Table 3 Bayesian repeated measures ANOVA results, post-hoc analysis: (a) number of safe choices, (b) probability of inconsistent choice, and (c) probability of correct IT answer.

From: Hierarchies and decision-making in groups: experimental evidence

(a) Model comparisona

Models

P(M)

P(M|data)

BFM

BF01

Error %

Null model (incl. subject and random slopes)

0.500

0.977

41.719

1.000

 

Hierarchy types

0.500

0.023

0.024

41.719

0.515

(b) Model comparisona

Models

P(M)

P(M|data)

BFM

BF01

Error %

Null model (incl. subject and random slopes)

0.500

0.829

4.860

1.000

 

Hierarchy types

0.500

0.171

0.206

4.860

0.565

(c) Model comparisona

Models

P(M)

P(M|data)

BFM

BF01

Error %

Null model (incl. subject and random slopes)

0.500

0.095

0.105

1.000

 

Hierarchy types

0.500

0.905

9.482

0.105

0.298

Post hoc testsb

Post hoc comparisons—correct answer IT

  

Prior odds

Posterior odds

BF10, U

Error %

Voting group

Random

0.320

0.153

0.480

0.031

 

Vote

0.320

0.072

0.226

0.037

 

Age

0.320

0.076

0.237

0.036

 

Merit

0.320

16.834

52.687

1.948 × 10−8

Random

Vote

0.320

0.069

0.215

0.037

 

Age

0.320

0.072

0.226

0.037

 

Merit

0.320

0.562

1.758

1.622 × 10−6

Vote

Age

0.320

0.060

0.189

0.038

 

Merit

0.320

0.954

2.987

8.137 × 10−7

Age

Merit

0.320

1.405

4.397

4.583 × 10−7

  1. aAll models include subject, and random slopes for all repeated measures factors.
  2. bThe posterior odds have been corrected for multiple testing by fixing to 0.5 the prior probability that the null hypothesis holds across all comparisons (Westfall et al., 1997). Individual comparisons are based on the default t-test with a Cauchy (0, r = 1/sqrt(2)) prior. The “U” in the Bayes factor denotes that it is uncorrected.
Back to article page