Regression analysis of participants' choice.

Number of subjects = 42; number of observations = 3360. Data from all trials (t = 80).
Log likelihood = −1944.3451, Wald chi2(3) = 596.32, Prob>chi2 = 0.000.
The table reports the coefficients estimated in the logistic regression of the choice made by participants. The dependent variable choice is equal to 1 if the participant chose the lottery 1 and 0 if the participant chose the lottery 2. The variable dev is the difference between the expected value of the first and second lottery (when participants maximize expected values the coefficient is positive); the variable dsd is the difference between the standard deviation of the first and second lottery (a negative coefficient indicates participants' risk averse behavior). The variable dsd * loss is equal to the latter when the expected value of the two lotteries is negative: this coefficient captures the loss aversion of participants.