Reanalyses of data from Horne, Powell, Hummel & Holyoak (2015)
git clone
Log | Files | Refs | README | LICENSE

commit 87d18b5bd320644e571faaab1b5ba6fc63f16908
parent 3a789db7f4bf3311dcede8a072f2c7c90cc75caa
Author: eamoncaddigan <>
Date:   Fri, 18 Sep 2015 18:37:16 -0400

Merge branch 'master' of

Diffstat: | 13++++++-------
1 file changed, 6 insertions(+), 7 deletions(-)

diff --git a/ b/ @@ -3,13 +3,12 @@ Data from Horne, Powell, Hummel &amp; Holyoak (2015). doi: 10.1073/pnas.15040191 *** -Currently using Bayesian Estimation and a hierarchical model of all output measures in a manner that avoids the need to correct for multiple comparisons. +I wrote two blog posts about this. -### TODO: +First, I used Bayesian estimation to estimate the size of the attitude shifts. You can read that [over here]( -- [x] Model individual questions -- [x] Add factors for intervention and interval plus interactions -- [x] Explore the full model with all interactions. Important stuff already replicated but this would allow me to look at the effects on each question. -- [x] Write up results and post online +Next, I did a followup where I explored how bootstrapping could be used to compare the interventions. This one is [here]( -[Online!]( +*** + +I *think* that's it for these data from me. Since this was my first time really diving into survey results I expect that I'll recycle code later. Please e-mail me if you have questions, or just to let me know if you've found any of this useful!