search

lottybrand / Prestige_2_Analysis

analysis scripts for 2nd prestige experiment via Dallinger
0 stars 0 forks source link

Compute correlations #2

Open lottybrand opened 4 years ago

lottybrand commented 4 years ago

Figure out the best way to check that domain-specific prestige (i.e. times chosen in Round 1 in the topic being answered) is more tightly correlated to domain-specific score than domain-general prestige or cross-domain prestige are. --> This is our prediction 1. Is it what we mean/want to do??

lottybrand commented 4 years ago

i think first just repeat correlations calculated from first experiment, e.g. are scores within topic more closely correlated than between and overall? then see if the prestige scores predict actual scores within topic better than over all topics (and on diff topics)

lottybrand commented 3 years ago

I think replicating the table from our preregistration document is fine, e.g. round 1 and round 2 score comparisons. Perhaps an exploratory analysis could check if prestige scores actually correlate with asocial scores within and between topics etc