Search

Psychologists have long been intrigued by how humans and nonhuman primates process magnitudes such as “how far, how fast, how much, how long, and how many.” This fascination perhaps stems from the fact that magnitude estimations of space, time, and number form the bedrock of most of the decisions that we make in daily life. Decisions about how many cookies to eat, how many payments to make, how many days to wait for a product, all have one thing in common: they require a fundamental ability to be able to discern discriminable elements of the type of stimulus in question. Although distinguishing one from many is an ability that is shared by humans and nonhuman primates, what makes research in this area particularly intriguing is the layer of complexity that arises when we take our ability to mathematically represent different quantities in different units (e.g., 1 month, 4 weeks, 30 days) and map it on to this more fundamental ability. The mapping of this numerical system onto a more generalized magnitude representational system allows us to raise the basic question: do magnitude estimates change when they are represented in a different unit or metric? The current collection of articles on numerosity and consumer behavior (appearing over the last two years) complements and adds to a growing body of work that has already appeared in JCR. The articles start with the assumption that a multitude of physical stimuli, regardless of the domain, can be represented in memory using a magnitude representational system. However, the overlay of a verbal, representational system that draws on our knowledge of mathematical symbols and provides a common basis for interpreting these quantities can sometimes produce effects that skew our ability to estimate things correctly. Thus, although there is no rational reason to suppose that a week might differ from 7 days, it does. The reasons for the occurrence of this numerosity effect are still unclear. Nevertheless, the effect leads to errors in estimation that affect not only how we make progress in goals that are set but also affect, at a more fundamental level, how we perceive things. The first two papers in this collection document this. The first article, by Pandelaere, Briers, and Lembregts, shows that expressing an attribute in a different unit leads to greater perceived difference if the unit is on an expanded or finer-grained scale. Thus, for example, consumers see a greater difference between products that express warranty information in months (84 months) as opposed to years (7 years). The effects are attributed to a tendency for people to focus on the numbers rather than the units in which the quantity is represented. One consequence of this tendency is that as the perceived difference increases, consumers are more likely to switch to the better quality option. The authors show the implications of this for decision making by tracking how consumers switch in a variety of domains that range from product decisions to picking a healthy snack. Interestingly, drawing attention to the possibility that this estimate can be represented in different units eliminates the effect. The second article in this series, by Bagchi and Li, examines the implications of scales that use a greater number of units in the context of loyalty programs (e.g., “earn 10 points per dollar spent and claim your reward when you reach 1,000 points” vs. “earn 1 point per dollar spent to claim a reward when you reach 100 points”). In their studies, not only does the scale representing the distance to be traveled to get the reward vary but also the step sizes to get there. At issue, then, is what consumers use in order to infer their goal progress relative to that of another person. Their findings suggest that when the step size information is ambiguous, goal progress is inferred largely from the magnitude of the scale. Consumers perceive themselves to be further away relative to someone else when the scale consists of a

This analysis introduces the Presenter’s Paradox. Robust findings in impression formation demonstrate that perceivers’ judgments show a weighted averaging pattern, which results in less favorable evaluations when mildly favorable information is added to highly favorable information. Across seven studies, we show that presenters do not anticipate this averaging pattern on the part of evaluators and instead design presentations that include all of the favorable information available. This additive strategy (“more is better”) hurts presenters in their perceivers’ eyes because mildly favorable information dilutes the impact of highly favorable information. For example, presenters choose to spend more money to make a product bundle look more costly, even though doing so actually cheapened its value from the evaluators’ perspective (study 1). Additional studies demonstrate the robustness of the effect, investigate the psychological processes underlying it, and examine its implications for a variety of marketing...

In public discourse and survey research, global climate change is sometimes referred to as ''global warming'' and sometimes as ''climate change.'' An analysis of web sites of conservative and liberal think tanks suggests that conservatives prefer to use the term ''global warming'' whereas liberals prefer ''climate change.'' A question wording experiment (N ¼ 2267) illustrates the power of these frames: Republi- cans were less likely to endorse that the phenomenon is real when it was referred to as ''global warming'' (44.0%) rather than ''climate change'' (60.2%), whereas Democrats were unaffected by question wording (86.9% vs. 86.4%). As a result, the partisan divide on the issue dropped from 42.9 percentage points under a ''global warming'' frame to 26.2 percentage points under a ''climate change'' frame. Theoretical and meth- odological implications are discussed.

Two studies examine how different emotions of the same valence influence product evaluation when products make specific emotional claims. Vacation products with adventurous (serene) appeals were evaluated more favorably when participants felt excited (peaceful) rather than peaceful (excited). This emotion‐congruency effect was not observed when participants were aware of the incidental nature of their feelings (study 1) and was mediated by the influence of feelings on participants’ expectations that the product will deliver what it promises (study 2). The findings show that consumers differentiate between distinct positive emotions and use them as information in assessing a product’s emotional claims.

Self-reports are a key source of information in consumer research. Unfortunately, self-reports are highly context dependent, and this problem is compounded when comparisons across cohorts or cultures are of interest. Age-related changes in cognitive functioning and cultural differences in cognition and communication influence the response process, resulting in differential context effects that may reverse the ordinal placement of cohorts or cultures on the measure of interest. Any observed difference between age groups or cultures may therefore reflect a meaningful difference in attitudes or behaviors, a difference in the response process, or an unknown mix of both. Copyright 2003 by the University of Chicago.

The activation of ageism and aging stereotypes in younger and older adults was investigated by manipulating both the valence and the stereotypicality of trait stimuli. Participants completed a lexical decision task in which the stimulus onset asynchronies (SOAs) between the prime and target stimuli were varied to examine the effects of automatic and controlled processing (300 and 2,000 ms, respectively). Both younger and older adults demonstrated strong stereotype activation for elderly stereotypes but relatively weak activation for young stereotypes. Both younger and older adults also demonstrated a positive bias toward older people, which was not moderated by SOA. These findings suggest that younger and older adults do not differ in their accessibility to aging stereotypes or to their age-based biases, which appear to be positive toward elderly people.

Emotion and rationality are fundamental elements of human life. They are abstract concepts, often difficult to define and grasp. Thus throughout the history of Western society, the head and the heart, concrete and identifiable elements, have been used as symbols of rationality and emotion. Drawing on the conceptual metaphor framework, we propose that people understand the abstract concepts of rationality and emotion using knowledge of a more concrete concept—the vertical difference between the head and heart. In six studies, we show a deep-seated conceptual metaphorical relationship linking rationality with “up” or “higher” and emotion with “down” or “lower.” We show that the association between verticality and rationality/emotion affects how consumers perceive information and thereby has downstream consequences on attitudes and preferences. We find the association to be most influential when consumers are unaware of it and when it applies to an unfamiliar stimulus. Because all visual formats—from the printed page to screens on a television, computer, or smartphone—entail a vertical placement, this association has important managerial implications. Our studies implement multiple methodologies and technologies and use manipulations of logos, websites, food advertisements, and political slogans.

Respondents of a mode experiment were asked to rate politicians along 11-point scales with different numeric values, ranging from "don't think very highly" (0 or -5) to "think very highly of this politician" (10 or +5). As predicted, all politicians received more favorable ratings along the -5 to +5 than along the 0 to 10 scale. Moreover, this effect was independent of the mode of data collection used, demonstrating that the impact of numeric values does not depend on their visual presentation. Schwarz, Knauper, Hippler, Noelle-Neumann, and Clark (1991) observed that the numeric values provided as part of a rating scale may influence respondents' interpretation of the endpoint labels, resulting in pronounced differences in the obtained responses. In one of their studies, German adults were asked to rate their success in life along an 11-point rating scale, with the endpoints labeled "not at all successful" and "extremely successful". When the numeric values ranged from 0 ("not at all successful") to 10 ("extremely successful"), 34 percent of the representative sample endorsed values between 0 and 5. However, only 13 percent endorsed formally equivalent values between -5 and 0, when the scale ranged from -5 ("not at all successful") to +5 ("extremely successful"). Subsequent experiments indicated that this difference reflected differential interpretations of the term "not at all successful." When the label was combined with the numeric value 0, respondents interpreted it to reflect the absence of success. However, when the same label was combined with the numeric value -5, they interpreted it to reflect the presence of failure. As this example illustrates, the specific numeric values presented as part of a rating scale may change the meaning of the scale's verbal endpoints, essentially resulting in responses to different questions. In the present paper, we provide a conceptual replication of this finding in a different content domain and compare the relative strength of the impact of numeric values under telephone and mail survey conditions. Specifically, we asked German respondents to evaluate politicians along 11-point rating scales, running from 0 (or 5) = "don't think very highly of this politician" to 10 (or +5) = "think very highly of this politician". We assume that the verbal label "don't think very highly of this politician" is interpreted as reflecting the absence of a positive evaluation if combined with the numeric value 0, but the presence of an explicitly negative evaluation if combined with the numeric value -5. Accordingly, the mean response should be more favorable if provided along a -5 to +5, rather than a 0 to 10 scale.

Questions assessing comparative judgments are often phrased as directed comparisons, that is, a stimulus A (subject) is to be compared to a stimulus B (referent); for example, "Is tennis more exciting than soccer or less exciting?" Tversky`s work on judgment of similarity indicated that comparing A to B may result in different similarity judgments than comparing B to A, The four studies reported in this article extend this work from judgments of similarity to evaluative judgments in general. The results demonstrate that the direction of comparison elicited by the wording of the question can have a strong impact on the obtained results. In some instances, a reversal in the direction of comparison (i.e., comparing A to B vs. B to A) resulted in a reversal of the ordinal ranking. Implications for question wording are discussed.

Data of a mode experiment demonstrate that the emergence of context effects may be independent of question order under mail survey conditions. Under telephone interview conditions, substantively related questions affected responses to the target question only when asked first. However, the same questions affected responses under mail survey conditions independent of whether they preceded or followed the target question. One of the key differences between modes of data collection pertains to the temporal order in which the material is presented (see Schwarz et al. 1991). Telephone and face-to-face interviews have a strict sequential organization, and respondents have to process the information in the temporal succession in which it is presented by the interviewer. As a result, preceding questions may influence answers given to subsequent questions, whereas subsequent questions cannot influence the answers already provided to preceding questions. In contrast, selfadministered questionnaires do not have a strict sequential ordering. Although the questions are presented in a predetermined order, respondents may read ahead, go back and forth between related questions, or may change previous answers. Accordingly, a selfadministered questionnaire format may attenuate the impact of question order. This, however, does not imply that the content of related The reported research was supported by grant SWFOO44 6 from the Bundeminister fur Forschung und Technologie of the Federal Republic of Germany to Norbert Schwarz. Both authors were affiliated with ZUMA, Mannheim, Federal Republic of Germany, at the time of the reported research. NORBERT SCHWARZ is now professor of psychology and research scientist, Institute for Social Research, University of Michigan. HANS-J. HIPPLER is now director of media and market research with the Gesellschaft fur Zeitungsmarketing, Frankfurt. Address correspondence to Norbert Schwarz, Institute for Social Research, University of Michigan, Ann Arbor, MI 48106-1248. Public Opinion Quarterly Volume 59:93-97 ? 1995 by the American Association for Public Opinion Research All rights reserved. 0033-362X/95/5901-0004$02.50 This content downloaded from 157.55.39.255 on Mon, 01 Aug 2016 05:44:41 UTC All use subject to http://about.jstor.org/terms 94 Norbert Schwarz and Hans-J. Hippler questions may not affect the answers given. Rather, it suggests that the emergence of context effects may be independent of question order in a self-administered format, allowing for influences of preceding as well as following questions. In line with this hypothesis, Bishop et al. (1988) observed that the operation of the norm of evenhandedness depended on question order under telephone, but not under self-administered, conditions. Replicating Schuman and Ludwig's (1983) Japanese trade items with a German sample, they asked respondents to report their opinion on (a) limiting Japanese imports to Germany and (b) limiting German exports to Japan. Under telephone interview conditions, German respondents were significantly more likely to favor limiting Japanese imports to Germany than they were to favor limiting German exports to Japan when each question was asked in the first position. However, support for limitations on German exports to Japan increased when this question was preceded by the question about restricting Japanese imports to Germany. This presumably reflects that endorsing restrictions on Japanese imports to Germany evoked a norm of evenhandedness (Schuman and Ludwig 1983). Thus, the data under telephone interview conditions replicated Schuman and Ludwig's findings based on a U.S. sample. But when respondents were asked these same questions under selfadministered conditions, the order in which they were presented had no significant effect on the results. Rather, these respondents reported considerable support for limiting German exports to Japan under both order conditions, suggesting that the norm of evenhandedness was evoked independent of question order. Data from a U.S. sample followed the same pattern, although the differences did not reach significance (see Bishop et al. 1988). These findings suggest that respondents either read ahead or returned to the previous question once they encountered the subsequent one. The latter may be particularly likely when the subsequent question draws attention to strongly held norms that have clear implications for a preceding question, as is the case for the Japanese trade items. Whether order-independent context effects would be observed in mail surveys under other conditions is an open issue. To address this issue, we assessed the emergence of question order effects under telephone and mail survey conditions in a mode experiment, using questions that do not evoke the norm of evenhandedness.

A theoretical model of the emergence of assimilation and contrast effects in part-whole question sequences is pre- sented. When one specific question precedes a general question and the two are not assigned to the same conversational context, respondents use the information primed by the specific question to form the general judgment. This results in part-whole assimila- tion effects. If both questions are perceived as belonging to- gether, however, conversational norms of nonredundancy pro- hibit the repeated use of information that has already been provided in response to the specific question when making the general judgment. Accordingly, respondents interpret the general question to refer to aspects other than the ones covered by the specific question. Contrast effects may emerge in that case under specified conditions. If several specific questions precede the general question, however, the general one is always interpreted as a request for a summary judgment. This results in assimilation effects, even under conditions that would foster contrast effects if only one specific question is asked. The model is supported by experimental data and provides a coherent account of apparently contradictory findings previously reported in the survey liter

Three experiments indicate that the numeric values provided as part of a rating scale may influence respondents' interpretation of the endpoint labels. In experiment 1, a representative sample of German adults rated their success in life along an 11-point rating scale, with the endpoints labeled "not at all successful" and "extremely successful." When the numeric values ranged from 0 ("not at all successful") to 10 ("extremely successful"), 34 percent of the respondents endorsed values between 0 and 5. However, only 13 percent endorsed formally equivalent values between 5 and 0, when the scale ranged from -5 ("not at all successful") to +5 ("extremely successful"). Experiment 2 provided an extended conceptual replication of this finding, and experiment 3 demonstrates that recipients of a respondent's report draw different inferences from formally equivalent but numerically different values. In combination, the findings indicate that respondents use the numeric values to disambiguate the meaning of scale labels, resulting in different interpretations and, accordingly, different subjective scale anchors. NORBERT SCHWARZ iS program director at ZUMA and Privatdozent of psychology at the University of Heidelberg. BARBEL KNAUPER is research associate at ZUMA. HANS-J. HIPPLER iS project director at ZUMA. ELISABETH NOELLE-NEUMANN is founder and director of the Institute fur Demoskopie Allensbach, and professor emeritus of mass communication at the University of Mainz. LESLIE CLARK iS assistant professor of psychology at Purdue University. The reported research was supported by grant SWFOO44-6 from the Bundesminister fur Forschung und Technologie of the Federal Republic of Germany to Norbert Schwarz. We thank George Bishop, Tory Higgins, and Tracy Wellens for stimulating discussions. Correspondence should be addressed to Norbert Schwarz, ZUMA, P.O. Box 12 21 55, D-6800 Mannheim, Germany. Public Opinion Quarterly Volume 55 570-582 ? 1991 by the American Association for Public Opinion Research All rights reserved 0033-362X/91/5504-0008$02.50 This content downloaded from 40.77.167.53 on Fri, 09 Sep 2016 04:16:31 UTC All use subject to http://about.jstor.org/terms Changes in the Meaning of Scale Labels 571 Rating scales with labeled endpoints are probably the most widely used measurement instrument in social and psychological research. Leaving some concerns about their psychometric properties aside (see Nunnally 1978), the use of these scales does not seem to be very controversial (see Dawes and Smith [1985] for a careful discussion of their general properties and for empirical and psychological justifications for their use). In general, 7-point scales seem to be best in terms of reliability, percentage of undecided respondents, and respondents' ability to discriminate between the scale values (e.g., Cox 1980). Thus, seven plus or minus two is the usual recommendation. Moreover, respondents are able to use rating scales consistently, even in telephone interviews without visual aids (e.g., Hormuth and Bruckner 1985). In addition, verbal rating scales, which provide a label for each scale point, have been found to be more reliable than scales that provide labels for the endpoints only (Krosnick and Berent 1990). Finally, researchers have observed that the terms used to label the endpoints, or to designate the separate values of verbal rating scales, may affect the obtained distribution (e.g., Rohrmann 1978; Wegner, Faulbaum, and Maag 1982; Wildt and Mazis 1978). This suggests that respondents pay close attention to the meaning of the labels provided to them, much as one would hope. Whereas the number of scale points, the inclusion or omission of a neutral point, and the choice of scale labels have received considerable attention in the literature (see Dawes and Smith 1985), the specific numeric values provided have, to our knowledge, not been the topic of theoretical analysis and empirical investigation. Apparently, researchers assume that, for example, a 7-point scale that ranges from " 1 " to "7" is equivalent to a 7-point scale that ranges from " 3" to " + 3," as long as the same endpoint labels are provided. In the present article, we will question this assumption. Drawing on survey data from the Allensbach archive, we will first demonstrate that the specific numeric values provided in a rating scale can have a dramatic impact on the obtained results. We will then discuss different underlying processes and will test their viability in laboratory experiments. Experiments 1 and 2: The Impact of Numeric Values

In the present paper we want to employ an information-processing perspective to explore the impact of the range of response categories provided to respondents on their behavioral reports and subsequent judgments. Central to the present perspective is the assumption that response scales are not simply "measurement devices" that respondents