PPBS

Prejudice [2016]

A Politico-Psychological Analysis

Report generated

March 23, 2026

1 Study Characteristics

1.1 Items: Prejudice

1.2 Sample

N=1500

To conduct a exploratory and a confirmatory large surveys during the general election, we hired a professional survey firm (SSI, a US-based market research company that recruits participants from a panel of 7,139,027 American citizens; more information can be found at www.surveysampling.com (now https://www.dynata.com/) to recruit a nationally representative sample of 1,500 Americans (50.7% women) who completed study materials during the general election from August 16-September 9, 2016. (Information about sampling and exclusion criteria is included in the Supplement). The age distribution was as follows: 18-24 (12.9%), 25-34 (17.6%), 35-44 (17.5%), 45-54 (19.5%), 55-65 (15.6%) and older than 65 (16.9%). The ethnic breakdown was: White/European American (82.5%), Black/African American (7.7%), Latino (5.9%) and “Other” (4.0%). Concerning religion, 67.6% identified as Christian, 17.1% as religiously affiliated but not Christian, and 15.3% as Atheist/Agnostic. With respect to education 35.1% indicated “high school only or lower,” 31.4 % indicated “some college,” and 33.6% indicated having received a “Bachelor” or “Graduate” degree. 2424 participants were directed to the survey, 1885 of which finished the survey (attrition rate 22%).

We followed recommendations to minimize the problem of careless responding in online studies. Specifically, we employed 10 random attention questions and time controls to check for data quality. There were 385 participants who failed more than one attention check or finished the survey in under ~22 minutes and were therefore excluded from the sample. For the 1500 participants who successfully finished the survey, completion time was 67 minutes on average (MD: 51min).

2 Descriptives

2.1 Means, SD, Range, & SE

Table 1: Descriptives
Descriptives for Prejudice
mean sd median se
Homo_negativity_mean 4.98 2.68 5.00 0.07
Symbolic_Racism_mean 4.58 1.13 4.75 0.03
Note:
n = 1500, min = 1, max = 9
Table 2: Descriptives – Item Level
Descriptives for Prejudice at Item Level
mean sd median se
Homo_negativity1 4.57 2.89 5.00 0.07
Homo_negativity2 4.99 2.91 5.00 0.08
Homo_negativity3 5.37 2.76 5.00 0.07
Symbolic_Racism1 4.18 2.62 4.00 0.07
Symbolic_Racism2 4.85 2.68 5.00 0.07
Symbolic_Racism3 4.26 2.77 4.00 0.07
Symbolic_Racism4 5.03 2.62 5.00 0.07
Homo_negativity_mean 4.98 2.68 5.00 0.07
Symbolic_Racism_mean 4.58 1.13 4.75 0.03
Note:
All items: n = 1500, min = 1, max = 9. Homo_negativity_mean: min = 1, max = 9. Symbolic_Racism_mean: min = 1, max = 9

2.2 Response Distribution

Figure 1: Proportion of each facet of Prejudice at item level: Homo_negativity
Figure 2: Proportion of each facet of Prejudice at item level: Symbolic_Racism
Figure 3: Homo_negativity
Figure 4: Symbolic_Racism

2.3 Correlations

2.3.1 Correlation Plots

Figure 5: Bivariate Spearman correlations: Prejudice
Figure 6: Bivariate Spearman correlations: Homo_negativity
Figure 7: Bivariate Spearman correlations: Symbolic_Racism

Note. Upper triangle shows Spearman’s rank-order correlations. *** p < .001, ** p < .01, * p < .05.

2.3.2 PIDF

Figure 8: Bivariate Spearman correlations by party identity: Prejudice
Figure 9: Bivariate Spearman correlations by party identity: Homo_negativity
Figure 10: Bivariate Spearman correlations by party identity: Symbolic_Racism

Note. Upper triangle shows Spearman’s rank-order correlations by party identity. *** p < .001, ** p < .01, * p < .05.

2.3.3 Correlation Matrix

3 Demographics

3.1 Social Class

Figure 11: Homo_negativity
Figure 12: Symbolic_Racism
Figure 13: Homo_negativity
Figure 14: Symbolic_Racism

Note on the Raincloud Plots

  • Statistical summary (top): Welch’s t-test (or ANOVA) results, effect size, confidence intervals, p-values, and sample sizes are shown above each plot.
  • Bayesian analysis (bottom): Log Bayes factor and credible intervals are reported below each plot.
Table 3
Table 4: Symbolic_Racism as grouped by SES
(a)
SES N Mean SD
Poor 38 -0.18 1.13
Lower Middle Class 298 0.07 0.99
Middle Middle Class 679 -0.04 1.01
Upper Middle Class 395 0.02 0.98
Rich 90 0.06 0.97
(b)
SES N Mean SD
Poor 38 0.19 0.86
Lower Middle Class 298 -0.07 1.07
Middle Middle Class 679 0.00 1.03
Upper Middle Class 395 0.04 0.91
Rich 90 0.00 0.99

3.2 Gender

Figure 15: Prejudice Facets, as grouped by Gender: Homo_negativity
Figure 16: Prejudice Facets, as grouped by Gender: Symbolic_Racism
Figure 17: Homo_negativity
Figure 18: Symbolic_Racism
Table 5
Table 6: Symbolic_Racism as grouped by Gender
(a)
Gender N Mean SD
Female 760 0.25 1.00
Male 740 -0.25 0.93
(b)
Gender N Mean SD
Female 760 0.03 0.99
Male 740 -0.03 1.01

3.3 Age

Figure 19: Homo_negativity
Figure 20: Symbolic_Racism
Figure 21: Homo_negativity
Figure 22: Symbolic_Racism
Table 7
Table 8: Symbolic_Racism as grouped by Age
(a)
Age N Mean SD
18-24 years 193 0.19 0.95
25-34 years 264 0.30 0.98
35-44 years 263 0.16 1.02
45-54 years 292 -0.06 0.95
55-64 years 234 -0.07 0.97
65+ 254 -0.49 0.93
(b)
Age N Mean SD
18-24 years 193 -0.01 1.03
25-34 years 264 0.05 0.96
35-44 years 263 -0.06 1.04
45-54 years 292 0.02 1.05
55-64 years 234 0.07 0.91
65+ 254 -0.06 1.00

3.4 Education

Figure 23: Homo_negativity
Figure 24: Symbolic_Racism
Figure 25: Homo_negativity
Figure 26: Symbolic_Racism
Table 9
Table 10: Symbolic_Racism as grouped by Education
(a)
Education N Mean SD
Less than High School 51 -0.02 1.07
High School 475 -0.05 0.98
Some College 471 -0.01 1.03
Bachelor 310 0.04 0.97
Graduate 193 0.10 1.00
(b)
Education N Mean SD
Less than High School 51 -0.07 0.98
High School 475 -0.03 1.09
Some College 471 -0.03 1.01
Bachelor 310 0.09 0.84
Graduate 193 0.00 0.97

3.5 Income Levels

Figure 27: Homo_negativity
Figure 28: Symbolic_Racism
Figure 29: Homo_negativity
Figure 30: Symbolic_Racism
Table 11
Table 12: Symbolic_Racism as grouped by Income Levels
(a)
Income Levels N Mean SD
Less than $15,000 178 0.00 1.03
$15,000-$24,999 180 -0.01 1.05
$25,000-$34,999 176 -0.02 1.03
$35,000-$49,999 227 0.01 1.00
$50,000-$74,999 292 0.08 1.00
$75,000-$99,999 192 -0.06 0.93
$100,000-$149,999 160 -0.03 1.01
$150,000 + 95 -0.03 0.95
(b)
Income Levels N Mean SD
Less than $15,000 178 -0.07 0.99
$15,000-$24,999 180 -0.03 1.07
$25,000-$34,999 176 -0.02 1.06
$35,000-$49,999 227 -0.10 0.97
$50,000-$74,999 292 0.03 0.99
$75,000-$99,999 192 0.10 0.92
$100,000-$149,999 160 -0.01 1.11
$150,000 + 95 0.15 0.80

3.6 Ethnicity

Figure 31: Homo_negativity
Figure 32: Symbolic_Racism

Note on the Okabe-Ito color palette. The Okabe-Ito color palette (seen above) is a set of colorblind-friendly categorical colors available in R. We are using this palette for graphs with non-ordered variables (e.g., groups, categories) for accessibility.

Figure 33: Homo_negativity
Figure 34: Symbolic_Racism
Table 13
Table 14: Symbolic_Racism as grouped by Ethnicity
(a)
Ethnicity N Mean SD
Caucasian/European origin 1237 -0.04 1.01
Black/African American 115 0.18 0.97
Latino 88 0.13 0.91
Asian/Pacific Islander 29 0.25 0.82
Native American 13 -0.08 0.97
Other 18 0.35 1.08
(b)
Ethnicity N Mean SD
Caucasian/European origin 1237 0.02 0.98
Black/African American 115 -0.19 1.00
Latino 88 -0.15 0.99
Asian/Pacific Islander 29 0.11 1.15
Native American 13 0.30 1.70
Other 18 -0.16 1.24

3.7 Occupation

Figure 35: Homo_negativity
Figure 36: Symbolic_Racism
Figure 37: Homo_negativity
Figure 38: Symbolic_Racism
Table 15
Table 16: Symbolic_Racism as grouped by Occupation
(a)
Occupation N Mean SD
Employed 768 0.06 0.99
Retired 268 -0.33 0.94
Unemployed 146 0.14 0.92
Parent 104 0.06 1.05
Disabled 98 -0.11 1.03
Student 85 0.31 1.06
Full-time caregiver 31 -0.04 1.06
(b)
Occupation N Mean SD
Employed 768 0.02 0.98
Retired 268 -0.11 1.00
Unemployed 146 0.03 1.02
Parent 104 0.00 1.05
Disabled 98 0.08 1.08
Student 85 0.01 0.97
Full-time caregiver 31 0.00 1.00

3.8 Area

Figure 39: Homo_negativity
Figure 40: Symbolic_Racism
Figure 41: Homo_negativity
Figure 42: Symbolic_Racism
Table 17
Table 18: Symbolic_Racism as grouped by Area
(a)
Area N Mean SD
Urban 955 0.09 0.99
Rural 545 -0.16 1.00
(b)
Area N Mean SD
Urban 955 0.01 1.00
Rural 545 -0.01 0.99

3.9 Religious Affiliation

Figure 43: Homo_negativity
Figure 44: Symbolic_Racism
Figure 45: Homo_negativity
Figure 46: Symbolic_Racism
Table 19
Table 20: Symbolic_Racism as grouped by Religious Affiliation
(a)
Religious Affiliation N Mean SD
Christian 1014 -0.20 0.95
Muslim 9 0.40 0.92
Jewish 52 0.21 1.06
Atheist/Agnostic 230 0.62 0.92
Not listed, No religion 195 0.25 0.94
(b)
Religious Affiliation N Mean SD
Christian 1014 -0.03 1.00
Muslim 9 -0.69 1.16
Jewish 52 0.18 0.76
Atheist/Agnostic 230 -0.01 0.97
Not listed, No religion 195 0.13 1.07

3.10 Section Summary

  Prejudice
Predictors Estimates CI p
(Intercept) 5.65 5.37 – 5.94 <.001
Age -0.16 -0.20 – -0.11 <.001
Income 0.02 -0.02 – 0.06 .262
Education 0.12 0.05 – 0.19 .001
Gender (Male) -0.54 -0.68 – -0.41 <.001
Observations 1500
R2 / R2 adjusted .086 / .084

4 Political Behavior

4.1 Political Orientation

Correlation with General Conservatism, Economic Conservatism, Social Conservatism

Figure 47: Homo_negativity
Figure 48: Symbolic_Racism
Figure 49: Homo_negativity
Figure 50: Symbolic_Racism
Figure 51: Homo_negativity
Figure 52: Symbolic_Racism

Note. The y-axis shows residuals after regressing the outcome on demographics (Age, Income, Education, Gender). This illustrates how much the focal construct explains beyond demographic factors.

Table 21: Prejudice and Political Orientation
Table 22: Prejudice models and Demographics

4.2 Partisanship

Figure 53: Homo_negativity
Figure 54: Symbolic_Racism
Table 23: Mean Homo_negativity by Party Identity (raw means)
Party Identity N Mean SD
Strong Democrat 351 6.71 2.46
Democrat 223 5.62 2.54
Leaning Democrat 124 6.38 2.27
Independent 84 5.17 2.16
Leaning Republican 111 3.95 2.28
Republican 255 4.05 2.19
Strong Republican 318 3.12 2.15
Table 24: Mean Symbolic_Racism by Party Identity (raw means)
Party Identity N Mean SD
Strong Democrat 351 4.63 1.13
Democrat 223 4.46 1.18
Leaning Democrat 124 4.50 1.10
Independent 84 4.76 1.08
Leaning Republican 111 4.43 1.04
Republican 255 4.55 1.14
Strong Republican 318 4.68 1.15
Table 25: Prejudice & Party Identity (7-point)
Table 26: Prejudice & Party Identity with Controls

4.3 Religiosity

Figure 55: Homo_negativity
Figure 56: Symbolic_Racism
Figure 57: Homo_negativity
Figure 58: Symbolic_Racism

Note. The y-axis shows residuals after regressing the outcome on demographics (Age, Income, Education, Gender). This illustrates how much the focal construct explains beyond demographic factors.

Figure 59: Homo_negativity
Figure 60: Symbolic_Racism

4.4 Candidate Preferences

Figure 61: Homo_negativity
Figure 62: Symbolic_Racism
Table 27: Candidate Preference and Homo_negativity [centered]
Candidate Preference N Mean SD
Donald Trump 444 -0.53 0.87
Hillary Clinton 371 0.34 0.91
Bernie Sanders 362 0.59 0.91
Ted Cruz 122 -0.69 0.72
Jeb Bush 83 -0.20 0.83
Gary Johnson 68 0.13 0.86
Rand Paul 44 -0.35 0.81
Table 28: Candidate Preference and Symbolic_Racism [centered]
Candidate Preference N Mean SD
Donald Trump 444 0.03 1.08
Hillary Clinton 371 0.06 0.96
Bernie Sanders 362 -0.07 1.01
Ted Cruz 122 0.02 0.89
Jeb Bush 83 -0.06 0.85
Gary Johnson 68 -0.18 0.94
Rand Paul 44 0.04 1.15
Table 29: Candidate Preference and Homo_negativity [raw means]
Candidate Preference N Mean SD
Donald Trump 444 3.57 2.33
Hillary Clinton 371 5.90 2.43
Bernie Sanders 362 6.56 2.43
Ted Cruz 122 3.12 1.92
Jeb Bush 83 4.44 2.24
Gary Johnson 68 5.31 2.29
Rand Paul 44 4.04 2.18
Table 30: Candidate Preference and Symbolic_Racism [raw means]
Candidate Preference N Mean SD
Donald Trump 444 4.62 1.21
Hillary Clinton 371 4.65 1.08
Bernie Sanders 362 4.50 1.13
Ted Cruz 122 4.60 1.00
Jeb Bush 83 4.52 0.95
Gary Johnson 68 4.38 1.05
Rand Paul 44 4.62 1.30

4.5 Party Preferences

Figure 63: Homo_negativity
Figure 64: Symbolic_Racism
Table 31: Party Preference and Homo_negativity [centered]
Party Preference N Mean SD
Democratic Party 560 0.50 0.92
Republican Party 508 -0.47 0.84
Libertarian Party 100 0.10 0.90
Green Party 40 0.87 0.89
Tea Party 68 -0.95 0.59
Constitution Party 14 -0.44 0.83
None 120 -0.04 0.91
Don't know 90 -0.11 0.91
Table 32: Party Preference and Symbolic_Racism [centered]
Party Preference N Mean SD
Democratic Party 560 -0.03 1.03
Republican Party 508 0.02 1.01
Libertarian Party 100 0.12 1.06
Green Party 40 0.18 0.98
Tea Party 68 0.03 0.85
Constitution Party 14 -0.36 0.80
None 120 -0.05 0.89
Don't know 90 -0.05 1.00
Table 33: Party Preference and Homo_negativity [raw means]
Party Preference N Mean SD
Democratic Party 560 6.31 2.47
Republican Party 508 3.71 2.25
Libertarian Party 100 5.26 2.42
Green Party 40 7.30 2.38
Tea Party 68 2.43 1.58
Constitution Party 14 3.79 2.23
None 120 4.87 2.44
Don't know 90 4.69 2.45
Table 34: Party Preference and Symbolic_Racism [raw means]
Party Preference N Mean SD
Democratic Party 560 4.55 1.16
Republican Party 508 4.60 1.14
Libertarian Party 100 4.72 1.19
Green Party 40 4.78 1.10
Tea Party 68 4.61 0.95
Constitution Party 14 4.18 0.90
None 120 4.52 1.00
Don't know 90 4.53 1.12

4.6 Voting

Table 35
Figure 65: Homo-negativity
Figure 66: Symbolic Racism

4.7 Voting & Party Identity

Figure 67: Homo_negativity
Figure 68: Symbolic_Racism
Table 36: Homo_negativity and Party Identity
Donald Trump Hilary Clinton
Strong Republican 282 7
Republican 166 24
Leaning Republican 58 7
Independent 17 16
Leaning Democrat 10 65
Democrat 27 129
Strong Democrat 4 323
Table 37: Symbolic_Racism and Party Identity
Donald Trump Hilary Clinton
Strong Republican 282 7
Republican 166 24
Leaning Republican 58 7
Independent 17 16
Leaning Democrat 10 65
Democrat 27 129
Strong Democrat 4 323
Table 38
Figure 69: Homo_negativity
Figure 70: Symbolic_Racism

4.8 Likeability

Figure 71: Homo_negativity
Figure 72: Symbolic_Racism
Figure 73: Homo_negativity
Figure 74: Symbolic_Racism

Note. The y-axis shows residuals after regressing the outcome on demographics (Age, Income, Education, Gender). This illustrates how much the focal construct explains beyond demographic factors.

Figure 75: Homo_negativity
Figure 76: Symbolic_Racism
Figure 77: Homo_negativity
Figure 78: Symbolic_Racism

Note. The y-axis shows residuals after regressing the outcome on demographics (Age, Income, Education, Gender). This illustrates how much the focal construct explains beyond demographic factors.

Figure 79: Homo_negativity
Figure 80: Symbolic_Racism
Figure 81: Homo_negativity
Figure 82: Symbolic_Racism

Note. The y-axis shows residuals after regressing the outcome on demographics (Age, Income, Education, Gender). This illustrates how much the focal construct explains beyond demographic factors.

5 Politico-Psychological correlates of Prejudice


5.1 Ideologies and Partisanship

Figure 83: Homo_negativity
Figure 84: Symbolic_Racism

5.2 Populism, Nationalism, Nativism, and Patriotism

Figure 85: Homo_negativity
Figure 86: Symbolic_Racism



5.3 Political Psychology

Figure 87: Homo_negativity
Figure 88: Symbolic_Racism

5.4 Social Justice Concerns, Empathy, and Prejudice

Figure 89: Homo_negativity
Figure 90: Symbolic_Racism



5.5 Values

Figure 91: Homo_negativity
Figure 92: Symbolic_Racism



5.6 Pot-Pourri

Figure 93: Homo_negativity
Figure 94: Symbolic_Racism



5.7 Positive and Negative correlates of Prejudice

Figure 95: Homo_negativity
Figure 96: Symbolic_Racism
Figure 97: Homo_negativity
Figure 98: Symbolic_Racism

6 Elastic Net Analysis

Figure 99: Elastic net predicting Prejudice: Political Psychology
Figure 100: Elastic net predicting Prejudice: Ideology & Partisanship
Figure 101: Elastic net predicting Prejudice: Demographics
Figure 102: Elastic net predicting Prejudice: Attitudes
Figure 103: Elastic net predicting Prejudice: Everything
Table 39: Elastic Net Summary
Elastic Net R² for predicting Prejudice
Variable Group # Predictors R²
Political Psychology 5 .333
Ideology & Partisanship 5 .284
Demographics 5 .096
Attitudes 6 .323
Everything 21 .421
PPBS
Psychology of Political Behavior Studies PI: Flavio Azevedo