I decided to power us for a 2x2 interaction, all between-subjects factors. pwr2::ss.2way(a = 2, b = 2, alpha = .05, beta = .1, f.A = .15, f.B = .15, B = 10000). This revealed that we need 118 in each group (472 total). So I opened both surveys to 250 (to reach 500, in case folks fail)

## 
##  Chi-squared test for given probabilities
## 
## data:  table(binary$binary)
## X-squared = 1, df = 1, p-value = 0.3

• Analyses
  • Manager: Comparing ratings of the Top woman vs. Top man
    • Managers’ ranking
    • Does managers’ gender interact with their top-selected worker?
  • Supervisee - Comparing women participants’ responses to men participants’ responses.
• Comparing Manager’s responses to Supervisee’s responses
  • Means and standard deviations
  • P values