Running head: Reasonable Reasons Reasonable Reasons for Waiting

Recent decision-making research claims to establish that, in violation of Savage’s normative sure-thing principle, individuals often wait to acquire noninstrumental information and subsequently base their decisions upon this information. The current research suggests that characterizing individuals as pursuing noninstrumental or useless information may be overstated. Through a series of experiments we establish, first, that many people choose to wait, even when waiting provides no additional information at all. Second, the longer people are allowed to wait before having to decide, the more people prefer to wait rather than decide immediately. Third, those individuals who choose to wait are the ones less confident about committing themselves to a decision. For them, the benefit from waiting may be especially valuable by allowing them to come to terms with a less-than-ideal decision. Reasonable Reasons 3 Reasonable reasons for waiting When facing an important decision conventional wisdom suggests seeking out all information that may have a bearing on subsequent choice. But not all information is relevant or worth pursuing. Information that is unlikely to change our preferences or alter our choices is probably not worth pursuing or waiting for. For example, an inexpensive, well-timed vacation package to Jamaica may be worth booking independent of the pending weather forecast from which the potential traveler will learn if the temperatures in Jamaica will be in the high 80s or the low 80s. If one would make the same decision regardless of how the uncertainty is resolved, then waiting for its resolution before deciding seems irrational (Savage, 1954). Despite the intuitive appeal of this “sure-thing” principle, recent decisionmaking research has demonstrated that it is often violated (Bastardi and Shafir, 1998, 2000; Croson, 1999; Shafir and Tversky, 1992; Tversky and Shafir 1992). In their 1998 article entitled, “On the Pursuit and Misuse of Useless Information”, Bastardi and Shafir used several scenario studies to document that people wait to pursue information that ultimately has no bearing on the decision they make. Many of their scenarios share the following format: in one “certain” scenario, the respondent is asked to decide between two options; there is no uncertainty. The second “uncertain” scenario makes two simultaneous additions: uncertainty about the resolution of a consideration potentially relevant to the decision task is introduced and the respondent is given a third option, namely, to wait until the resolution of uncertainty before deciding. For example, one pair of their scenarios entails the decision of Reasonable Reasons 4 whether to accept a college applicant who has supportive letters of recommendation and is active in extra-curricular activities, but who has a B average from high school in the certain version or an A or a B average, subject to clarification, in the uncertain version. In this paper, we select one of their scenarios, to be referred to as the course registration scenario, and demonstrate the difficulties inherent in comparing responses from a two-option scenario with no uncertainty to a three-option scenario containing uncertainty. In the course registration scenario (problem 1 in Bastardi and Shafir, 1998) the participants were asked if they would register for an interesting course in their major, knowing that the popular professor who usually teaches it (who is reputed to be an excellent teacher) will be on a sabbatical leave. A less popular professor will instead teach the course. In this “certain” version 82% of the respondents indicated that they would register for the course while only 18% chose not to register. A separate group of respondents were exposed to an “uncertain” version of this scenario: participants were told that the popular professor who usually teaches the course may be on sabbatical and they would not know for certain until tomorrow whether he or a less popular professor will teach the course. The participants had three response options to choose from, namely, register, not register, and wait until tomorrow (after finding out who will be teaching the course) to decide about registering. The “waiting” option was preferred by 56% of the participants, with only 42% deciding to register, and 2% deciding not to register. According to Bastardi and Shafir, the initial finding that 82% chose to register for the course even though they knew for certain that it would be Reasonable Reasons 5 taught by the less popular professor indicates that, “the missing information regarding the professor had no instrumental value for more than 80% of participants and, thus, less than 20% of participants were expected wait.” (p. 21). This raises the question, why do more than half of the respondents in the uncertain version choose to wait? The first point we would like to make is that people may have different motives that could lead them to delay their decision other than, or in addition to, the pursuit of more information. Sometimes people prefer to wait simply to think matters over, to come to terms with a difficult decision, or to consult with friends or family. The need for further contemplation may be particularly intense when one has to choose the lesser of two evils. The participants in the certain version of the course registration scenario are faced with choosing between registering for a poorly taught course and passing up a course containing very interesting subject matter. This dilemma may have introduced serious doubts in the minds of the respondents. Some of them may have chosen to take the course despite these doubts simply because they were forced to decide between registering and not registering. Given the chance, these individuals may have preferred an option such as “I will need more time to think about it”. Unfortunately, this choice was unavailable to them. Their response options placed them in a deterministic, binary mold, which made them appear determined even if in fact they were not. In view of the above, the reference to the 82% of the respondents who chose to register for the course in the two-option, certain condition as those who have no value for waiting may be overstated. Similarly, the choice of waiting in the uncertain version cannot be automatically equated with waiting Reasonable Reasons 6 for more information. Some of the respondents may have chosen to wait because they felt that they needed more time for contemplation. Although Bastardi and Shafir acknowledge that people may have different reasons for waiting, their methodology, in which they compare a two-option, fullinformation world with a three-option, uncertain one, does not afford any insight into possible motives other than information seeking. The first study reported here introduces a proper control condition for Bastardi and Shafir’s three-option, uncertain condition by including a threeoption, certain condition. We show that even when no new information is forthcoming, many people like to wait. Furthermore, increasing the waiting period enhances the attractiveness of waiting. Bastardi and Shafir (1998) argued further that people who pursue noninstrumental information misconstrue it as instrumental and in so doing allow it to affect their choices in accordance with the resolution of uncertainty. Indeed, in their uncertain version when the participants who chose to wait were told, “It is now the next day and you found out that the less popular professor will be teaching the course, do you a) decide to register for the course? b) decide not to register for the course?”, nearly half of those who waited decided not to register for the course. Following our second set of studies we suggest a different way to interpret these results. In experiments 2, 3, and 4, we demonstrate that the option to postpone deciding has the greatest appeal for those who lack confidence in their decision. We therefore reason that the lack of confidence that makes waiting appealing also serves to sensitize the individual to any incoming information that may be instrumental in resolving doubts. Thus, we Reasonable Reasons 7 suggest that the effect of the added information should be attributed to the psychological state of its recipients who chose to wait, rather than to a process by which this information is endowed with extra weight simply because one waited for it. Experiment 1 Two-hundred and seventy-five students at Ben-Gurion University were asked to read and respond to one of five versions of the course registration scenario adapted from Bastardi and Shafir (1998). Following are the versions used, and the participants’ responses to each version. Versions 1 and 2: Replication The first two versions were identical to those employed by Bastardi and Shafir (translated to Hebrew by the authors). These were included in an attempt to replicate previous findings. It is important to note that in Israel, as in many universities in the United States (where the original study was conducted), students are offered a large selection of “elective courses” from which to choose. Unlike required courses, many of the elective courses are not offered on a regular basis. Thus, the decision not to register for an elective course could mean that the student will not have another opportunity to take it, but instead must select a different elective course. Such a decision does not delay the student’s completion of the degree. Given the similarity in the course systems in Israel and in the U.S., the adaptation of the course registration scenario for Israeli students required merely a translation. Following are the versions used, and the participants’ responses to each version. Reasonable Reasons 8 Version 1 (Certain): “You are considering registering for a course in your major that has very interesting subject matter and will not be offered again before you graduate. While the course is reputed to be taught by an excellent professor, you have just discovered that he will be on leave, and that a less popular professor will be teaching the course. Do you a) decide to register for the course? b) decide not to register for the course?” Version 2 (Uncertain): “You are considering registering for a course in your major that has very interesting subject matter and will not be offered again before you graduate. While the course is reputed to be taught by an excellent professor, you have just discovered that he may be on leave. It will not be known until tomorrow if the regular professor will teach the course or if a less popular professor will”. In addition to the response options of “register” and “not register”, this version included a third option; “c) wait until tomorrow (after finding out if the regular professor will be teaching the course) to decide about registering for the course.” Results Table 1a: Proportion of Participants Choosing Each of the Options Version N % Register % Not register % Wait 1. Certain 55 81.82 18.18 --------2. Uncertain 56 48.21 3.57 48.21 Following Bastardi and Shafir (1998), participants who chose to wait were asked to respond to the second part of the uncertain version questionnaire: “It is now the next day, and you find out that the less popular Reasonable Reasons 9 professor will be teaching the course. Do you a) decide to register for the course? b) decide not to register for the course?” Table 1b: Decisions Following the Waiting Period Version % Register % Not register Total 2. Uncertain 21.42 26.79 48.21 The results of versions 1 and 2 successfully replicated Bastardi and Shafir’s (1998) original findings. In the two-option, certain version, their results were precisely duplicated with 82% of respondents choosing to register and 18% who decided not to register. In the uncertain, three-option version, as in Bastardi and Shafir, a large proportion (nearly half) of the participants chose to wait for the professor information. The effect of the resolution of uncertainty in the second part of the questionnaire on participants’ decisions was marginally significant: a total of 29.36% chose not to register in the three-option, uncertain version (3.57% before and 26.79% after pursuing the information) in contrast to only 18.18% of the respondents in the two-option, certain version, χ(1)=2.23, p<.07. Version 3: Three-Option Certain It may be that even in the absence of uncertainty, some participants might welcome the opportunity to postpone their decision in order to contemplate which outcome is more acceptable, forgoing an interesting course or attending a poorly taught one. To investigate this hypothesis a third, “three-option certain” version was created. This version was identical to version 1 (that is, certain knowledge that the course will be taught by the less Reasonable Reasons 10 popular professor), with one addition, the inclusion of a third alternative, namely, “c) wait until tomorrow to decide about registering for the course.” Participants who chose this option were subsequently asked to respond to the second part of the questionnaire and indicate their decisions the following