VS Field Study Report

Load data

Dataset Information

Peek at the dataset:

the final dataset includes 42 complete triads and 57 dyads from the 195 total triads.

Abbreviations

  • “S.” = “Supervisor”.
  • “FE.” = “Focal Employee”.
  • “CW.” = “Coworker”.
  • “VQ” = “Voice Quality”.
  • “VS” = “Voice Solicitation”.
  • “NS” = “Needs-Supply fit”.

Gender information

Demographic Breakdown

Gender Breakdown in Clean Dataset

How many coworker-focal employee pairs had different genders or the same gender?

## 
## DifferentGenders      SameGenders 
##               46               32

Correlations

Pre-registered Analyses

Supervisor ratings of voice quality

Main Effects

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: supervisor_voice_quality ~ employee_gender_f + (1 | triad_id)
##    Data: full_data
## 
## REML criterion at convergence: 312.1
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -2.22122 -0.35124  0.01865  0.34890  1.57679 
## 
## Random effects:
##  Groups   Name        Variance Std.Dev.
##  triad_id (Intercept) 1.4150   1.1895  
##  Residual             0.4577   0.6766  
## Number of obs: 98, groups:  triad_id, 64
## 
## Fixed effects:
##                      Estimate Std. Error      df t value Pr(>|t|)    
## (Intercept)            5.0657     0.1889 86.7940   26.82   <2e-16 ***
## employee_gender_fMan   0.1872     0.1891 53.9865    0.99    0.326    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr)
## emply_gnd_M -0.486

Controls

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## supervisor_voice_quality ~ employee_gender_f + s_gender_f + self_expertise +  
##     gender_diverse_pair + (1 | triad_id)
##    Data: full_data
## 
## REML criterion at convergence: 201.3
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -1.8863 -0.3073  0.1070  0.3510  1.4455 
## 
## Random effects:
##  Groups   Name        Variance Std.Dev.
##  triad_id (Intercept) 1.6416   1.2813  
##  Residual             0.5958   0.7719  
## Number of obs: 61, groups:  triad_id, 35
## 
## Fixed effects:
##                                Estimate Std. Error       df t value Pr(>|t|)
## (Intercept)                     4.80130    0.52276 50.13745   9.185 2.53e-12
## employee_gender_fMan            0.30379    0.28329 36.45065   1.072    0.291
## s_gender_fMan                   0.18551    0.48854 31.83884   0.380    0.707
## self_expertise                  0.08974    0.08748 38.71167   1.026    0.311
## gender_diverse_pairSameGenders -0.26754    0.49809 31.44497  -0.537    0.595
##                                   
## (Intercept)                    ***
## employee_gender_fMan              
## s_gender_fMan                     
## self_expertise                    
## gender_diverse_pairSameGenders    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) emp__M s_gn_M slf_xp
## emply_gnd_M -0.420                     
## s_gendr_fMn -0.534 -0.100              
## self_exprts -0.578  0.480  0.005       
## gndr_dvr_SG -0.414  0.029  0.065  0.046

Coworker’s/Employee’s self-ratings of voice solicitation

Main Effects

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: self_voice ~ employee_gender_f + (1 | triad_id)
##    Data: full_data
## 
## REML criterion at convergence: 365.3
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -1.5416 -1.0127  0.0724  0.7335  1.9234 
## 
## Random effects:
##  Groups   Name        Variance Std.Dev.
##  triad_id (Intercept) 0.000    0.000   
##  Residual             3.575    1.891   
## Number of obs: 89, groups:  triad_id, 60
## 
## Fixed effects:
##                      Estimate Std. Error      df t value Pr(>|t|)    
## (Intercept)            3.9149     0.2758 87.0000  14.194   <2e-16 ***
## employee_gender_fMan  -0.5518     0.4015 87.0000  -1.374    0.173    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr)
## emply_gnd_M -0.687
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')

Controls

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: self_voice ~ employee_gender_f + s_gender_f + gender_diverse_pair +  
##     self_expertise + (1 | triad_id)
##    Data: full_data
## 
## REML criterion at convergence: 265.9
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -1.86180 -0.80290  0.03008  0.67272  2.42014 
## 
## Random effects:
##  Groups   Name        Variance Std.Dev.
##  triad_id (Intercept) 0.000    0.000   
##  Residual             3.433    1.853   
## Number of obs: 66, groups:  triad_id, 37
## 
## Fixed effects:
##                                Estimate Std. Error      df t value Pr(>|t|)    
## (Intercept)                      3.4310     0.6365 61.0000   5.391  1.2e-06 ***
## employee_gender_fMan            -0.4880     0.4759 61.0000  -1.025   0.3092    
## s_gender_fMan                   -0.4336     0.4744 61.0000  -0.914   0.3643    
## gender_diverse_pairSameGenders  -0.2655     0.4796 61.0000  -0.554   0.5819    
## self_expertise                   0.2723     0.1424 61.0000   1.912   0.0606 .  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) emp__M s_gn_M gn__SG
## emply_gnd_M -0.441                     
## s_gendr_fMn -0.446 -0.145              
## gndr_dvr_SG -0.302  0.004  0.061       
## self_exprts -0.716  0.237  0.051  0.017
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')

Interactions

I tested if the following variables moderated our effects: > Employee/Coworker’s self-rated demands-abilities fit (note: there was no main effect of gender on this variable, even with controls)
> Employee/Coworker’s self-rated needs-supply fit (note: there was no main effect of gender on this variable, even with controls)
> Supervisor’s gender
> Employee/Coworker’s self-rated expertise in DEI

We did not see any significant or marginal interactions on these variables.

Exploratory analyses - Analyses with gender diverse pairs

Voice Quality

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: supervisor_voice_quality ~ employee_gender_f + (1 | triad_id)
##    Data: gender_diverse_pair_data
## 
## REML criterion at convergence: 118.9
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.1391 -0.4068  0.1235  0.3223  1.2337 
## 
## Random effects:
##  Groups   Name        Variance Std.Dev.
##  triad_id (Intercept) 1.2254   1.1070  
##  Residual             0.4751   0.6893  
## Number of obs: 39, groups:  triad_id, 22
## 
## Fixed effects:
##                      Estimate Std. Error      df t value Pr(>|t|)    
## (Intercept)            5.1952     0.2847 28.1499  18.250   <2e-16 ***
## employee_gender_fMan   0.2183     0.2317 17.8512   0.942    0.359    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr)
## emply_gnd_M -0.391

Voice Quality - Control

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## supervisor_voice_quality ~ employee_gender_f + s_gender_f + self_expertise +  
##     (1 | triad_id)
##    Data: gender_diverse_pair_data
## 
## REML criterion at convergence: 119.8
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -2.11058 -0.34307  0.07657  0.31325  1.34328 
## 
## Random effects:
##  Groups   Name        Variance Std.Dev.
##  triad_id (Intercept) 1.3194   1.1486  
##  Residual             0.4655   0.6823  
## Number of obs: 39, groups:  triad_id, 22
## 
## Fixed effects:
##                      Estimate Std. Error       df t value Pr(>|t|)    
## (Intercept)           4.65954    0.57418 33.84096   8.115  1.9e-09 ***
## employee_gender_fMan  0.34295    0.28608 21.63333   1.199    0.244    
## s_gender_fMan         0.44821    0.55053 20.48455   0.814    0.425    
## self_expertise        0.07708    0.10172 25.77087   0.758    0.455    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) emp__M s_gn_M
## emply_gnd_M -0.531              
## s_gendr_fMn -0.600  0.014       
## self_exprts -0.649  0.596  0.052

Voice Solicitation

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: self_voice ~ employee_gender_f + (1 | triad_id)
##    Data: gender_diverse_pair_data
## 
## REML criterion at convergence: 166.5
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -1.8671 -0.9614  0.1794  0.8214  1.4950 
## 
## Random effects:
##  Groups   Name        Variance Std.Dev.
##  triad_id (Intercept) 0.000    0.00    
##  Residual             2.925    1.71    
## Number of obs: 43, groups:  triad_id, 23
## 
## Fixed effects:
##                      Estimate Std. Error      df t value Pr(>|t|)    
## (Intercept)            4.1932     0.3646 41.0000  11.500 2.08e-14 ***
## employee_gender_fMan  -1.0979     0.5218 41.0000  -2.104   0.0415 *  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr)
## emply_gnd_M -0.699
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')

Voice Solicitation - Control

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: self_voice ~ employee_gender_f + s_gender_f + self_expertise +  
##     (1 | triad_id)
##    Data: gender_diverse_pair_data
## 
## REML criterion at convergence: 165.7
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.0066 -0.8444  0.1390  0.7228  1.8195 
## 
## Random effects:
##  Groups   Name        Variance Std.Dev.
##  triad_id (Intercept) 0.000    0.000   
##  Residual             2.934    1.713   
## Number of obs: 43, groups:  triad_id, 23
## 
## Fixed effects:
##                      Estimate Std. Error      df t value Pr(>|t|)    
## (Intercept)            3.8051     0.7898 39.0000   4.818 2.22e-05 ***
## employee_gender_fMan  -0.8714     0.5645 39.0000  -1.544    0.131    
## s_gender_fMan         -0.3572     0.5460 39.0000  -0.654    0.517    
## self_expertise         0.1855     0.1689 39.0000   1.098    0.279    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) emp__M s_gn_M
## emply_gnd_M -0.598              
## s_gendr_fMn -0.541  0.070       
## self_exprts -0.772  0.378  0.142
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')

Interesting Interactions

summary(lmer(self_voice_quality~employee_gender_f*self_demands_abilities+(1|triad_id), gender_diverse_pair_data))
## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: self_voice_quality ~ employee_gender_f * self_demands_abilities +  
##     (1 | triad_id)
##    Data: gender_diverse_pair_data
## 
## REML criterion at convergence: 111.4
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.2039 -0.3940  0.1821  0.5328  1.7144 
## 
## Random effects:
##  Groups   Name        Variance Std.Dev.
##  triad_id (Intercept) 0.1944   0.4409  
##  Residual             0.5831   0.7636  
## Number of obs: 42, groups:  triad_id, 23
## 
## Fixed effects:
##                                             Estimate Std. Error       df
## (Intercept)                                  2.64943    0.48555 37.41723
## employee_gender_fMan                         0.65335    0.76640 36.97580
## self_demands_abilities                       0.55404    0.09654 36.71277
## employee_gender_fMan:self_demands_abilities -0.28633    0.16779 37.49984
##                                             t value Pr(>|t|)    
## (Intercept)                                   5.457 3.31e-06 ***
## employee_gender_fMan                          0.852   0.3994    
## self_demands_abilities                        5.739 1.45e-06 ***
## employee_gender_fMan:self_demands_abilities  -1.707   0.0962 .  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) emp__M slf_d_
## emply_gnd_M -0.640              
## slf_dmnds_b -0.919  0.610       
## emply__M:__  0.558 -0.949 -0.603
sjPlot::plot_model(lmer(self_voice_quality~employee_gender_f*self_demands_abilities+(1|triad_id), gender_diverse_pair_data),
                   type = "int",
                   mdrt.values = "meansd",
                   axis.title = c("Employee Gender", "Self-rated Voice Quality"),
                   legend.title = "Self-rated demands-abilities")

summary(lmer(self_voice_quality~employee_gender_f*s_gender_f+(1|triad_id), gender_diverse_pair_data))
## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: self_voice_quality ~ employee_gender_f * s_gender_f + (1 | triad_id)
##    Data: gender_diverse_pair_data
## 
## REML criterion at convergence: 132.8
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -2.19992 -0.57360  0.06099  0.51265  1.65591 
## 
## Random effects:
##  Groups   Name        Variance Std.Dev.
##  triad_id (Intercept) 0.1842   0.4292  
##  Residual             1.2134   1.1016  
## Number of obs: 43, groups:  triad_id, 23
## 
## Fixed effects:
##                                    Estimate Std. Error       df t value
## (Intercept)                         5.00748    0.41758 38.64792  11.992
## employee_gender_fMan                0.03842    0.55537 21.44175   0.069
## s_gender_fMan                       0.17109    0.52364 38.56653   0.327
## employee_gender_fMan:s_gender_fMan -1.20360    0.69952 20.73747  -1.721
##                                    Pr(>|t|)    
## (Intercept)                        1.35e-14 ***
## employee_gender_fMan                  0.945    
## s_gender_fMan                         0.746    
## employee_gender_fMan:s_gender_fMan    0.100    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) emp__M s_gn_M
## emply_gnd_M -0.665              
## s_gendr_fMn -0.797  0.530       
## empl__M:__M  0.528 -0.794 -0.658
sjPlot::plot_model(
  lmer(self_voice_quality~employee_gender_f*s_gender_f+(1|triad_id), gender_diverse_pair_data),
  type = "int",
  legend.title = "Supervisor Gender",
  axis.title = c("Employee Gender", "Self-rated Voice Quality")
  )