Hi all,
Attached is a dataex file with two dependent variables (r_invest & e_invest) with binary outcomes (risk==1) and a set of independent variables (playera playerb income gender birthyear_cohort).
To give you an idea of the data structure, all players (a and b) played all treatments (r_invest and e_invest). However, based on the design of my experiment, playera is the only one making a decision in e_invest (playerb makes no decision there).
I hypothesise that, for player a: P(X=1) e_invest < P(X=1) r_invest
Only the data of player a having values for e_invest leaves me with 241 observations to compare the behaviour of player a between these two treatments.
Essentially, I am trying to find an efficient way to investigate whether risk-taking, for player a, is significantly different between r_invest and e_invest. However, when I started analysing the data, "tabulate e_invest r_invest if playera==1, chi2 expected" showed no significant association.
My question is twofold:
1) Am I approaching my data structure in the correct way? I am unsure since chi-square is insignificant
2) If not, how would I restructure my data to answer my hypothesis efficiently?
Best,
Lucas
Attached is a dataex file with two dependent variables (r_invest & e_invest) with binary outcomes (risk==1) and a set of independent variables (playera playerb income gender birthyear_cohort).
To give you an idea of the data structure, all players (a and b) played all treatments (r_invest and e_invest). However, based on the design of my experiment, playera is the only one making a decision in e_invest (playerb makes no decision there).
I hypothesise that, for player a: P(X=1) e_invest < P(X=1) r_invest
Only the data of player a having values for e_invest leaves me with 241 observations to compare the behaviour of player a between these two treatments.
Essentially, I am trying to find an efficient way to investigate whether risk-taking, for player a, is significantly different between r_invest and e_invest. However, when I started analysing the data, "tabulate e_invest r_invest if playera==1, chi2 expected" showed no significant association.
My question is twofold:
1) Am I approaching my data structure in the correct way? I am unsure since chi-square is insignificant
2) If not, how would I restructure my data to answer my hypothesis efficiently?
Best,
Lucas
