[ Pobierz całość w formacie PDF ]
.Since results are contradictory, only replication and refinement will make it possible tocomprehend the relationships.Homogeneity-heterogeneity.In all but one study of the homogerteity-heterogeneity dimension,heterogeneity is artificially created by experimenters' remarks.Festinger and Thibaut (41) found for discussion of football but not juvenile delinquency significantlygreater readiness of homogeneous groups to shift opinion as a function of receiving notes from others.Gerard (48) created homogeneity and heterogeneity similarly for discussion of federal aid toeducation, and found changes in opinion to be unrelated to the homogeneity-heterogeneity dimension.Festinger, Gerard, Hymovitch, Kelley, and Raven (40) report no differences in opinion change amonggroups told that "experts" were-239-present compared with control groups.Since there were no influence attempts directly traceable to"experts," this variation appears to have been a manipulation of homogeneity-heterogeneity in the propersense rather than a variation in the prestige dimension (see above). Summary.Differences in susceptibility are related to characteristics of the persons creating theinfluence.Age differences found include greater influence exerted on children by children than by adults,and on younger children by older children than on older by younger children.Several studies havereported greater susceptibility to pressures created by acquaintances than by strangers.Other differencesreported do not reveal a clear pattern.Greater influences are exerted when the other person has a higher status than the subject.This findingis obtained consistently.Uniform results have not been obtained from experiments designed to evaluatethe influence of homogeneity-heterogeneity among members.PERSONAL ACCOUNTABILITY AND COMMITMENTTwo related dimensions of influence exertion are evaluated in tests of these hypotheses: conditionspermitting anonymity should decrease the extent of influence; and personal commitment in a prior,private performance should increase resistance.Anonymity.Mouton, Blake, and Olmstead (103) found that influence was significantly greater whencollege subjects disclosed their personal identity than when conditions permitted anonymity (see above).Deutsch and Gerard (35) and Asch (3) obtained similar results (see above).Reduced susceptibility to pressures under conditions of anonymity has been clearly demonstrated; thisexplains use of the secret ballot as a basic instrument of personal expression.Commitment.Of two studies varying the personal commitment factor, one by Deutsch and Gerard (35)introduced three variations.The highest frequency of shifting occurred when recording was not required,with lowest frequencies resulting when the responses were recorded, personally signed, and given to theexperimenter.Fisher, Rubinstein, and Freeman (43) used a tachistoscopic presentation for the task of indicating thenumber of dots exposed on a trial-by-trial basis.The conditions were: first, judging without partners;second, judging only after the partner; third and fourth, judging both prior to and following the partner.No differences in the frequency of shifting were found for the experimental conditions-240-except for the last ten trials of the second, when the confederate's judgments were five standarddeviations above the mean of the standardizing group: subjects' responses were significantly higher thanthose under commitment conditions.For the fourth condition, most influence was found on the intertrialbasis when the confederate's report was one standard deviation above that of the critical subject.Althoughsubjects did not shift significantly for any one trial, judgments given prior to the confederate steadilyincreased.These results demonstrate the stabilizing influence of personal commitment on a within-session basis, with the partner's influence exerted in an anticipatory fashion.Summary.Anonymity and commitment have opposite effects on an individaal's reaction.An individualwho makes a definite commitment prior to being subjected to pressures resists and maintains his position more strongly.The degree of public commitment is positively related to the degree of resistance toinfluence.INTERACTION WITH OTHER PARTICIPANTS PRIOR TO THE EXPERIMENTAL SERIESSeveral studies have evaluated susceptibility as a function of the kind of relationships existingbetween the leader and other members or among the members themselves.Variations in leadership stylesare of two different types: leader centered vs.group-centered classroom teaching methods and aparticipatory style vs.the supervisory style.Variations in relations among members that have beenstudied are perceived independence on a partner to attain a goal, and perceived success jointly with thepartner in working toward a goal.Leader-Centered vs.Group-Centered Interaction.Approximately half the subjects in the study byBovard (22) participated in "leadercentered" classes, i.e., routine lecture and question-answer typeinteraction and the others in "group-centered" classes with frequent interaction among members.Individual judgments showed significant convergence toward the announced group norm in all groups.No differences were observed for judgments made on the first day.Both the dispersion of initialjudgments and degree of convergence after the announced group norm and during the last week of thecourse were significantly greater for group-centered classes.Participatory vs.Supervisor Leadership Behavior.Maier and Solem (94) arranged for half the groupsto have leaders instructed to encourage member participation, and the other half, observers free-241-to participate only in member roles.No differences in initial judgments were found.After an eight-minutediscussion of the Maier Horse Trading Problem, groups with leaders significantly increased their per centof correct answers.Preston and Heintz (108) had students first give individual rankings of the names of twelve prominentmen for their desirability as President of the United States; next, a group ranking of the twelve names infour- to five-person groups having either participatory or supervisory leaders; and a final individualranking.Final individual rankings of participatory leaders and followers correlated significantly higherwith the group rankings.There also was more shifting from initial to final rankings for those workingunder participatory leaders [ Pobierz całość w formacie PDF ]
zanotowane.pl doc.pisz.pl pdf.pisz.pl milosnikstop.keep.pl
.Since results are contradictory, only replication and refinement will make it possible tocomprehend the relationships.Homogeneity-heterogeneity.In all but one study of the homogerteity-heterogeneity dimension,heterogeneity is artificially created by experimenters' remarks.Festinger and Thibaut (41) found for discussion of football but not juvenile delinquency significantlygreater readiness of homogeneous groups to shift opinion as a function of receiving notes from others.Gerard (48) created homogeneity and heterogeneity similarly for discussion of federal aid toeducation, and found changes in opinion to be unrelated to the homogeneity-heterogeneity dimension.Festinger, Gerard, Hymovitch, Kelley, and Raven (40) report no differences in opinion change amonggroups told that "experts" were-239-present compared with control groups.Since there were no influence attempts directly traceable to"experts," this variation appears to have been a manipulation of homogeneity-heterogeneity in the propersense rather than a variation in the prestige dimension (see above). Summary.Differences in susceptibility are related to characteristics of the persons creating theinfluence.Age differences found include greater influence exerted on children by children than by adults,and on younger children by older children than on older by younger children.Several studies havereported greater susceptibility to pressures created by acquaintances than by strangers.Other differencesreported do not reveal a clear pattern.Greater influences are exerted when the other person has a higher status than the subject.This findingis obtained consistently.Uniform results have not been obtained from experiments designed to evaluatethe influence of homogeneity-heterogeneity among members.PERSONAL ACCOUNTABILITY AND COMMITMENTTwo related dimensions of influence exertion are evaluated in tests of these hypotheses: conditionspermitting anonymity should decrease the extent of influence; and personal commitment in a prior,private performance should increase resistance.Anonymity.Mouton, Blake, and Olmstead (103) found that influence was significantly greater whencollege subjects disclosed their personal identity than when conditions permitted anonymity (see above).Deutsch and Gerard (35) and Asch (3) obtained similar results (see above).Reduced susceptibility to pressures under conditions of anonymity has been clearly demonstrated; thisexplains use of the secret ballot as a basic instrument of personal expression.Commitment.Of two studies varying the personal commitment factor, one by Deutsch and Gerard (35)introduced three variations.The highest frequency of shifting occurred when recording was not required,with lowest frequencies resulting when the responses were recorded, personally signed, and given to theexperimenter.Fisher, Rubinstein, and Freeman (43) used a tachistoscopic presentation for the task of indicating thenumber of dots exposed on a trial-by-trial basis.The conditions were: first, judging without partners;second, judging only after the partner; third and fourth, judging both prior to and following the partner.No differences in the frequency of shifting were found for the experimental conditions-240-except for the last ten trials of the second, when the confederate's judgments were five standarddeviations above the mean of the standardizing group: subjects' responses were significantly higher thanthose under commitment conditions.For the fourth condition, most influence was found on the intertrialbasis when the confederate's report was one standard deviation above that of the critical subject.Althoughsubjects did not shift significantly for any one trial, judgments given prior to the confederate steadilyincreased.These results demonstrate the stabilizing influence of personal commitment on a within-session basis, with the partner's influence exerted in an anticipatory fashion.Summary.Anonymity and commitment have opposite effects on an individaal's reaction.An individualwho makes a definite commitment prior to being subjected to pressures resists and maintains his position more strongly.The degree of public commitment is positively related to the degree of resistance toinfluence.INTERACTION WITH OTHER PARTICIPANTS PRIOR TO THE EXPERIMENTAL SERIESSeveral studies have evaluated susceptibility as a function of the kind of relationships existingbetween the leader and other members or among the members themselves.Variations in leadership stylesare of two different types: leader centered vs.group-centered classroom teaching methods and aparticipatory style vs.the supervisory style.Variations in relations among members that have beenstudied are perceived independence on a partner to attain a goal, and perceived success jointly with thepartner in working toward a goal.Leader-Centered vs.Group-Centered Interaction.Approximately half the subjects in the study byBovard (22) participated in "leadercentered" classes, i.e., routine lecture and question-answer typeinteraction and the others in "group-centered" classes with frequent interaction among members.Individual judgments showed significant convergence toward the announced group norm in all groups.No differences were observed for judgments made on the first day.Both the dispersion of initialjudgments and degree of convergence after the announced group norm and during the last week of thecourse were significantly greater for group-centered classes.Participatory vs.Supervisor Leadership Behavior.Maier and Solem (94) arranged for half the groupsto have leaders instructed to encourage member participation, and the other half, observers free-241-to participate only in member roles.No differences in initial judgments were found.After an eight-minutediscussion of the Maier Horse Trading Problem, groups with leaders significantly increased their per centof correct answers.Preston and Heintz (108) had students first give individual rankings of the names of twelve prominentmen for their desirability as President of the United States; next, a group ranking of the twelve names infour- to five-person groups having either participatory or supervisory leaders; and a final individualranking.Final individual rankings of participatory leaders and followers correlated significantly higherwith the group rankings.There also was more shifting from initial to final rankings for those workingunder participatory leaders [ Pobierz całość w formacie PDF ]