40 Charles River (cd) rats were trained to run down a 15-ft. alley and bar press for food in a 2 × 2 analysis of variance design with alcohol and control treatments and two conflict-producing methodologies. In Exp. 1, conflict was induced by shocking S when it bar pressed; in Exp. II, when S touched the food dish. After approach-avoidance conflict was established, half the Ss in each experiment were injected intraperitoneally with 2.0 cc/100 gm. body weight of 1.0 gm. alcohol/Kg. and half with an equitonic control dextrose solution. When Ss were returned to the alley with shock off, 5 alcohol and 1 control Ss in Exp. I and 4 alcohol and 2 control Ss in Exp. II resolved the conflict to resume running, bar pressing and eating. Analysis of variance revealed that alcohol was significant in determining distance run. Compared with controls, alcohol-treated is ran farther and more frequently touched the dish and the bar. Data support the thesis that alcohol attenuates conflict in rats and suggest that more discrete dependent variables, such as bar pressing, temporally and spatially mote closely associated with the conflictful goal, may be less significantly affected by alcohol than grosser measures of conflict such as running.
What do instrumental variable models deliver with discrete dependent variables?
