
- p. 12 footnote 11
fielded in the winter of 2013
Only the authors can say what is right here.
Judge's note: Read page 14 image for 'winter of 2012'; footnote 11 (page 12) and the Discussion (page 16) checked in the text. The two years may describe one winter (December 2012 to early 2013), so this is an unresolved ambiguity and not a demonstrated wrong date, but the paper gives one study two different years. Stated dates elsewhere: Study 1 July 2012, Study 3 September 2012; Study 2 has no date, so 'spring 2013' has no stated referent.
Second check: another model agreed
Report a problem with this finding
23 minor (typos and the like; most checked by the judge only)
Wording The abstract says "affective polarization based on party is just as strong as polarization based on race" and the introduction says "just as significant as affect and discrimination based on race", while the introduction also says partisan animus "exceeds racial hostility" and the results say the race effect "(Cohen's d = .61) was not nearly as strong as the corresponding effect of party (Cohen's d = .95)".
- p. 1 Abstract
affective polarization based on party is just as strong as
- p. 7 Comparing Partisan and Racial Affect
race (Cohen’s d = .61) was not nearly as strong as the
- p. 2 left column
in the American public exceeds racial hostility
- p. 1 Abstract
just as strong as
Judge's note: Read the abstract on the page 1 crop and the page 7 results on the page image; the page 1 introduction and page 2 wording were checked in the extracted text. The paper states the headline comparison both as equality and as party exceeding race. 'Just as strong as' can be read as a cautious 'at least as strong', so this is a wording inconsistency, not a wrong number.
Second check: another model agreed
Report a problem with this finding
Wording "Outgroup animosity is more consequential than favoritism for the ingroup" is stated without qualification, but in the Study 4 dictator game the "copartisan bonus of $0.67 (95% confidence interval [.50, .84])" is slightly larger than the "penalty of $0.63 (95% confidence interval [–.80, –.46])"; the penalty exceeds the bonus only in the trust game ($0.63 against $0.30).
- p. 14 Study 4, end of section
Outgroup animosity is more consequential than
- p. 14 Study 4, results paragraph
tisan bonus of $0.67 (95% confidence interval [.50, .84])
- p. 14 Study 4, results paragraph
penalty of $0.63 (95% confidence interval [–.80, –.46])
- p. 2 Introduction, 'Second, through three experiments' paragraph
outgroup animus than ingroup favoritism
Only the authors can say what is right here.
Judge's note: Read page 14 image. The four Study 4 effects are as quoted: dictator bonus 0.67 and penalty 0.63, trust bonus 0.30 and penalty 0.63. The surrounding paragraph is about the trust game, so the authors may mean the closing sentence for that game only, but neither it nor the introduction's 'based more on outgroup animus than ingroup favoritism' (page 2) says so. An overstatement in wording; the numbers are not in question.
Second check: another model agreed
Report a problem with this finding
Wording The Discussion says the party-exceeds-race pattern "applies to both explicit and implicit measures of group affect", but the paper reports an explicit measure (feeling thermometers) for parties only; race is compared with party only through the BIATs and the behavioural tasks.
- p. 14 Discussion, first paragraph
explicit and implicit measures of group affect
- p. 2 Introduction
fair comparison, we use implicit measures
Judge's note: Read page 14 image and searched the text for 'explicit' and 'thermometer'. Study 1 compares the partisan and race D-scores (implicit); the explicit thermometer is the difference between ratings of Democrats and Republicans, and the paper says it uses implicit measures to avoid under-reporting on 'explicit questions about race relations'. No explicit racial-affect result is given. The authors may count the overt choices in Studies 2 and 3 as explicit, so this is a wording point.
Second check: another model agreed
Report a problem with this finding
Wording The thermometer analysis "among Independents in the SSI sample" reports group sizes n = 1,340, 297, 697, 222 and 624, which sum to 3,180, more than the Study 1 sample ("a sample of 2,000 adults", "left with a sample of 1,971"); Democrats number 1,340 here against 374 + 466 = 840 weak and strong Democrats in the D-score analysis.
- p. 7 Partisan Affect among Independents and Leaners
Democrats (mean = –48.84, s.e. = 0.93, n = 1,340)
- p. 7 same paragraph
mean = 3.52, s.e. = 0.84, n = 697
- p. 4 Study 1 Design and footnote 2
We recruited a sample of 2,000 adults
- p. 5 Study 1 Results
s.e. = .02, n = 374
Only the authors can say what is right here.
Judge's note: Read on the page 4, 5, 7 and 8 images. I searched the article text for a statement that this section pools respondents across studies or uses another sample and found none; the section sits inside Study 1 and says only 'the SSI sample'. The article says 'Measure and sample information are included in the supporting information', which is not among the files and may explain the larger n (for example pooling the SSI samples of several studies), so this is a question for the authors, not a demonstrated error. The second reviewer point (explicit polarization scores for left and right 'generally indistinguishable' against thermometer means of 48.84 and 42.58) is a loosely worded discussion claim with no test reported; not confirmed as a separate defect.
Second check: another model agreed
Report a problem with this finding
Wording Discussion: "our results replicate during times of relatively low political conflict (i.e., winter 2013 and spring 2013)"; no study is reported as fielded in spring 2013, and Study 4 is dated "the winter of 2013" in footnote 11 but "the winter of 2012" in the Study 4 text.
- p. 16 Discussion, limitations paragraph
spring 2013) and times of stronger political conflict
- p. 12 Footnote 11
We fielded the study in September 2012
- p. 4 Footnote 1
The study was fielded in July 2012.
Only the authors can say what is right here.
Judge's note: Read the Discussion sentence on the page 16 image and searched the extracted text for every fielding date. Those stated are July 2012 (Study 1, footnote 1, and a pretest in footnote 11), September 2012 (Study 3, footnote 11), and Study 4, given as "winter of 2013" in footnote 11 and "winter of 2012" in the Study 4 section. No date is given for Study 2 and nothing is said to have run in spring 2013. Study 2 may have been fielded then, but the paper does not say so; the supporting information is not among the files and might settle it. The two dates for Study 4 could be one winter spanning the year end. The fielding-date passages on pages 4, 12 and 14 were read in the extracted text only.
Second check: another model agreed
Report a problem with this finding
Wording "There were traces of intransitivity in the explicit indicator, as strong Democrats were slightly less biased against Republicans than weak Democrats (mean = –.21, s.e. = .02, n = 466)": the means compared are partisan D-scores, which the paper calls its implicit measure.
- p. 5 Study 1 Results, first paragraph
There were traces of intransitivity in the explicit indicator, as strong Democrats were slightly
- p. 5 Study 1 Results, first paragraph
the explicit indicator, as strong Democrats were slightly
- p. 5 Study 1 results
There were traces of intransitivity in the explicit indicator
- p. 5 Study 1 Results
traces of intransitivity in the explicit indicator
Only the authors can say what is right here.
Judge's note: Read on the page 5 image. The values –.26 and –.21 are D-score means from Figure 2, top panel (implicit). The paper elsewhere uses 'explicit' for the feeling thermometer, introduced only in the next paragraph. 'Explicit indicator' could conceivably mean the self-reported party-strength scale against which the D-score is non-monotonic, so the authors should say which is meant; 'implicit' looks intended.
Second check: another model agreed
Report a problem with this finding
Wording "Democratic leaners are significantly less affectively polarized than self-identified Democrats, t(466.93)= –15.02, p < .001), but significantly more polarized than Independents, t(516.52) = 9.98, p < .001": the two statistics do not correspond to the comparisons they are attached to when recomputed from the printed means, standard errors and group sizes; they appear to be transposed.
- p. 7 Partisan Affect among Independents and Leaners
t(466.93)= –15.02, p < .001), but significantly more polarized than Independents, t(516.52) = 9.98, p < .001
- p. 7 Partisan Affect among Independents and Leaners
t(466.93)= –15.02, p < .001), but significantly more polarized than Independents, t(516.52) = 9.98
- p. 7 Partisan Affect among Independents and Leaners, right column
–15.02, p < .001), but significantly more polarized
- p. 7 Partisan Affect among Independents and Leaners
self-identified Democrats, t(466.93)= –15.02, p < .001), but significantly more polarized than Independents, t(516.52) = 9.98
Only the authors can say what is right here.
Judge's note: Read on the page 7 image. The paper does not name the test, but the fractional degrees of freedom indicate unequal-variance (Welch) t-tests. From the printed values, leaners vs Democrats gives t about 9.83 with about 517 degrees of freedom, and leaners vs Independents (taking the Independent mean as –3.52, see C007) gives t about 15.01 with about 467 degrees of freedom. The printed pairing is the reverse. The same recomputation reproduces the Republican-side statistic in its printed place (14.40/2.27 = 6.34 against 6.33), so the method fits the paper's own numbers. Both P values are below .001 either way. The stray closing parenthesis raised by one reviewer is covered in C011.
Second check: another model agreed
Report a problem with this finding
Wording "Democratic leaners (b = –1.35, s.e. = .33, p < .05)" prints the same standard error as "Democrat (b = –1.04, s.e. = .33, p < .01)" although the text says leaners are fewer and more variable and Figure 6 shows a much wider interval for Lean Democrat; could the authors check the standard error against the full model in the supporting information?
- p. 10 Study 2 Results, leaners paragraph
Democratic leaners (b = –1.35, s.e. = .33, p < .05)
- p. 10 Study 2 Results, partisans paragraph
Democrat (b = –1.04, s.e. = .33, p < .01)
Only the authors can say what is right here.
Judge's note: Read page 10 image and crop. The four coefficients are printed with s.e. and a significance level: b/s.e. is 3.15 for Democrats (p < .01), 4.10 for Republicans (p < .001), 2.02 for Republican leaners (p < .05) and 4.09 for Democratic leaners, which is nevertheless marked only p < .05, the level used for a ratio near 2. The printed p < .05 is not false under either reading, so this is not a P-value contradiction; it is a question about the s.e. The same sentence says leaners show 'larger variance' because of smaller numbers, yet the s.e. equals the Democrats' value. In Figure 6 (equally qualified panel) the Lean Democrat bar runs from roughly .06 to .36, against the printed [.14, .29] for Democrats; figure values are read by eye. The full model is in the supporting information (footnote 8), which is not among the files and would settle it.
Second check: another model agreed
Report a problem with this finding
Wording "three times as large for Democrats ($0.63) than Republicans ($0.24) in the trust game": the ratio of the printed values is about 2.6.
- p. 13 Results, Study 3
three times as large for Democrats ($0.63)
- p. 14 Table 4, trust game
$4.71 ... $4.08 ... $4.45 ... $4.21
Judge's note: Page 13 text and Table 4 on page 14: 4.71 − 4.08 = 0.63 and 4.45 − 4.21 = 0.24, both as printed; 0.63/0.24 = 2.63. Loose rounding in the wording; the numbers themselves are right.
Second check: another model agreed
Report a problem with this finding
Wording "with the opposing partisan penalty amounting to almost twice the copartisan bonus": with the printed trust-game values ($0.63 penalty, $0.30 bonus) the ratio is 2.1, slightly more than twice.
- p. 14 Study 4, right column
amounting to almost twice the copartisan bonus
- p. 14 Study 4, results paragraph
$0.30 (95% confidence interval
Judge's note: Page 14 image. The sentence is in the trust-game discussion, so the dictator-game ratio (0.63/0.67) is not the comparison meant. 0.63/0.30 = 2.1; unrounded values could give a ratio just under 2, so either form could be legitimate.
Second check: another model agreed
Report a problem with this finding
Typo The last sentence of Study 4, "Outgroup animosity is more consequential than favoritism for the ingroup", ends without a full stop; the Greenwald, McGhee and Schwartz (1998) reference also lacks the period before the closing quotation mark ("The Implicit Association Test” Journal of").
- p. 14 last sentence before Discussion
favoritism for the ingroup
- p. 14 Study 4 results
Outgroup animosity is more consequential than favoritism for the ingroup
- p. 14 Study 4 results
copartisan bonus of $0.67
- p. 2 Introduction
discrimination against the outgroup is based more on outgroup animus than ingroup favoritism
Judge's note: Page 14 image: the sentence ends at 'ingroup' with no punctuation before the Discussion heading. Reference crop on page 17: other entries close the title with a period inside the quotation mark, this one does not. The cluster also carries points judged elsewhere: the substantive claim about animosity versus favouritism (C028), 'singe' (C017) and 'patterns applies' (C027).
Second check: none (judge only)
Report a problem with this finding
Typo Missing space: "implement theBIAT in a web browser".
- p. 4 right column, top
implement theBIAT in a web browser
Judge's note: Visible on the page 4 image, top of the right column: 'theBIAT' is set as one word.
Second check: none (judge only)
Report a problem with this finding
Typo Figure 4, top panel, x-axis label is misspelled "Partisn D-Score".
- p. 7 Figure 4, top panel, x-axis label
Partisn D-Score
Judge's note: Visible in the figure itself on the page 7 image; Figure 2 on page 5 has 'Partisan D-Score'.
Second check: none (judge only)
Report a problem with this finding
Typo Unmatched closing parentheses after two test statistics: "t(466.93)= –15.02, p < .001), but" and "t(461.55) = 6.33, p < .001), and significantly".
- p. 7 Partisan Affect among Independents and Leaners
–15.02, p < .001), but
- p. 8 top of left column
6.33, p < .001), and significantly
- p. 8 top of left column (continuation of p 696)
t(461.55) = 6.33, p < .001), and significantly more polarized
Judge's note: Checked on the page 7 and page 8 images. In both places the statistic is introduced with a comma and no opening parenthesis (on page 8 the '(mean = 42.58 ... n = 624)' bracket is already closed), so the ')' after 'p < .001' has no partner.
Second check: none (judge only)
Report a problem with this finding
Typo Figure 4 note: "This figure shows the distributions of implicit partisan affect (top) and implicit racial affect (bottom) with 95% confidence intervals", but the panels plot group means with interval bars, not distributions.
- p. 7 Figure 4 note
shows the distributions of implicit partisan affect
Judge's note: Page 7 image: each row of Figure 4 is one dot with an error bar. The notes to Figures 2 and 5 say 'The dots are the means for each group'. Loose wording in the note; nothing in the data is affected.
Second check: none (judge only)
Report a problem with this finding
Typo "This remarkable patterns applies to both explicit and implicit measures": should be 'pattern'. Also the x-axis label of the upper panel of Figure 4 reads "Partisn D-Score" for 'Partisan D-Score'.
- p. 14 Discussion, first paragraph
This remarkable patterns applies to both
- p. 14 Discussion
This remarkable patterns applies
- p. 10 footnote 9
a singe “equally qualified” condition
- p. 7 Figure 4 axis label
Partisn D-Score
Judge's note: Page 14 image shows 'This remarkable patterns applies'. The Figure 4 crop on page 7 shows the axis label 'Partisn D-Score'. The third item in the cluster, 'singe', is C017.
Second check: none (judge only)
Report a problem with this finding
Typo Footnote 9: "combined to produce a singe “equally qualified” condition"; 'singe' should be 'single'.
- p. 10 Footnote 9
combined to produce a singe
Judge's note: Visible on the page 10 image in footnote 9.
Second check: none (judge only)
Report a problem with this finding
Typo "additional evidence that lea ners are in fact closet partisans": the word 'leaners' is printed with a space inside it.
- p. 10 Study 2 Results, leaners paragraph
lea ners are in
Judge's note: The crop shows a clear gap between 'lea' and 'ners', wider than the letter spacing elsewhere on the line; it is in the typeset page, not only the text layer.
Second check: none (judge only)
Report a problem with this finding
Typo Figure 7 labels the participant group "White" on the y-axis while the facet title in the same figure and the accompanying text say "European American".
- p. 11 Figure 7, y-axis category
White
Judge's note: Page 11 image: y-axis categories are 'White' and 'African American'; the third facet is 'European American More Qualified' and the text on the page uses 'European American' throughout. Inconsistent naming for one group, no effect on the data; the paper also uses 'white' for Player 2 in Study 3, so either term is legitimate.
Second check: none (judge only)
Report a problem with this finding
Typo "discourage the manifestation of behavior that may be constructed as discriminatory": "constructed" should be "construed".
- p. 15 Discussion, left column
festation of behavior that may be constructed as discrim-
Judge's note: Read on the page 15 image, left column: the word printed is "constructed". The sentence is about egalitarian norms discouraging behaviour that could be interpreted as discriminatory, so "construed" is the word meant. Wrong word, cosmetic.
Second check: none (judge only)
Report a problem with this finding
Typo Reference entry reads "Fowler, John H., and Cindy D. Kam. 2007. 'Beyond the Self: Social Identity, Altruism, and Political Participation.' Journal of Politics 69(3): 813–27."; the first author's given name appears to be James, not John.
- p. 17 References, Fowler and Kam
Fowler, John H., and Cindy D. Kam. 2007.
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: that the author of this Journal of Politics article is James H. Fowler. The entry is printed as quoted on the page 17 image. Nothing in the files gives the author's first name elsewhere, so it cannot be verified here.
Second check: none (judge only)
Report a problem with this finding
Typo Reference-list slips: Johnson and Mislin URL printed "http://ssrncom/abstract1315325." (dot and "=" missing); "Kirchheimer, Otto. (1966)." is the only entry with the year in parentheses; Greenwald, McGhee and Schwartz title ends "The Implicit Association Test” Journal of" with no full stop; Cai et al. title reads "The Implicit Association Tests D Measure" (apostrophe missing).
- p. 17 References, Johnson and Mislin
http://ssrncom/abstract1315325.
- p. 18 References, Kirchheimer
Kirchheimer, Otto. (1966).
- p. 17 References, Greenwald, McGhee and Schwartz
The Implicit Association Test” Journal of
- p. 17 References, Cai et al.
“The Implicit Association Tests D Measure
Judge's note: All four are visible as quoted on the page 17 and page 18 images, so none is an extraction artefact. The URL as printed is not a valid address. The further point in one note (Klofstad et al. dated 2012 with Political Behavior 35(3)) rests on outside knowledge of that volume's year and is not confirmed here.
Second check: none (judge only)
Report a problem with this finding
Typo Reference entry reads "Cohen, Jack. 1988. Statistical Power Analysis for the Behavioral Sciences."; the author's name is usually given as Jacob Cohen.
- p. 17 References
Cohen, Jack. 1988.
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: that this book is published under the name Jacob Cohen. The entry is printed as "Cohen, Jack" on the page 17 image. "Jack" is a familiar form of the same name, so this is cosmetic.
Second check: none (judge only)