Choice Dilemma Items Research Articles

The relationship between familiarity, gender, disagreement, and status and bouts of solitary and joint laughter

The current research treated laughter as an indexical with two closely allied properties: to designate talk as non-serious and to serve as a mode of address signalling a preference for solidarity. These properties gave rise to four discrete forms of laughter bout, solitary speaker, solitary listener, speaker-initiated joint, and listener-initiated joint laughter, which were examined using 55 same-gender pairs discussing three choice dilemma items. By exploring the associations between the wider contextual factors of familiarity, gender, disagreement and status, and the frequencies of each form of bout within the dyad, it was hoped to establish whether laughter was related to how participants modulated their social relationships. Neither familiarity nor disagreement had any effect on any of the forms of laughter bout, while females were found to demonstrate higher frequencies of joint speaker laughter than males. In unequal status pairs, high status female staff joined in the laughter of their low status female student interlocutors less often than the reverse, a finding comparable with the exchange of other terms of address, such as second person pronouns in European languages. It was concluded that joint laughter was a signal of solidarity and solitary speaker laughter was a declared preference for solidarity, but the significance of solitary listener laughter, beyond an acknowledgement of the speaker’s non-serious talk, remained less clear. It was also noted that norms associated with the setting and topic of interaction were influential in determining the extent to which laughter would be used to modulate the relationships between interlocutors.

Group polarization and choice‐dilemmas: how important is self‐categorization?

AbstractSelf‐categorization proponents (e.g., Turner, 1991) assume that group polarization occurs because discussants wish to differentiate themselves from outgroup positions and implicitly think of such groups even when they are not specifically mentioned. Ingroup/outgroup salience is thought to heighten such effects. To examine this view, we had participants discuss Choice Dilemma items either with or without explicit knowledge of outgroup positions. Contrary to a self‐categorization account, this manipulation of outgroup salience did not affect the degree of group polarization. In addition, rating measures revealed little spontaneous consideration of outgroup positions on the part of participants, nor was consideration of outgroup positions related to degree of polarization. Group members did show evidence of ingroup identification, but this identification was unrelated to participants' post‐discussion conformity to the group consensus. Taken as a whole, these results suggest distinct limits to the self‐categorization interpretation of group polarization involving Choice Dilemmas. Copyright © 2006 John Wiley & Sons, Ltd.

Gender differences in explicit and implicit risk attitudes: A socially facilitated phenomenon

The present study (a) examined the question of whether gender differences in hypothetical risk decisions might be socially facilitated by the presence of gender-homogenous groups and (b) investigated the conscious and non-conscious motivators of risk-taking through the application of both explicit and implicit measures of risk attitude. Using hypothetical choice dilemma items, no gender difference was found at an individual level; however, when placed in-groups, males expressed a stronger pro-risk position than females. While males self-reported a stronger pro-risk position than did females on two explicit measures of risk-attitude, no gender differences were found on two parallel implicit measures. However, a newly developed implicit measure of risk-attitude showed its utility in the form of convergent, predictive and incremental validity with respect to a behavioural outcome.

The validity of choice dilemma response scales

Most choice shift research uses choice dilemma items as stimuli with an “odds of success” response scale. This research shows that the odds scale produces substantial measurement error while a Likert response scale for choice dilemma items does not. The implications of these findings for past and future choice shift research is discussed.

Response Scale Ambiguity as a Moderator of the Choice Shift

The vast majority of choice shift inquiry employs choice dilemma (CD) items. In the main, responses to CD items are scaled by asking participants to report the odds of the risky response being successful before they would attempt it. Recent research challenges the validity of the odds scale. This article reports the results of two experiments that demonstrate that response scale ambiguity influences the interpretation of choice shift processes. Specifically, response scale ambiguity moderates the impact of social comparison and persuasive arguments processes. The former is more potent when the ambiguous odds scale is used, whereas the latter exerts a stronger impact when the less ambiguous Likert scale is employed.

The effects of two forms of decision reappraisal on the perception of pertinent arguments

The experiment examined a central assumption of the informational influence explanation of group polarization: that arguments have inherent qualities which elicit from subjects shared and unchanging evaluations. Subjects were required to make binary decisions either on choice dilemma (CD) items or on attitude items and then assess the relevance of two comprehensive sets of arguments to that decision. On each item, one set of arguments favoured one course of action (CD items) or pole (attitude items) and the second set, the alternative course or pole. Subjects then either reconsidered their decision, or reconsidered it and wrote a passage justifying their current decision, and then reassessed the relevance of the arguments. It was found that after their initial decision, subjects discriminated significantly more clearly between the opposing sets of arguments on the attitude items compared with those on the CD items. Secondly, the trend to accentuate this perceived difference following decision reappraisal was significantly greater when subjects had to justify their decision. Thirdly, this accentuation process was more marked on the CD items than on the attitude items irrespective of the form of decision reappraisal. The informational influence explanation appeared incapable of explaining these results and an alternative interpretation based on social identity theory is offered.

Role of Potential Gain and Loss in Choice-Dilemma Decisions

The effects of potential gain and loss on choice-dilemma decisions were assessed by experimentally manipulating the content of four choice-dilemma items. Results indicated that the loss manipulation had a major impact on both individual decisions and group shifts, but the gain manipulation did not.

Novel argumentation and attitude change: The case of polarization following group discussion

AbstractThis study is concerned with attitude polarization as a function of two properties of a persuasive message: (a) its validity or acceptability and (b) its novelty. The latter is defined as the extent to which the message contains new arguments unlikely to have been already considered by the individual. Acceptability is assumed to be a necessary condition for inducing attitude change; the impact of novelty, therefore, was expected to be most pronounced for arguments of high validity. This hypothesis was tested in two related studies using arguments produced in response to choice dilemma items, widely used in research on polarization. First, it was shown that arguments rated as both valid and novel were perceived as more persuasive than arguments rated either as highly valid but obvious (non‐novel) or as low in validity (non‐valid) but novel. Second, when subjects read samples of valid arguments, their attitudes polarized in the direction advocated by the novel arguments rather than by the non‐novel ones. These findings are considered relevant to the polarization of attitudes in groups. Other research demonstrates that this phenomenon is the result of persuasive arguments raised during group discussion, The present study suggests why such arguments may be persuasive.

What a person thinks upon learning he has chosen differently from others: Nice evidence for the persuasive-arguments explanation of choice shifts

Small shifts in choice occur even without discussion, when individuals merely know each other's preference. This appears to support an interpersonal comparison explanation of group induced shifts in choice and to refute explanations based on persuasive argumentation. The present study demonstrates the contrary, that such effects are consistent with the persuasive-arguments formulation and are obtained under particular conditions specified only by the latter theory, to wit: Knowledge of other's choices is assumed to lead a person to think of reasons (arguments) others might have had for their choices-reasons which ordinarily would not come to mind without this knowledge. Such reasoning functions in the same way as persuasive argumentation during group discussion; it causes the person to persuade himself that an alternative course of action now has greater merit than the one he initially preferred. To test this analysis, an experiment was performed in which subjects responded to choice-dilemma items under three different conditions: Following their own choice (I) they learned what several others had chosen and then wrote arguments in support of alternatives given in that same item; (II) they learned what several others had chosen and then wrote arguments in support of alternatives given in a different item; and (III) they received no information about others' choices but merely wrote arguments on that item. As predicted, shifts in choice occurred only if the person knew what others chose and had an opportunity to think about the latter (condition I); they did not occur if an opportunity to think of others' choices was denied (condition II), nor if knowledge of others' choices was withheld (condition III). Content analysis of the arguments subjects produced in conditions I and II completely supported the hypothesis, as did analyses of responses to postexperimental questionnaire which directly asked the subjects about their feelings and thoughts upon learning what others had chosen.

Age, Group Decisions on Risk‐related Topics and the Prediction of Choice Shifts

Assertions in the literature that the elderly are less willing to make decisions involving risk than are the young were tested by comparing the decision‐making of two samples: one of 36 undergraduates and one of 26 subjects aged 55–65. Neither on their initial decisions nor in their group discussions of choice dilemma items did the older differ from the younger sample. The predictive validity of Stoner's Value Ranking Instrument was tested with both samples; and in neither was the instrument successful in indicating which items subsequently produced cautious shifts and which shifts to greater risk following group discussion. One reason for the inadequacy of Stoner's method of examining particular cultures' value for risk may lie in the abstractness of the instrument he has developed.

Informational influence in group discussion

An informational influence explanation of group-induced shift on choice dilemma items was examined by experimental manipulation and by a mathematical model based on information weighing assumptions. Although the exchange of arguments in an interactive discussion context produced significant response change, passive reading of arguments did not. Examination of the model revealed that at a molar level the mean model prediction for an item corresponded closely with the mean shift observed on that item following discussion. At a more molecular level, the informational model failed to predict the magnitude of specific group shifts on particular items. A suggestion as to how the informational influence explanation could be refined to accommodate these findings was taken from theory and research on the role of cognitive learning and cognitive rehearsal in attitude change.

Normative and Informational Effects of Group Interaction

Sources of group-induced shift on choice dilemma items were examined by comparing the typical discussion treatment with normative and informational treatments. Control condition groups responded to three risky choice dilemma items without prior treatment. Social comparison condition groups responded after being exposed to the percentage distribution of initial responses by control subjects. Relevant arguments condition groups responded after discussing the items without pretest and without awareness of the probability scale. Both experimental treatments produced significant effects when compared to control responses; These and other findings are interpreted as supportive of a two-process explanation of group-induced shift and as contrary to explanations of group shift based on group decision models.

The cross-situational validity of risk as a value

Demand characteristics and risk are confounded on the choice dilemma items. Thus, it is difficult using that instrument to determine if individuals view themselves as more risky than others because of a cultural value of risk or instructional demand characteristics. The present study asked Ss to make judgments for themselves and for others on either the choice dilemmas (with risk-oriented or risk-neutral instructions) or on one of three other measures of risk taking (behavior prediction scale, risk of aversive consequences, zero expected value bets) which have exhibited the risky shift but do not have the instructional problems of the choice dilemmas. Only on the choice dilemmas did Ss view themselves as being more risky. On all other measures, Ss viewed themselves as either equal to or more conservative than others. These findings cast doubt on the validity of value type theories as general explanations for the risky shift.

Interpersonal Comparison Processes in Choice Dilemma Responding

Summary Reliable group-induced response changes have been observed on choice dilemma items and, in extensions of this research, with other stimulus materials as well. The search for a theoretical explanation of these findings requires analysis of the processes that determine pregroup responses. It has been proposed that interpersonal comparison considerations determine initial responses to choice dilemma items by inducing the subject to compromise between his perception of the group norm and his own ideal. This study demonstrated that the order of measurement of “self,” “other,” and “admired” responses significantly affects these responses. The findings are interpreted as contrary to the interpersonal comparison assumption that initial choices are an attempt to present oneself favorably, but consistent with the idea that after one has acted, there is a tendency to perceive one's action as more admirable than that of the average peer.

Effects of discussion on opinions concerning illegal behavior.

Summary In an attempt to generalize recent research on the effects of group discussion and to further elucidate mechanisms responsible for discussion-induced response change, 15 groups of institutionalized males convicted of felony and 14 groups of college males responded, before and after discussion, to three choice dilemma items and three ethical-legal dilemmas. It was predicted that (a) the enhancement of mean initial tendency observed on choice dilemma items would extend to the ethical-legal dilemmas, and (b) shift to increased risk on both types of items would occur if, and only if, subjects tended to perceive themselves as initially riskier than their average peer. On the choice dilemmas (with a Likert type response format) only small nonsignificant shifts occurred. On the ethical-legal dilemmas both the inmate and the college samples significantly increased their preference for the legally deviant action following discussion, although initially perceiving themselves to favor it less than their peers.

SELF-OTHER DISCREPANCY IN ETHICAL RISK TAKING

There is a body of literature which shows a change in the amount of S's risk taking from an individual to a group situation. A currently popular explanation for the risky shift is Brown's (1) view that there is a norm in our society to be risky. Upon entering a group, one or more of the individuals in the group find that they are not as risky as the other members and change their judgments accordingly. If Brown's cultural value of risk exists, then to live up to the norm an individual should view himself as being at least as risky as some hypothetical other individual. This discrepancy has been found by Wallach and Wing (8) using the choice-dilemma items. Similarly, Levinger and Schneider (4) found Ss tended to admire a position more risky than their own. Rettig's ( 5 ) Behavior Prediction Scale has also been used as a measure of risk taking. The scale is somewhat unique in that (a) it involves unethical behavior and ( b ) it breaks down the risk taking situation into several components. Rettig has found from an individual to a group condition an increase in both total risk taking and in emphasis of the reinforcement value of gain (RVgn). Horne ( 3 ) and Schulman ( 7 ) have found respectively a self-other discrepancy and an admire-shift for total scale's risk taking in a direction opposite to what one would predict from Brown's value position. Both Horne and Schulman ascribe their findings to the unethical content of Rettig's scale. Neither author, however, reports any data for the RVgn component of the scale. Twenty-three undergraduate introductory psychology students from Bowdoin College volunteered to take part in the study. Ss were first requested to respond as individuals to Rettig's scale under the usual instructions. After completing all 16 items, they were then requested to go back and for each item mark an M where they thought an average introductory psychology student at Bowdoin would respond. Consistent wlth Hotne's ( 3 ) finding, significantly (z = 3.06, p < .01, 2-tail) more Bowdoin Ss judged their fellow students to be more risky than themselves on the total score. Of 23 Ss, 3 had positive discrepancies and 18 had negative discrepancies in Total score; 2 showed no discrepancy. These results seem to indicate that Brown's value hypothesis does not explain the increase in Total risk taking on the scale. However, Ss themselves were significantly ( z = 2.62, p < .01, 2-tail) more sensitive to the RVgn component than was the hypothetical average other student. On RVgn, 17 students had positive discrepancies and 4 had negative discrepancies. These data indicate that the value interpretation may explain the increase in S's sensitivity to RVgn. There are three other component scores to the scale. None of these have shown a group shifc in past literature and none showed a significant sel€-other discrepancy in the present study, suggesting values may be a mediator of the shift on the RVgn component. On the total score, the unethical nature of the act may be in conflict with the value of risk or some other force acting to produce the risky shift. Presence of conflict is supported by the observation that shifts in total score are not as consistent between studies as is the shift in reinforcement value of gain (2, 6 ) .

Enhancement of Dominant Risk Tendencies in Group Discussion

Much recent research indicates that discussion predictably affects responses to choice-dilemma items. In the present experiment, 12 choice-dilemma items were discussed to consensus by 40 female groups of varying size (2, 3, 5, or 7 members). Group size did not significantly affect shift scores. Over all groups, the mean of initial risk taking on an item was an excellent predictor of the mean amount of shift that item elicited ( r = –.89), a finding consistent with certain models of group decision making as well as with the idea that discussion arguments enhance dominant values. Further analyses of the present and past research indicated that the group decision making models could not account for the observed shifts.

Risky Shift in Marketing Decision Making: A Nonconfirmation

Choice-dilemma items and a marketing case-study task were administered to 48 MBA candidates at the University of Texas in a test of the risky-shift phenomenon. Using consensus decisions, groups were more risk-oriented than the individuals comprising them in the instance of the choice-dilemma items, but not when a familiar, realistic task was performed. Hence, it is necessary to use caution when attempting to generalize the risky shift to decision-making situations that use more realistic measures of risk than those represented by the standard choice-dilemma instrument.

The relationship between riskiness and confidence in choice dilemma decisions

AbstractThe relationship between riskiness of decisions and confidence in decisions was examined using Choice Dilemma items. Graphs show a clear curvilinear relationship between riskiness and confidence; on both risky items and cautious items greater confidence is associate with extremely risky and extremely cautious decisions than with moderate ones. it was also demonstrated that individual subjects have more confidence in their extreme decisions than in their moderate ones.

Responsibility diffusion and drive enhancement effects on risky shift1

Summary Discussion produced expected risky and cautious shifts on choice dilemma items It was predicted from responsibility diffusion theory that, compared with low-anxious, high-defensive groups, high-anxious, low-defensive groups produce greater risky shift It was predicted from drive enhancement theory that, compared with low-anxious groups, high-anxious groups produce greater risky shift on risk-dominant items and greater cautious shift on caution-dominant items The results of two experiments failed to confirm either interpretation Surprisingly, significant shifts were obtained in a control condition