Description of studies
Both studies are our standard vignettes (whose voice would you solicit about a role? Participants than saw four candidates).
In Study 1 (24_03.13-QualificationsDraft2), we keep the context constant (whose voice should we solicit about a DEI role). However, we manipulate who is the expert: NOBODY (‘’none’’ condition), a focal man (‘’man’‘) condition or a focal woman (’‘woman’’). > Note: PD is the focal man. JA was the other man. MC was the focal woman. LM was the other woman
In Study 2 (24_03.28 - Qualifications4), participants were told about a DEI position (‘’none’’ and ‘’man’‘) or a’‘traditional’’ workplace position (about the participant’s job function).
Study Analyses
Voice Quality
Summary
Study 1
## Linear mixed model fit by REML. t-tests use Satterthwaite's method ['lmerModLmerTest']
## Formula: vq ~ deiexpert_f * person + (1 | pid)
## Data: qual_clean_long
##
## REML criterion at convergence: 7237
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.618 -0.541 0.039 0.612 2.643
##
## Random effects:
## Groups Name Variance Std.Dev.
## pid (Intercept) 0.57 0.755
## Residual 1.53 1.236
## Number of obs: 2068, groups: pid, 517
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 5.856 0.113 1683.285 51.62 < 0.0000000000000002 ***
## deiexpert_fNo one has DEI Quals. -1.391 0.159 1683.285 -8.77 < 0.0000000000000002 ***
## deiexpert_fWoman Has DEI Quals. -1.951 0.156 1683.285 -12.51 < 0.0000000000000002 ***
## personFocalWoman -1.264 0.137 1542.000 -9.23 < 0.0000000000000002 ***
## personOtherMan -1.475 0.137 1542.000 -10.78 < 0.0000000000000002 ***
## personOtherWoman -1.187 0.137 1542.000 -8.67 < 0.0000000000000002 ***
## deiexpert_fNo one has DEI Quals.:personFocalWoman 1.928 0.191 1542.000 10.07 < 0.0000000000000002 ***
## deiexpert_fWoman Has DEI Quals.:personFocalWoman 3.316 0.188 1542.000 17.61 < 0.0000000000000002 ***
## deiexpert_fNo one has DEI Quals.:personOtherMan 1.572 0.191 1542.000 8.22 0.00000000000000044 ***
## deiexpert_fWoman Has DEI Quals.:personOtherMan 1.639 0.188 1542.000 8.71 < 0.0000000000000002 ***
## deiexpert_fNo one has DEI Quals.:personOtherWoman 1.833 0.191 1542.000 9.58 < 0.0000000000000002 ***
## deiexpert_fWoman Has DEI Quals.:personOtherWoman 1.589 0.188 1542.000 8.44 < 0.0000000000000002 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) dx_NohDEIQ. dx_WHDEIQ. prsnFW prsnOM prsnOW d_NohDEIQ.:F d_WHDEIQ.:F d_NohDEIQ.:OM d_WHDEIQ.:OM d_NohDEIQ.:OW
## dx_NohDEIQ. -0.716
## dxp_WHDEIQ. -0.727 0.520
## persnFclWmn -0.603 0.432 0.439
## persnOthrMn -0.603 0.432 0.439 0.500
## prsnOthrWmn -0.603 0.432 0.439 0.500 0.500
## d_NohDEIQ.:F 0.432 -0.603 -0.314 -0.716 -0.358 -0.358
## d_WHDEIQ.:F 0.439 -0.314 -0.603 -0.727 -0.364 -0.364 0.520
## d_NohDEIQ.:OM 0.432 -0.603 -0.314 -0.358 -0.716 -0.358 0.500 0.260
## d_WHDEIQ.:OM 0.439 -0.314 -0.603 -0.364 -0.727 -0.364 0.260 0.500 0.520
## d_NohDEIQ.:OW 0.432 -0.603 -0.314 -0.358 -0.358 -0.716 0.500 0.260 0.500 0.260
## d_WHDEIQ.:OW 0.439 -0.314 -0.603 -0.364 -0.364 -0.727 0.260 0.500 0.260 0.500 0.520Study 2
## Linear mixed model fit by REML. t-tests use Satterthwaite's method ['lmerModLmerTest']
## Formula: vq ~ deiexpert_f * person + (1 | pid)
## Data: qual_clean_long4
##
## REML criterion at convergence: 6794
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.304 -0.551 0.049 0.628 3.160
##
## Random effects:
## Groups Name Variance Std.Dev.
## pid (Intercept) 0.834 0.913
## Residual 1.283 1.133
## Number of obs: 1988, groups: pid, 497
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 6.094 0.121 1348.115 50.26 <0.0000000000000002 ***
## deiexpert_fDEI-None -1.547 0.160 1348.115 -9.67 <0.0000000000000002 ***
## deiexpert_fTrad 0.262 0.167 1348.115 1.56 0.1183
## personOtherMan -1.941 0.134 1482.000 -14.54 <0.0000000000000002 ***
## personwoman1 -1.285 0.134 1482.000 -9.62 <0.0000000000000002 ***
## personwoman2 -1.319 0.134 1482.000 -9.89 <0.0000000000000002 ***
## deiexpert_fDEI-None:personOtherMan 1.817 0.176 1482.000 10.31 <0.0000000000000002 ***
## deiexpert_fTrad:personOtherMan -0.081 0.184 1482.000 -0.44 0.6601
## deiexpert_fDEI-None:personwoman1 2.006 0.176 1482.000 11.39 <0.0000000000000002 ***
## deiexpert_fTrad:personwoman1 -0.571 0.184 1482.000 -3.10 0.0020 **
## deiexpert_fDEI-None:personwoman2 2.080 0.176 1482.000 11.80 <0.0000000000000002 ***
## deiexpert_fTrad:personwoman2 -0.489 0.184 1482.000 -2.65 0.0081 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) dx_DEI-N dxpr_T prsnOM prsnw1 prsnw2 d_DEI-N:O d_T:OM d_DEI-N:1 dx_T:1 d_DEI-N:2
## dxprt_DEI-N -0.758
## dexprt_fTrd -0.724 0.549
## persnOthrMn -0.550 0.417 0.399
## personwomn1 -0.550 0.417 0.399 0.500
## personwomn2 -0.550 0.417 0.399 0.500 0.500
## dx_DEI-N:OM 0.417 -0.550 -0.302 -0.758 -0.379 -0.379
## dxprt_fT:OM 0.399 -0.302 -0.550 -0.724 -0.362 -0.362 0.549
## dxp_DEI-N:1 0.417 -0.550 -0.302 -0.379 -0.758 -0.379 0.500 0.274
## dxprt_fTr:1 0.399 -0.302 -0.550 -0.362 -0.724 -0.362 0.274 0.500 0.549
## dxp_DEI-N:2 0.417 -0.550 -0.302 -0.379 -0.379 -0.758 0.500 0.274 0.500 0.274
## dxprt_fTr:2 0.399 -0.302 -0.550 -0.362 -0.362 -0.724 0.274 0.500 0.274 0.500 0.549Emmeans
Study 1
Study 2
Mixed-Model Anova comparing EVERY candidate to each other
Study 1
Study 2
Pairwise comparisons (comparing conditions across candidates)
Study 1
Study 2
Pairwise comparisons (comparing candidates across conditions)
Study 1
Study 2
Graph
Study 1
Study 2
Voice Solicitation
Summary
Study 1
## Linear mixed model fit by REML. t-tests use Satterthwaite's method ['lmerModLmerTest']
## Formula: vs ~ deiexpert_f * person + (1 | pid)
## Data: qual_clean_long
##
## REML criterion at convergence: 7595
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.249 -0.518 0.061 0.618 2.781
##
## Random effects:
## Groups Name Variance Std.Dev.
## pid (Intercept) 0.934 0.966
## Residual 1.710 1.308
## Number of obs: 2068, groups: pid, 517
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 5.828 0.127 1496.028 45.76 < 0.0000000000000002 ***
## deiexpert_fNo one has DEI Quals. -1.331 0.178 1496.028 -7.48 0.00000000000012729 ***
## deiexpert_fWoman Has DEI Quals. -1.861 0.175 1496.028 -10.63 < 0.0000000000000002 ***
## personFocalWoman -1.141 0.145 1542.000 -7.88 0.00000000000000624 ***
## personOtherMan -1.331 0.145 1542.000 -9.19 < 0.0000000000000002 ***
## personOtherWoman -0.801 0.145 1542.000 -5.53 0.00000003813708947 ***
## deiexpert_fNo one has DEI Quals.:personFocalWoman 1.884 0.202 1542.000 9.31 < 0.0000000000000002 ***
## deiexpert_fWoman Has DEI Quals.:personFocalWoman 3.179 0.199 1542.000 15.96 < 0.0000000000000002 ***
## deiexpert_fNo one has DEI Quals.:personOtherMan 1.419 0.202 1542.000 7.01 0.00000000000356045 ***
## deiexpert_fWoman Has DEI Quals.:personOtherMan 1.501 0.199 1542.000 7.53 0.00000000000008302 ***
## deiexpert_fNo one has DEI Quals.:personOtherWoman 1.593 0.202 1542.000 7.87 0.00000000000000667 ***
## deiexpert_fWoman Has DEI Quals.:personOtherWoman 1.639 0.199 1542.000 8.23 0.00000000000000039 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) dx_NohDEIQ. dx_WHDEIQ. prsnFW prsnOM prsnOW d_NohDEIQ.:F d_WHDEIQ.:F d_NohDEIQ.:OM d_WHDEIQ.:OM d_NohDEIQ.:OW
## dx_NohDEIQ. -0.716
## dxp_WHDEIQ. -0.727 0.520
## persnFclWmn -0.569 0.407 0.414
## persnOthrMn -0.569 0.407 0.414 0.500
## prsnOthrWmn -0.569 0.407 0.414 0.500 0.500
## d_NohDEIQ.:F 0.407 -0.569 -0.296 -0.716 -0.358 -0.358
## d_WHDEIQ.:F 0.414 -0.296 -0.569 -0.727 -0.364 -0.364 0.520
## d_NohDEIQ.:OM 0.407 -0.569 -0.296 -0.358 -0.716 -0.358 0.500 0.260
## d_WHDEIQ.:OM 0.414 -0.296 -0.569 -0.364 -0.727 -0.364 0.260 0.500 0.520
## d_NohDEIQ.:OW 0.407 -0.569 -0.296 -0.358 -0.358 -0.716 0.500 0.260 0.500 0.260
## d_WHDEIQ.:OW 0.414 -0.296 -0.569 -0.364 -0.364 -0.727 0.260 0.500 0.260 0.500 0.520Study 2
## Linear mixed model fit by REML. t-tests use Satterthwaite's method ['lmerModLmerTest']
## Formula: vs ~ deiexpert_f * person + (1 | pid)
## Data: qual_clean_long4
##
## REML criterion at convergence: 7020
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.196 -0.560 0.044 0.609 2.676
##
## Random effects:
## Groups Name Variance Std.Dev.
## pid (Intercept) 1.05 1.02
## Residual 1.40 1.18
## Number of obs: 1988, groups: pid, 497
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 6.0417 0.1304 1276.8118 46.33 <0.0000000000000002 ***
## deiexpert_fDEI-None -1.5674 0.1721 1276.8118 -9.11 <0.0000000000000002 ***
## deiexpert_fTrad 0.2917 0.1800 1276.8118 1.62 0.1054
## personOtherMan -1.9514 0.1396 1482.0000 -13.98 <0.0000000000000002 ***
## personwoman1 -1.2431 0.1396 1482.0000 -8.91 <0.0000000000000002 ***
## personwoman2 -1.2361 0.1396 1482.0000 -8.86 <0.0000000000000002 ***
## deiexpert_fDEI-None:personOtherMan 1.8895 0.1842 1482.0000 10.26 <0.0000000000000002 ***
## deiexpert_fTrad:personOtherMan 0.0017 0.1926 1482.0000 0.01 0.9929
## deiexpert_fDEI-None:personwoman1 2.1606 0.1842 1482.0000 11.73 <0.0000000000000002 ***
## deiexpert_fTrad:personwoman1 -0.5022 0.1926 1482.0000 -2.61 0.0092 **
## deiexpert_fDEI-None:personwoman2 2.2232 0.1842 1482.0000 12.07 <0.0000000000000002 ***
## deiexpert_fTrad:personwoman2 -0.5438 0.1926 1482.0000 -2.82 0.0048 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) dx_DEI-N dxpr_T prsnOM prsnw1 prsnw2 d_DEI-N:O d_T:OM d_DEI-N:1 dx_T:1 d_DEI-N:2
## dxprt_DEI-N -0.758
## dexprt_fTrd -0.724 0.549
## persnOthrMn -0.535 0.405 0.388
## personwomn1 -0.535 0.405 0.388 0.500
## personwomn2 -0.535 0.405 0.388 0.500 0.500
## dx_DEI-N:OM 0.405 -0.535 -0.294 -0.758 -0.379 -0.379
## dxprt_fT:OM 0.388 -0.294 -0.535 -0.724 -0.362 -0.362 0.549
## dxp_DEI-N:1 0.405 -0.535 -0.294 -0.379 -0.758 -0.379 0.500 0.274
## dxprt_fTr:1 0.388 -0.294 -0.535 -0.362 -0.724 -0.362 0.274 0.500 0.549
## dxp_DEI-N:2 0.405 -0.535 -0.294 -0.379 -0.379 -0.758 0.500 0.274 0.500 0.274
## dxprt_fTr:2 0.388 -0.294 -0.535 -0.362 -0.362 -0.724 0.274 0.500 0.274 0.500 0.549Emmeans
Study 1
Study 2
Mixed-Model Anova comparing EVERY candidate to each other
Study 1
Study 2
Pairwise comparisons (comparing conditions across candidates)
Study 1
Study 2
Pairwise comparisons (comparing candidates across conditions)
Study 1
Study 2
Graph
Study 1
Study 2
Rank
Note: all values are reverse-coded, because Qualtrics codes the values ranked at the bottom of the list as lower numerical values. So someone who is ranked “1” is coded as “4”, who is ranked “2” is coded as “3”, etc.
Summary
Study 1
## Call:
## clm2(location = rank_rf ~ deiexpert_f * person + (1 | pid), data = qual_clean_long)
##
## Location coefficients:
## Estimate Std. Error z value Pr(>|z|)
## deiexpert_fNo one has DEI Quals. 0.456 0.217 2.102 0.03554
## deiexpert_fWoman Has DEI Quals. -0.488 0.206 -2.371 0.01774
## personFocalWoman 0.844 0.201 4.192 0.00002762464449352
## personOtherMan 0.307 0.217 1.416 0.15683
## personOtherWoman -0.042 0.207 -0.202 0.83975
## deiexpert_fNo one has DEI Quals.:personFocalWoman -1.042 0.293 -3.552 0.00038
## deiexpert_fWoman Has DEI Quals.:personFocalWoman 2.463 0.305 8.062 0.00000000000000075
## deiexpert_fNo one has DEI Quals.:personOtherMan -0.676 0.296 -2.283 0.02241
## deiexpert_fWoman Has DEI Quals.:personOtherMan -0.206 0.286 -0.719 0.47190
## deiexpert_fNo one has DEI Quals.:personOtherWoman 0.061 0.287 0.212 0.83237
## deiexpert_fWoman Has DEI Quals.:personOtherWoman 0.620 0.276 2.248 0.02456
##
## No scale coefficients
##
## Threshold coefficients:
## Estimate Std. Error z value
## 1|2 -0.889 0.157 -5.649
## 2|3 0.294 0.157 1.867
## 3|4 1.556 0.161 9.671
##
## log-likelihood: -2700.83
## AIC: 5429.66
## Condition number of Hessian: 508.05Study 2
## Call:
## clm2(location = rank_rf ~ deiexpert_f * person + (1 | pid), data = qual_clean_long4)
##
## Location coefficients:
## Estimate Std. Error z value Pr(>|z|)
## deiexpert_fDEI-None -3.707 0.276 -13.415 < 0.0000000000000002
## deiexpert_fTrad 0.527 0.360 1.462 0.1436
## personOtherMan -4.826 0.297 -16.235 < 0.0000000000000002
## personwoman1 -3.578 0.284 -12.620 < 0.0000000000000002
## personwoman2 -3.575 0.280 -12.754 < 0.0000000000000002
## deiexpert_fDEI-None:personOtherMan 4.282 0.349 12.253 < 0.0000000000000002
## deiexpert_fTrad:personOtherMan 0.083 0.421 0.196 0.8444
## deiexpert_fDEI-None:personwoman1 4.721 0.344 13.707 < 0.0000000000000002
## deiexpert_fTrad:personwoman1 -1.081 0.414 -2.608 0.0091
## deiexpert_fDEI-None:personwoman2 4.978 0.345 14.428 < 0.0000000000000002
## deiexpert_fTrad:personwoman2 -0.553 0.410 -1.349 0.1774
##
## No scale coefficients
##
## Threshold coefficients:
## Estimate Std. Error z value
## 1|2 -4.624 0.250 -18.490
## 2|3 -3.301 0.246 -13.442
## 3|4 -1.686 0.237 -7.115
##
## log-likelihood: -2316.79
## AIC: 4661.58
## Condition number of Hessian: 815.36
## (8 observations deleted due to missingness)Emmeans
Study 1
Study 2
Mixed-Model Anova comparing EVERY candidate to each other
Study 1
Study 2
Pairwise comparisons (comparing conditions across candidates)
Study 1
Study 2
Pairwise comparisons (comparing candidates across conditions)
Study 1
Study 2
Graph
Study 1
Study 2
Rank woman as #1
Study 1
Logistic regression (w/ confidence interval of coefficients)
##
## Call:
## glm(formula = rank_wom ~ deiexpert_r, family = "binomial", data = qual_clean)
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.105 0.153 -0.69 0.491
## deiexpert_r2. Man -0.488 0.224 -2.18 0.029 *
## deiexpert_r3. Woman 1.817 0.256 7.09 0.0000000000013 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 706.93 on 516 degrees of freedom
## Residual deviance: 605.41 on 514 degrees of freedom
## AIC: 611.4
##
## Number of Fisher Scoring iterations: 4## 2.5 % 97.5 %
## (Intercept) 0.4069 0.1946
## deiexpert_r2. Man 0.9302 0.0507
## deiexpert_r3. Woman 1.3257 2.3323Odds ratio
## (Intercept) deiexpert_r2. Man deiexpert_r3. Woman
## 0.9000 0.6138 6.1508Study 2
Logistic regression (w/ confidence interval of coefficients)
##
## Call:
## glm(formula = rank_wom ~ orderings, family = "binomial", data = qual_clean4)
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 2.025 0.244 8.30 <0.0000000000000002 ***
## orderings2. Man -3.704 0.380 -9.75 <0.0000000000000002 ***
## orderings3. Trad -3.803 0.396 -9.60 <0.0000000000000002 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 464.33 on 334 degrees of freedom
## Residual deviance: 263.42 on 332 degrees of freedom
## (166 observations deleted due to missingness)
## AIC: 269.4
##
## Number of Fisher Scoring iterations: 4Odds ratio
## (Intercept) orderings2. Man orderings3. Trad
## 2.4192 0.4838 0.4777Rank woman as #2
Study 1
Ranking
##
## Call:
## glm(formula = rankwom2 ~ deiexpert_r, family = "binomial", data = qual_clean)
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.887 0.183 4.84 0.0000013 ***
## deiexpert_r2. Man 0.937 0.303 3.10 0.002 **
## deiexpert_r3. Woman -0.464 0.250 -1.86 0.064 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 508.57 on 431 degrees of freedom
## Residual deviance: 483.23 on 429 degrees of freedom
## (85 observations deleted due to missingness)
## AIC: 489.2
##
## Number of Fisher Scoring iterations: 4Odds
## (Intercept) orderings2. Man orderings3. Trad
## 2.081 1.170 1.076Study 2
Ranking
##
## Call:
## lm(formula = rank_wom2 ~ orderings, data = qual_clean4, family = "binomial")
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.890 0.110 0.194 0.267 0.267
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.7329 0.0310 23.62 < 0.0000000000000002 ***
## orderings2. Man 0.1568 0.0467 3.36 0.00086 ***
## orderings3. Trad 0.0728 0.0456 1.60 0.11084
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.394 on 424 degrees of freedom
## (74 observations deleted due to missingness)
## Multiple R-squared: 0.0259, Adjusted R-squared: 0.0213
## F-statistic: 5.64 on 2 and 424 DF, p-value: 0.00383Odds
## (Intercept) orderings2. Man orderings3. Trad
## 2.081 1.170 1.076