Social-evaluative influences moderate the relationship between price and perceived quality

Abstract

People often perceive products that cost more as having higher quality. Two experiments tested the hypothesis that the effect of price on perceived quality is attenuated when people believe that their judgments of product quality will be shared with other people. Shoppers rated wines that they thought sold for a low or high price, believing that they might have to explain their ratings or that their ratings were private. The prospect of making public ratings eliminated the tendency to rate higher-price wines more positively, but this effect occurred only when participants were told that their judgments would be public before tasting the wines. The findings show that social-evaluative concerns moderate the effects of price on perceived quality.

Keywords::

• Price-perceived quality effect
• Self-presentation
• Impression management
• Social evaluation
• Consumer judgments

People use the price of a product to infer its quality, judging products that cost more as being of higher quality than less-expensive ones (Brucks, Zeithaml, & Naylor, 2000; Gerstner, 1985; Noël & Hanna, 1996; O’Neill & Lambert, 2001). Although using price as a heuristic for inferring quality is not unreasonable, people perceive the association between price and quality to be stronger than it actually is (Agarwal & Teas, 2002; Kardes, Posavac, Cronley, & Herr, 2008; Lichtenstein & Burton, 1989). For example, although the price of wine does relate to its quality (as assessed by experts), increases in price do not map onto equal increases in quality (Horowitz & Lockshin, 2002). In fact, for certain products, price and quality are inversely related (Boyle & Lathrop, 2009).

In a meta-analysis, Volckner and Hofmann (2007) found that the correlation-based effect size for the relationship between price and perceived quality across 71 effects was .286, showing that nearly 9% of the variance in quality ratings was related to the product's price. The relationship between price and perceived quality was moderated by a number of variables. The link between price and judgments of quality was stronger when people were less familiar with the product, for fast-moving consumer goods than for services and durable goods, and for products that cost more. People are also more likely to infer quality from price when unambiguous indicators of quality are not available (Riesz, 1978).

Although research has examined the effects of various factors on the price-perceived quality effect, little attention has been devoted to interpersonal factors that influence people's tendency to infer quality from price despite evidence that consumers’ attitudes and behavior are sensitive to a variety of social variables. For example, people's buying patterns and judgments of products are influenced by the attitudes and behaviors of their friends, the norms of their reference groups, the values of their social class, their culture and religion, and even the mere presence of other people (Argo, Dahl, & Manchanda, 2005; Orth & Kahle, 2008; Puntoni & Tavassoli, 2007; Ratner & Kahn, 2002; Schlosser & Shavitt, 2002). Consumers’ choices are particularly influenced by how they believe other people view particular products and those who buy them (Batra, Homer, & Kahle, 2001; Chow, Celsi, & Abel, 1990; Escalas & Bettman, 2005). Among other things, people may buy products that are unique when they believe that appearing to be different will make a positive impression (Ariely & Levav, 2000; Ratner & Kahn, 2002).

Research also shows that people sometimes express more moderate attitudes than they privately feel when they expect to explain or defend their judgments (Cialdini, Levy, Herman, & Evenbeck, 1973; Cialdini, Levy, Herman, Kozlowski, & Petty, 1976; Sedikides & Herbst, 2002; Sedikides, Herbst, Hardin, & Dardis, 2002). When people are concerned with others’ evaluations of them, they may offer evaluations that are not unusual or extreme and that can be easily explained and defended. Importantly, these effects may occur in judgments of products as in other domains. People wish to be viewed as reasonable, informed, and astute, and thus desire to express opinions of products that portray them as discriminating consumers (Scholl, 2009). Because offering an incorrect or unusual assessment of a product may result in negative social evaluations, people may express less extreme judgments of products than they actually feel.

The effects of real or implied social evaluation on moderation of expressed attitudes and evaluations have been demonstrated in both basic and consumer research, but the implications of these effects for judgments of product quality have not been examined. Specifically, when social-evaluative concerns lead people to moderate their judgments, those concerns may attenuate the effects of extraneous influences that normally affect people's evaluations, including the effects of price on perceived quality.

Our primary hypothesis was that concerns with social evaluations may attenuate or eliminate the relationship between price and perceived quality. When people express moderate evaluations of a target to avoid having to justify or defend their judgments, their evaluations of both low- and high-price items should be less extreme and thus tend to converge. As a result, low- and high-price items will be evaluated more similarly than they otherwise would. In addition, when people think that their judgments may be scrutinized by others, they may consider a product's merits more carefully and be affected less by extraneous factors such as price. These considerations suggest that expecting that one will explain one's evaluation of a product to another person should attenuate the relationship between the price of a product and ratings of its quality.

Study 1

In Study 1, shoppers rated wines that they thought sold for a low or high price, believing that they might have to explain their ratings or that their evaluations were private and anonymous. Based on research on attitude moderation, we predicted that participants’ ratings of the wines would be less extreme when they thought that they would explain their ratings to another person than when their judgments were private. As a result, people who believe that they might have to explain their evaluations should show a smaller difference in their ratings of low- and high-price wines, thereby attenuating the effects of price on perceived quality.

Method

Participants

A total of 85 participants (26 men, 38 women, and 21 who did not report gender) were recruited from shoppers at a retail wine store in central North Carolina. They ranged in age from 25 to 71 years (M = 46.0, SD = 12.25).

Procedure

The study was conducted near the entrance of a mid-sized wine store. The research area was situated such that shoppers could be approached soon after entering the store. The location also ensured that new participants would not be influenced by those who had already participated in the study.

A male researcher greeted the shopper, introduced himself, and invited the shopper to participate in a wine-tasting study. If the shopper agreed, he or she signed an informed consent form and completed a background questionnaire. Participants indicated their sex and age and rated the degree to which they were knowledgeable about wine (1 = not at all; 7 = extremely), felt confident in their ability to tell wines apart (1 = not at all; 7 = extremely), and were knowledgeable about wine compared to everyone who shopped at the wine store the same day (1 = I know much less about wine; 7 = I know much more about wine). Participants also answered questions regarding how many glasses of wine they consume in an average week, the number of wine tastings they had attended, and how much they spend on wine in an average month.

The researcher randomly assigned each participant to one of four experimental conditions in a 2 (price: low vs high) by 2 (social context: public vs private) factorial design. Two wines, one red and one white, were used to ensure that effects were not specific to a particular wine.1 Although both wines sold for $14.99 in this store, the cost of each wine was experimentally varied by displaying its ostensible price on a placard placed in front of each wine bottle on a small table. In the low-price condition, the two wines were labeled as $9.49 and $9.99, and in the high-price condition, the wines were labeled as $19.49 and $19.99. For each participant, both wines were always in the low-price condition or in the high-price condition. Two low prices ($9.49 and $9.99) and two high prices ($19.49 and $19.99) were used to minimize suspicion, with the prices counterbalanced between the two wines. (Analyses revealed no order effects.) The prices were changed from the low-price to the high-price condition (and vice versa) when shopper traffic slowed to a point at which changing prices would not be noticed by shoppers who were eligible to participate. The wine bottles were in wine bags so that the label would not affect ratings, and a one-ounce pourer was used so that the amount of wine offered for each tasting was the same.

Before tasting the wines, participants read an instruction sheet that reiterated the price of the wine and varied the way in which participants would report their ratings. For participants in the public context condition, the instructions stated “After rating the wine, please be prepared to explain your ratings to the wine associate who poured the wine. He may ask specific questions about your ratings of the wine.” For participants in the private context condition, the instructions said “After rating the wine, please put your rating sheet in the envelope so that your ratings will be completely private and anonymous.”

Participants then tasted the first of two wines (counterbalanced between red and white; analyses revealed no order effects) and completed a rating questionnaire. They rated the extent to which they felt that the wine was high-quality (1 = strongly disagree; 7 = strongly agree) and rated the wine from 1 (very poor) to 100 (excellent). Particpants then indicated the maximum price that they would be willing to pay for the wine by circling one of 18 prices that increased in $1.00 increments from $6.99 to $23.99. Participants then rated the wine on various gustatory sensations: pleasant aftertaste, acidic, fresh, harsh, aromatic, bitter, hot (i.e., alcohol taste is too strong), smooth, and watery/thin.

After participants rated the first wine, they were offered water to cleanse their palate and then repeated the procedure with the second wine. After rating the second wine, they were debriefed regarding the true cost of the wines and the purpose of the study.

Results

Participant characteristics

On average, our participants rated themselves as “moderate” in their knowledge about wine (M = 4.2, SD = 1.23) and as having moderate confidence in their ability to tell wines apart (M = 4.0, SD = 1.19). They also rated themselves as “about average” in knowledge about wine compared to other shoppers (M = 4.1, SD = 1.11), not showing the “better-than-average effect” often found in comparative self-ratings (Alicke & Govorun, 2005). The data on wine consumption, wine tastings, and amount spent on wine per month were skewed, with a few participants reporting very high scores on all three measures. Inspection of the median values showed that participants reported drinking an average of 6.0 glasses of wine per week (range 1–20 glasses), spending an average of $75 on wine per month (range $1–$400), and attending an average of 25 wine tastings in their lifetime (range 2–250 tastings). Thus, our sample reflected a diverse but reasonably experienced group of wine consumers.

Wine judgments

Ratings of the two wines were correlated (rs = .48 for wine rating and .39 for overall impression). To control for differences in participants’ tendency to rate both wines positively or negatively, the data were analyzed using 2 × 2 × 2 repeated-measures ANOVA in which price (low vs high) and social context (public vs private) were between-participants factors and wine (red and white) was a within-participants factor. (Preliminary analyses revealed no effects that included gender; ps = .13 − .97).

Wine score

A main effect of price on the wine score (out of 100) indicated that participants rated the wines lower in the low-price condition (M = 69.7, SD = 19.20) than in the high-price condition (M = 78.0, SD = 15.48), revealing a price-perceived quality effect, F(1, 76) = 6.23, p < .05. The main effect of social context was not significant, p = .53.

As predicted, a price by social context interaction was obtained, F(1, 76) = 5.17, p < .05. As depicted in Figure 1, participants in the private context condition rated the wines lower in the low-price condition (M = 67.0, SD = 20.21) than in the high-price condition (M = 82.8, SD = 12.60), revealing a price-perceived quality effect, F(1, 38) = 11.37, p < .01. However, the relationship between price and perceived quality was eliminated in the public context condition in which average wine scores did not differ significantly as a function of price, F(1, 38) = 0.02, p = .88.

Figure 1. Wine score rating as a function of price and social context (Study 1).

Quality

Identical effects were obtained on participants’ ratings of the extent to which the wines were high-quality. Participants rated the quality of the wines lower in the low-price condition (M = 4.4, SD = 1.32) than in the high-price condition (M = 5.0, SD = 1.31), revealing a price-perceived quality effect, F(1, 80) = 5.05, p < .05. In addition a nearly significant interaction of price and social context was obtained, F(1, 80) = 3.31, p = .07. In the private context condition, participants rated the quality of the low-price wines (M = 4.4, SD = 1.49) significantly lower than the high-price wines (M = 5.4, SD = 1.23), revealing a price-perceived quality effect, F(1, 43) = 7.31, p < .05. In the public context condition, price did not affect ratings of quality, F(1, 37) = 0.11, p = .74.

Taste characteristics

The negatively-worded gustatory attributes were reverse-scored so that higher scores reflected positive ratings, and the ratings were summed to index the pleasantness of the wine's attributes (Cronbach's α = .86 for ratings of the red wine and .91 for ratings of the white wine). Participants reported less-positive taste characteristics in the low-price condition (M = 58.7, SD = 10.44) than in the high-price condition (M = 63.2, SD = 11.15), revealing a price-perceived quality effect, F(1, 81) = 5.22, p < .05. Neither the main effect of social context, p = .10, nor the interaction of price and social context, p = .44, was significant.

Maximum price

Participants were willing to pay significantly less for the low-price wines (M = $11.26, SD = $3.11) than for the high-price wines (M = $15.17, SD = $3.62), F(1, 80) = 38.57, p < .001. In addition, the predicted interaction between price and social context was obtained, F(1, 80) = 3.96, p = .05. As seen in Figure 2, in the private context condition, participants were willing to pay less for the wine in the low-price condition (M = $11.05, SD = $2.83) than in the high-price condition (M = $16.17, SD = $3.50), F(1, 42) = 38.10, p < .001. In the public context condition, participants also indicated that they were willing to pay less in the low-price condition (M = $11.49, SD = $3.40) than in the high-price condition (M = $14.12, SD = $3.46), but the size of this difference was significantly smaller in the public than private context condition.

Figure 2. Maximum price as a function of price and social context (Study 1).

Self-perceived expertise

Seeing that shoppers’ ratings were influenced by both the social context and the wines’ ostensible price raises the question of whether these effects are moderated by the wine-related expertise of the participants. An oenophile might insist that people with confidence in their knowledge about wine would not be influenced by the wine's price or concerned about expressing their judgments to others.

To test this possibility, participants’ ratings of their knowledge about wine, their ability to tell different wines apart, and their knowledge compared to others were averaged, after standardizing the scores, to create an index of self-perceived wine expertise (α = .92). Analyses were conducted in which price (low vs high) and social context (public vs private) were dichotomous between-participants factors, self-perceived wine expertise was a continuous between-participants factor, and wine (red and white) was a within-participants factor. Self-perceived wine expertise was unrelated to all three global ratings of the wines (wine score, quality, and maximum price), and it did not moderate any of the effects of price or social context on participants’ evaluations.

Discussion

As expected, price influenced participants’ expressed judgments of the wines, as it does the perceived quality of a variety of other products (Volckner & Hofmann, 2007). As predicted, however, the effect of price on perceived quality was attenuated—and sometimes eliminated—when participants believed that they might have to explain their ratings to the researcher. Participants consistently expressed less-extreme judgments of the wines when they believed that they might discuss their ratings than when they thought that their ratings were private and anonymous. These results extend laboratory research in which the possibility of making incorrect judgments leads people to endorse more moderate attitudes (Cialdini et al., 1973, 1976; Sedikides & Herbst, 2002; Sedikides et al., 2002). The public context eliminated the influence of price on perceived quality by causing ratings of the low- and high-price wines to be less extreme. These effects were not subtle; whereas the wines were rated significantly differently in the private context condition, the effects were highly non-significant in the public context condition.

Two distinct processes could be responsible for the effects of public context on elimination of the price-perceived quality effect. On one hand, the prospect of social evaluation might have made participants think more carefully about the wines’ gustatory characteristics as they tasted each wine. As a result, their judgments were influenced more by the wine's characteristics and less by its ostensible price. On the other hand, social evaluation may have merely led participants to rate the wines less extremely to avoid making an indefensible judgment. Because participants in the public context condition were told that they might explain their ratings before they tasted the wines, the results of Study 1 do not allow us to distinguish between these two explanations.

STUDY 2

Study 2 was designed to determine whether expecting to explain their ratings influenced the tasting process or the rating process. Did the prospect of social evaluation engender a different process of tasting and evaluating the wine, or did it moderate participants’ ratings only at the time that they completed the rating form? If the price–quality effect was eliminated in the public context condition in Study 1 because participants thought about the wines’ attributes more carefully during tasting, then the price–quality effect should be eliminated when participants are told about the possibility of discussing their ratings before they taste the wine. However, if the price–quality effect was eliminated in the public context condition in Study 1 because participants expressed less-extreme ratings on the questionnaire, then the price–quality effect should be eliminated only when participants learn that their ratings might be public after tasting the wine but before providing their ratings.

Method

Participants

A total of 81 participants (30 men, 42 women, and 9 who did not report gender) were recruited at the same wine store as Study 1 (although the store's location had changed). They ranged in age from 23 to 78 years (M = 46.3, SD = 13.91).

Procedure

The procedure was generally the same as in Study 1 except as indicated below. Because the goal of the study was to test explanations of the effect obtained in the public context condition of Study 1, all participants in Study 2 were told that they should be prepared to answer any questions the researcher might have about their ratings. However, participants were randomly assigned to receive this information either before or after tasting the wine.

This modification of the design precluded having participants rate two wines after receiving the manipulation because the second wine would always be tasted after the public context information was provided for the first wine. Thus, to make Study 2 as similar to Study 1 as possible, participants again rated two wines but were told that the first wine was a “practice” tasting to familiarize them with the rating form. After participants tasted and rated the first wine, the researcher set their practice ratings aside without looking at them.

The researcher then randomly assigned each participant to one of four conditions in a 2 (price: low vs high) by 2 (timing of public context information: before-tasting vs after-tasting) factorial design. Participants were told that they would taste and rate a wine that sold for either $9.99 (low price) or $19.99 (high price), and the price was displayed on a placard. Although a different wine was used than in Study 1, it sold in the store for $14.99.2 As in Study 1, the bottle was in a wine bag, and a pourer assured that the amount of wine was the same for each participant.

Participants in the before-tasting condition were told to turn the first page of the rating questionnaire before tasting the wine, where they were immediately informed that they should be prepared to answer any questions that the researcher might have about their ratings. They then tasted the wine and completed their ratings. Participants in the after-tasting condition were instructed to taste the wine before turning the page of their questionnaire booklet and rating the wine. When these participants had tasted the wine and turned the page, they were informed on the next page that they should be prepared to answer any questions that the researcher might have about their ratings. They then rated the wine.

Participants rated their overall impression of the quality of the wine (1 = very negative; 7 = extremely positive), rated the wine from 1 (very poor) to 100 (excellent), and indicated the maximum price that they would be willing to pay for the wine (from $6.99 to $23.99). Participants also rated the wine's taste qualities as in Study 1. Particpants then indicated the amount for which the wine sold in the store (the researcher concealed the placard showing the ostensible price before participants reached this manipulation check question). Finally, participants were debriefed regarding the true cost of the wine and the purpose of the study.

Results

Participant characteristics and manipulation check

As in Study 1, our participants rated themselves as “moderate” in their knowledge about wine (M = 4.2, SD = 1.04), having moderate confidence in their ability to tell wines apart (M = 3.9, SD = 1.10), and average in knowledge about wine compared to other shoppers (M = 4.0, SD = 1.05). The data on wine consumption, wine tastings, and amount spent on wine were also similar to Study 1. Inspection of the median values showed that participants reported drinking an average of 6.5 glasses of wine per week (range 0–23 glasses), spending an average of $100 on wine per month (range $0–$500), and attending an average of 75 wine tastings in their lifetime (range 0–2000). Of the 81 participants, 78 correctly reported how much they were told the wine cost; the other three did not answer this question.

Wine judgments

Ratings were analyzed with a 2 (low vs high price) × 2 (public context introduced before vs after tasting) ANOVA. (Preliminary analyses revealed no effects that included gender; ps = .08–1.00.)

Wine score

Participants rated the wine lower in the low-price condition (M = 67.4, SD = 26.35) than in the high-price condition (M = 79.5, SD = 14.07), revealing a price-perceived quality effect, F(1, 73) = 8.21, p < .01. A main effect of the timing of the public context manipulation was qualified by the interaction of price by timing, F(1, 73) = 7.62, p < .01. As shown in Figure 3, in the before-tasting condition, no relationship was obtained between price and perceived quality, F(1, 38) = 0.01, p = .92. However, in the after-tasting condition, participants rated the wine lower in the low-price condition (M = 55.5, SD = 32.61) than in the high-price condition (M = 81.1, SD = 10.94), revealing a price-perceived quality effect, F(1, 35) = 10.44, p < .01.

Figure 3. Wine score rating as a function of price and timing of public context information (Study 2).

Overall quality

Participants rated the wine lower in the low-price condition (M = 4.6, SD = 1.63) than in the high-price condition (M = 5.4, SD = 1.18), revealing a price-perceived quality effect, F(1, 73) = 7.29, p < .01. The main effect of the timing of the public context manipulation was not significant, p = .57, but the price by timing interaction was obtained, F(1, 73) = 5.65, p < .05. As shown in Figure 4, in the before-tasting condition, participants’ quality ratings did not differ as a function of price, F(1, 39) = 0.06, p = .80. However, in the after-tasting condition, participants rated the low-price wine lower (M = 4.1, SD = 1.94) than the high-price wine (M = 5.7, SD = 0.83), revealing a price-perceived quality effect, F(1, 34) = 10.54, p < .01.

Figure 4. Overall impression of quality as a function of price and timing of public context information (Study 2).

Taste characteristics

Ratings of the taste characteristics were summed (after reverse-scoring negatively-worded characteristics) to produce an index of the pleasantness of the wine's attributes (α = .86). Neither the main effect of price, p = .14, nor the main effect of the timing of the public context manipulation, p = .86, was significant, but the price by timing interaction was obtained, F(1, 75) = 4.03, p < .05. In the before-tasting condition, participants’ ratings of the taste characteristics did not differ as a function of price, F(1, 38) = 0.17, p = .68. In the after-tasting condition, participants rated the wine's taste less positively in the low-price condition (M = 39.0, SD = 12.06) than in the high-price condition (M = 46.3, SD = 7.94), revealing a price-perceived quality effect, F(1, 37) = 5.04, p < .05.

Maximum price

Participants were willing to pay significantly less for the wine in the low-price condition (M = $11.74, SD = $3.83) than in the high-price condition (M = $15.64, SD = $4.80), F(1, 76) = 15.93, p < .001. Neither the main effect of the timing of the public context manipulation, F(1, 76) = 0.03, p = .87, nor the price by timing interaction, F(1, 76) = 1.47, p = .23, was significant.

Discussion

The moderating effects of public context on the price-perceived quality effect observed in Study 1 were obtained in Study 2 only when participants learned before they tasted the wine that they might have to answer questions about their ratings. When participants learned this information after tasting the wine, their ratings were influenced by price.

General Discussion

Price affected perceived quality when participants thought that their ratings were private (Study 1) or when they learned that their ratings would be public only after they tasted the wine (Study 2). However, the effects of price on perceived quality were eliminated when participants were told that they might have to explain their ratings (Study 1), and this effect occurred only when participants learned this information before tasting the wine (Study 2). Unlike most research on the price-perceived quality effect in which participants rate products that they do not actually use or consume, participants in the present study actually tasted the product, yet their ratings of the wines were still influenced by price under certain circumstances.

When participants knew that their judgments were public and open to discussion, they might have focused more intently on the taste of the wines in order to render a more accurate and defensible assessment. Paying extra attention to the sensory qualities of the wine then reduced the influence of extraneous factors such as the wine's ostensible price. Most cognitive biases are not easily modified when people think consciously about their decisions, and thinking consciously about taste judgments lowers the validity of judgments of jams (Wilson & Schooler, 1991). Even so, the tendency to infer quality from price may be weakened when people pay greater attention to the attributes of the product.

According to goal processing theory, contextual factors can activate a variety of goals that influence how people process information (Russo, Carlson, Meloy, & Yong, 2008). Although these goals help people process information in ways that promote accurate and efficient judgments based on information that is relevant to a particular decision, activated goals can also compromise the quality of decisions by making certain information more or less salient and leading people to construe information in goal-congruent ways (Cunha & Shulman, 2011; Russo et al., 2008). Introducing the prospect of social evaluation into people's ratings may activate processing goals that are not active when people rate targets privately. And activating goals associated with social evaluation may lead people to process information about a product differently out of a concern for the interpersonal risks associated with making extreme judgments.

Previous research on attitude moderation (Cialdini et al., 1973, 1976; Sedikides & Herbst, 2002; Sedikides et al., 2002) was not designed in a way that allows strong conclusions regarding precisely when and why participants’ positions became less extreme when they expected to discuss them. The present studies suggest that research is needed to disentangle cognitive effects that occur when people think about a topic from interpersonal effects that occur at the time that people describe their views to others.

Analyses showed that the degree to which participants believed that they were knowledgeable about wines did not moderate the effects. Even if their self-views were inaccurate, one might expect that shoppers with greater confidence would be less susceptible to the price-perceived quality effect (because they trust their taste buds more than cues such as price) and less concerned about making strong claims in the public evaluation context (because they assume that their ratings are defensible).

The results of these two studies identify a boundary condition for the tendency to infer quality from price. Making people's judgments public eliminated the price-perceived quality effect with respect to wine, but research is needed to determine the degree to which this effect generalizes to other objects and products. We can imagine reasons why the effects obtained in this study might be both weaker and stronger when people judge other things. On one hand, taste judgments are inherently subjective and relatively unconstrained by objective indicators of quality, so evaluations of food and drink might be especially susceptible to extraneous influences. On the other hand, the immediate experience of a product's taste may anchor people's evaluations and make them less prone to contextual influences. Furthermore, taste is always a matter of personal preference, so people may be less likely to change taste evaluations out of a concern for making incorrect judgments.

Of course, researchers are well aware of the ways in which people's judgments can be influenced by other individuals (Bearden, Netemeyer, & Teel, 1989; Tetlock, Skitka, & Boettger, 1989) and of the effects of social-evaluative concerns on people's judgments. Yet, the present findings go beyond the general impact of social factors on people's evaluations to show that such concerns can moderate the influence of other factors, such as price, that are known to affect people's judgments.

Notes

¹The red wine was a Windmill Cabernet Savignon, and the white wine was a Windmill Chardonnay. The white wine was stored in a frozen ceramic sleeve to maintain an appropriate temperature.

² The first (practice) wine was a Rock and Vine Cabernet Sauvignon, and the second (target) wine was an Earth, Zin, and Fire Old Vine Zinfandel.