Conversation

Neuroskeptic

@Neuro_Skeptic

·

Jul 10

Consider an experiment to measure effect of speaker sex on perceived competence. 400 participants take part, in 40 groups of 10. Each group is randomized to listen to 1 of 4 speakers (2 male 2 female); then group discusses the talk, then rate the speaker. What's the effective n?

• 400

  22.3%
• 40

  36.1%
• 4

  16.5%
• Other/Not enough info

  25.1%

696 votes·Final results

14

4

5

'(·)

@allgebrah

Replying to

@Neuro_Skeptic

group discussion and choice of topic seem like they'd ruin generalizability of the finding each on their own, so I'm not sure n is even meaningful

2:22 PM · Jul 10, 2022·Twitter Web App

1

Like