Context in source publication

Context 1
... we control for the degree of risk aversion and what the subjects observed in previous rounds (History), as we did in the case of depositor 1. Our estimates for this specification are reported in Table 3, which also presents the regressions in which men and women's behavior is studied separately. 20 20 We note that the dummy variables F1ObsWa, WObsWa and WF1ObsWa have been eliminated from the analysis because (i) men who observed a waiting and know that their actions will be observed, and (ii) women who observe a waiting (regardless of whether they are linked with depositor 3) never withdraw their money from the bank; i.e., F1ObsWa, WObsWa and WF1ObsWa predict waiting perfectly. ...