30-06-2014
That would never happen to me! The role of perceived probability in the perception of negative eWOM narratives
Remy de Jong
Student number: 10022244
Master’s Thesis Business Studies (Marketing Track) Supervisor: dhr. dr. Alfred Zerres
Abstract
In this study an attempt was made to apply construal level theory to the phenomenon of electronic word of mouth. An online experiment (N = 247) was conducted to find out whether temporal distance (something happened a long time ago), spatial distance (something
happened far away) and social distance (something happened to a dissimilar person) lead to hypothetical distance, i.e. a lower perceived probability of occurrence, and whether
hypothetical distance mediated the relationship between temporal, spatial and social distance and consumer evaluations and purchase intentions. No support was found for the hypothesis. Possible shortcomings of the experiment and other explanations of the results are discussed and directions for future research are given.
Introduction
People can only directly experience the here and now. However, people remember the past and make predictions about the future, imagine going to other places, put themselves in other people’s places and fantasize about unlikely scenarios. Fairly recently, a theory has emerged that tries to explain how we transcend the here and now to include distal entities (Trope & Liberman, 2010). It is called construal level theory (CLT), and it states that as an event is far away from our direct experience, it is construed using central and abstract features, while events that are close to our direct experience are construed using more concrete and
contextual features (Liberman & Trope, 1998). The distance between an event and someone’s direct experience is called psychological distance, and the four dimensions of psychological distance are temporal, spatial, social and hypothetical distance.
The link between psychological distance and level of construal has many different implications. Research has shown that the different dimensions similarly affect prediction, preference and action (Trope & Liberman, 2008). Because the different dimensions of distance all have the same egocentric reference point (someone’s direct experience), they should be related to each other and similarly affect and be affected by level of construal (Trope & Liberman, 2010). Previous research has found support for most directions of
influence between the different dimensions of psychological distance (e.g. Stephan, Liberman & Trope, 2010). However, no research to date has looked at the influence of temporal, spatial and social distance on hypothetical distance, i.e. whether people think events that happen or have happened far from the present, happen in far away places and concern people that are perceived as dissimilar are perceived as less likely to occur to them. If this is the case, it could provide interesting insights in the perception of online or electronic word of mouth (eWOM) narratives.
Word of mouth is informal advice passed between customers (East, Hammond & Lomax, 2008), and it plays a big role in customers’ buying decisions (Richins & Rootshaffer, 1988; Buttle, 1998). Because WOM is unsponsored, it is perceived as more credible and reliable than traditional advertising (Arndt, 1967), and it has been shown to have an impact on consumer attitudes (e.g. Brucks, 1985) and purchase intentions (e.g. Weinberger & Dillon, 1980). Since the rise of the internet, more and more WOM communication occurs online. The most widely used format of eWOM is the consumer-opinion platform, where consumers can post messages with their opinions and experiences and read about those of others (Hennig-Thurau et al., 2004). As the different dimensions should all be related to each other, events described in messages that evoke a high sense of temporal, spatial or social distance should be perceived as less likely to occur to the reader than messages that evoke a low sense of
distance on these dimensions. If someone reading an eWOM narrative thinks it is unlikely that something similar will occur to him, the effect of the narrative on his attitude towards the brand and his purchase intentions should be smaller compared to when the reader thinks the narrative portrays a situation that the reader could find himself in. This would be an important finding, as many companies (e.g. Amazon) have reduced or even eliminated their spending on traditional advertisement, because they believe that eWOM is better at influencing the
behavior of their customers (Thompson, 2003).
In the present research, an experiment is conducted to test whether temporal, spatial and social distance decrease the perceived probability of an event, and whether perceived probability mediates the relation between an eWOM narrative and a reader’s subsequent attitudes and purchase intentions. Next, the literature about CLT is reviewed, and possible implications for eWOM narrative perception are discussed. After that, the experiment conducted to test the predictions is described. Following the methods section, results of the
experiment are discussed. Finally, the results are discussed and directions for future research are given.
Theoretical framework
Construal level theory was initially called temporal construal theory, and it focused only on temporal distance: it stated that central and abstract features are used to construe events that will occur in the distant future, while events that will occur in the near future are construed by using more concrete and contextual features (Liberman & Trope, 1998). Using abstract features to construe an event is called a high-level construal, while using concrete features to construe an event is called a low-level construal. As an example to illustrate the difference, imagine students sitting in a classroom. A lower-level construal would focus on concrete things such as the students’ age, the color of the room or the clothes of the lecturer, while a high-level construal of the same activity could simply be ‘studying’. Support for this hypothesis, among others, was offered by Liberman, Sagristiano & Trope (2002), who showed that as temporal distance increased, people categorized objects into fewer, broader categories and used simpler structures to organize preferences for events.
Later on, it was found that several other dimensions of distance from direct experience had a similar effect on construal as the distance increased. The dimensions that were added later on to the theory are spatial distance, social distance and hypothetical distance. A study that showed that spatial distance had the same effect as temporal distance was conducted by Fujita et al. (2006), who showed that spatial distance leads people to represent events by their central, abstract features as the reported distance of events increases, whereas they represent events by their concrete, peripheral features when the reported spatial distance of an event decreases. Liviatan, Trope & Liberman (2008) researched whether the manner in which other
people’s actions are represented and judged is influenced by perceived similarity to the one judging his or her actions. Their work was based on the assumption that interpersonal similarity is a form of social distance. They found that as other people were perceived as similar, people focused more on subordinate and secondary features of information when judging their actions. The use of polite, formal language is another way of creating social distance. As Brown (1987) noted, because people address strangers more politely than they address friends, the use of polite language creates a sense of social distance. The last dimension of psychological distance is hypothetical distance, which has to do with the likelihood of the occurrence of an event. A hypothetically near event is an event that is likely to occur, while a hypothetically distant event is one that is unlikely to occur. Wakslak et al. (2006) showed that unlikely events are construed using general, abstract features, while likely events are construed using concrete, specific features. It must be noted that hypothetical distance is a little different than the others. While temporal, spatial and social distance are fairly objective measures of distance, hypothetical distance takes entirely place in the mind. Because of this, one could argue whether this is really a form of ‘distance’ at all. However, for purposes of clarity and consistency with previous literature, the term will be used in this article (interchangeably with ‘perceived probability’).
The link between psychological distance and level of construal has many different implications. Research has shown that the different dimensions have a similar effect on prediction, preference and action (Trope & Liberman, 2008). When predicting future events, little attention is given to low-level contextual circumstances (Wilson, Wheatley, Meyers, Gilbert & Axsom, 2000). According to CLT, this is because as events are further removed from the present, people focus on central and schematic features and tend to neglect
situational details (Trope & Liberman, 2008). CLT’s influence on evaluation stems from the fact that we do not evaluate things themselves, but we evaluate things based on the way we
represent them in our minds. When considering an action to pursue, we can think of its desirability (high-level construal) or its feasibility (low-level construal). Liberman & Trope (1998) showed that people focused more on the desirability of an action when it was in the distant future, while their attention was directed more towards feasibility concerns when it was in the near future. Behavioral intentions are also affected by psychological distance. Eyal et al. (2009) hypothesized that, because of their abstract nature, values will have greater impact on how people plan things in the distant future than the near future. They found support for this hypothesis; values were better predictors of behavioral intentions for
situations far removed from the present than events closer to the present. Also, while values were good predictors of distant future situations, feasibility concerns were better predictors of near future situations.
Because the different dimensions of distance all have the same egocentric reference point (someone’s direct experience), they should be cognitively related to each other and similarly affect and be affected by level of construal (Trope & Liberman, 2010). Research by Stephan, Liberman & Trope (2010) showed that social distance between people (created by using polite language) lead them to believe that the target of communication was both spatially and temporally more distant. Evidence was also found for the reverse direction of influence: when participants wrote messages to be read by tourists either far away (spatial distance) or in the distant future (temporal distance), they tended to use more polite and formal language (Stephan et al., 2010). In another study, it was shown that participants expected unlikely events (hypothetical distance) to happen in remote locations and at a later point in time (Wakslak, 2012). Bar-Anan et al. (2007), using a series of picture-word Stroop tasks (Stroop, 1935), found that participants were faster when the dimensions of distance were congruent than when they were incongruent. These studies support the notion that the
where psychological distance refers to how far an object or event is removed from the
perceiver’s direct experience (Bar-Anan et al., 2007). Research in neuroscience is concurrent with this view, which shows that both temporal and social distance activate the same brain network involving the prefrontal cortex and the medial temporal lobe (Eddis & Schacter, 2008; Schacter, Addis & Bucker, 2007; Buckner & Carroll, 2007).
Most of the directions of influence between the dimensions of psychological distance have been researched. However, there is one direction that has not received attention yet in the literature, and that is the influence of temporal, spatial and social distance on hypothetical distance. In other words, do people think events are less likely to occur when they are events that will happen in the distant future (or have happened a long time ago), in a faraway place and/or by people that are not perceived as similar to themselves. As was shown by Wakslak (2012), there is evidence for the reverse direction (i.e. people expect unlikely events to occur at greater temporal and spatial distance). Her reasoning was that as distant times and locations are thought about more abstractly, a low likeliness of occurrence, which is also thought about abstractly, will match these distant contexts. Some of the studies mentioned above showed that interrelations between the other dimensions of distance work in both directions (e.g. Stephan, Liberman & Trope, 2010). This is because all dimensions of psychological distance lead people to construe events more abstractly. For example, when people think about an event that happened a long time ago, they do so in an abstract way, and these abstract thoughts will lead to higher perceived distance on the other dimensions. Therefore, because hypothetical distance has been shown to increase both perceived temporal and spatial distance (because of the abstract thoughts it evokes), it would be logical that as temporal or spatial distance increases, hypothetical distance will increase as well. This is because a low
probability of occurrence best matches the abstract thoughts that are evoked by the temporal and spatial distance. Wakslak (2012) did not look at social distance, because previous
empirical research already established this pattern for social distance, which was
operationalized as the distinction between the self and someone else (Chambers, Windschitl & Suls, 2003; Kruger & Burrus, 2004). She concluded that even when social distance is not operationalized this way, it should still be positively related to hypothetical distance
(Wakslak, 2012).
CLT and eWOM perception
One area of research where psychological distance, and especially the influence of temporal, spatial and social distance on hypothetical distance, might provide interesting insights, is online or electronic word of mouth (eWOM). As Sweeney, Soutar & Mazzarol (2008) noted, most of the research on eWOM focuses on the sender, and the receiver of eWOM
communication has received fewer attention in the academic literature. As such, there is plenty of knowledge about such topics as what motivates people to articulate themselves on the internet (e.g. Hennig-Thureau et al., 2004) or what the determinants of eWOM are (Chu & Kim, 2011). However, there has been no research to date about the effect of psychological distance evoked by an eWOM narrative, while this could prove to be of importance when looking at eWOM message perception.
Most of the time, consumers write an eWOM product/service review to either
encourage or discourage others from consuming the product/service. Therefore, the reviewers offer positive arguments in favor of the product or negative arguments against it (Sen & Lerman, 2007). While little is known about eWOM message perception, there are many studies about offline message perception and word of mouth. Studies in offline settings have consistently indicated that negative information has a bigger influence on impression
judgments than positive information (Skowronski & Carlston, 1989; see East, Hammond & Lomax, 2008 for an exception). There are different theories that offer explanations for this
negativity effect. Prospect theory states that ‘losses loom larger than gains’ (Kahneman & Tversky, 1979). Translating this finding to eWOM perception, customers who are seeking information about a new product online are more likely to be influenced by negative
information than by positive information. This is because people are driven more strongly to avoid losses than to achieve gains, i.e. driven more strongly to avoid buying a bad product than towards buying a good product (Kahneman, 2011). Another explanation for the negativity dominance can be found in our social environment. Because someone’s social environment contains more positive cues than negative cues, the negative ones will be more salient and also perceived as counter normative. This causes the negative cues to attract more attention and be attributed to the stimulus that caused them more so than positive cues
(Kanouse & Hanson, 1972; Feldman, 1966; Zajonc, 1968). Evidence for this negativity dominance was found in in an offline context by Weinberger and Dillon (1980), who showed that negative product ratings had a bigger influence on purchase intentions than positive ratings. Brand evaluations also have been shown to be more influenced by negative WOM than positive WOM (e.g. Mizerski, 1982; Richins, 1983). Park & Lee (2006) found that this negativity effect appears in an online setting as well, as they showed that eWOM’s impact is bigger for negative eWOM than for positive eWOM. This finding was replicated by Chevalier & Mayzlin (2006), who found that the impact of negative online reviews on book sales was bigger than the impact of positive reviews.
Construal level theory has the potential to offer new insights regarding the perception of negative eWOM narratives and their influence on consumer behavior and attitudes. Negative eWOM narratives can vary on all of the dimensions of psychological distance described above. They can concern events that happened a long time ago or recently, in a place far from the reader or close to him/her, and the sender of the message can be perceived as similar or dissimilar to the reader of the message. If the dimensions of psychological
distance are all interrelated, greater temporal, spatial or social distance of an eWOM narrative should also lead to greater hypothetical distance. That means that if someone reads a message about a particular event that happened a long time ago, happened far away from that person or was written by a person who is not perceived to be similar to him, this person should also feel that this event is less likely to occur to him. Lower perceived likelihood of occurrence of an event should have an effect on the influence of a message describing such an event.
Regarding social distance, it would seem logical to assume that greater interpersonal similarity leads to greater influence of a reviewer, but there is such a great number of
dimensions on which people can differ from each other that general statements about the effect of similarity on persuasiveness are difficult to make (O’Keefe, 2002). One line of research suggests that similarity does not have a direct influence on persuasiveness, but that this relationship is mediated by the receiver’s liking for the communicator (O’Keefe, 2002). Here, a different line of reasoning is adopted: as the communicator differs more from the reader, the message will be thought about more abstractly, and an improbable chance of occurrence matches these abstract thoughts better than a likely chance of occurrence. If the reader of a negative eWOM narrative does not feel that something similar can happen to him, the effect of the narrative should be small, whereas when the reader perceives the narrative as something that could happen to him just the same, the effect of the narrative should be bigger. The same argument holds for the other types of distance: temporal and spatial distance lead to high-level construals using abstract thoughts, and an unlikely chance of occurrence best matches these abstract thoughts. Events that are perceived as unlikely to occur to the reader should have a smaller impact than events that are perceived as likely to occur.
Based on the literature reviewed above, it is hypothesized that the impact of temporal, spatial and social distance in a negative eWOM narrative on brand attitudes and purchase intentions is mediated by the perceived likelihood of occurrence. This is because as temporal,
spatial or social distance increases, readers of a negative eWOM narrative should feel less likely to experience something similar, which should reduce the impact the narrative has on their evaluations and purchase intentions.
H1: Temporal, spatial and social distance in a negative eWOM narrative are positively related to consumer attitudes and purchase intentions, and this relationship is mediated by the perceived probability of occurrence of the described event.
Method
Participants
247 people (58% female, Mage = 26.4 years, age range: 17-80 years) voluntarily
participated in an online experiment to test the hypotheses. 89 participants were excluded beforehand because they did not finish the experiment. Participants were recruited via social media or email in a non-random way. 63.2 percent were students and 27.5% were employed for wages. 78 percent were Dutch, with the rest coming from 23 other countries from different continents. 42.5 percent listed a Bachelor’s degree as their highest degree completed, 13% listed a Master’s degree and 15.8% listed grammar school as their highest level of schooling completed. The participants were randomly assigned to one of the eight conditions of the experiment.
Materials and procedure
The experiment had a 2 (temporal distance: low/high) by 2 (spatial distance: low/high) by 2 (social distance: low/high) between-subjects factorial design. Participants were asked to imagine being assigned to a six-month work project, and after arriving at the work location, feeling a sharp pain in their lower back. Because of this, they wanted to see a physiotherapist
to find out what is wrong with them. They were then shown a web page of a (real)
physiotherapist with some basic information about the practice (see appendix A). After seeing the website, they were shown a negative online review of a customer who had been to the same physiotherapist and was not happy with the way he was treated (see appendix B for the full review). After this, participants were asked to answer the questions about the dependent variables, some demographic questions and finally some control questions. After filling in these questions, participants were thanked and asked to tell friends to participate.
Temporal distance was manipulated by varying how long ago the reviewer said he went to the physiotherapist (last week in the low temporal distance condition and last year in the high temporal distance condition). The date the review was posted was also varied (March 2nd, 2014 in the low temporal distance condition and March 2nd, 2013 in the high temporal distance condition). Spatial distance was manipulated by varying the location of the work project (Amsterdam in the low spatial distance condition and Sydney, Australia in the high spatial distance condition). The choice for Sydney, Australia was made because Sydney is, like Amsterdam, a modern, first world city. By choosing a modern city, people would not make any assumptions about the quality of physiotherapists located there that could bias the results. The location of the physiotherapist was repeated several times on the webpage shown to the participants. To manipulate social distance, a method similar to Kim, Zhang & Li (2008) was used. In a similar study on online word of mouth, they manipulated social distance by varying the name of the reviewer, i.e. by giving him a common Chinese name (as the study was done in China) or an uncommon, non-Chinese name. In this study, the reviewer in the low social distance condition was a 26-year old called Mark de Vries (a common Dutch name, as most of the participants were expected to be Dutch) and the reviewer in the high social distance condition is a 46-year old person called Bogdan Kozlowski. As most of the
was in reality 26.4 years), adding their respective ages meant to increase the social distance between the two reviewers.
Dependent and control variables
Hypothetical distance was measured by asking how likely participants thought it was that they would have a similar experience with the particular physiotherapist. Participants answered on three 7-point bipolar scales anchored by the adjectives “unlikely-likely”, “improbable-probable” and “impossible-possible” (Yi, 1990). Consumer evaluations were measured by three 7-point scales: “bad-good”, “unfavorable-favorable”, “unsatisfactory-satisfactory” (Hensel, 1992). Purchase intentions were measured by three 7-point scales: “unlikely-likely”, “improbable-probable” and “impossible-possible” (Yi, 1990). Cronbach’s alphas for the 3 hypothetical distance, 3 consumer evaluations and 3 purchase intentions items were .899, .859 and .952, respectively. No items were removed, as this would not have lead to an increase in reliability.
Several variables were added as control variables. These were importance of online reviews when buying a new product/service to participants, whether they had ever been to Amsterdam/Sydney before, and whether they had ever been to a physiotherapist (and if yes, how often they went).
Results
Manipulation check
To check whether the social distance manipulation was effective, a pre-test was conducted (N = 40). Participants were asked to what extent they felt a 26-year old person called Mark de Vries or a 46-year old person called Bogdan Kozlowski was similar to themselves, a typical in-group member and psychologically close. All three questions were
An independent sample t-test showed that Mark de Vries (M = 3.98, SD = 1,75) was perceived as socially closer than Bogdan Kozlowski (M = 2.3, SD = 1,03), t(39) = 3,736, p < .001. This supported the effectiveness of the social distance manipulation.
Correlations
Means, standard deviations, reliabilities and correlations for the three independent variables and hypothetical distance, consumer evaluations and purchase intentions can be found in table 1. Temporal distance was not positively related to hypothetical distance, consumer evaluations and purchase intentions, r = .036, p >.05; r = .09, p > .05; r = .064, p > .05, respectively. Spatial distance was also not significantly related to either of hypothetical distance, consumer evaluations and purchase intentions, r = -.077, p > .05; r = .09, p > .05; r = .051, p > .05, respectively. Social distance was also not significantly related to hypothetical distance, consumer evaluations and purchase intentions, r = .02, p > .05; r = -.015, p > .05; r = .051, p > .05, respectively. Consumer evaluations were strongly related to purchase
intentions, r = .704, p < .01. Hypothetical distance was not significantly related to both consumer evaluations and purchase intentions, r = .031, p > .05; r = -.025, p > .05.
Hypothesis testing
To test whether temporal, spatial and social distance are positively related to hypothetical distance, independent sample t-tests were conducted with temporal distance, spatial distance and social distance as grouping variables and hypothetical distance as the dependent variable. There was no significant difference between the low (M = 4.56, SD = 1.22) and high (M = 4.65, SD = 1.32) temporal distance conditions, t(245) = -.568, p > .05. There was also no significant difference between the low (M = 4.70, SD = 1.23) and high (M = 4.50, SD = 1.30) spatial distance conditions, t(245) = 1.213, p > .05. Finally, there was no
Table 1
Means, Standard Deviations and Correlations of the Variables (Cronbach’s Alphas on the diagonal)
Variables Mean S.D. 1 2 3 4 5 6 1. Temporal Distance 2. Spatial Distance 3. Social Distance 4. Hypothetical Distance 4.60 1.27 .036 -.077 .02 (.899) 5. Consumer Evaluations 3.42 1.08 .011 .09 -.015 .031 (.859) 6. Purchase Intentions 3.50 1.42 .064 .063 .051 -.025 .704** (.952) Note. N=247. * p<.05, ** p<.01 .09 .03 .02 (.02)
Figure 1. Regression coefficients for the relationship between temporal distance and consumer evaluations with perceived probability Temporal Distance
Perceived Probability of
Occurrence
significant difference between the low (M = 4.58, SD = 1.34) and high (M = 4.63, SD = 1.19) social distance conditions, t(245) = -.321, p > .05. This means that people thought it was equally likely for them to have a similar experience with this physiotherapist, regardless of whether the event described in the eWOM narrative happened a long time ago or recently, happened in a faraway place or close by, or was experienced by someone perceived as socially distant or socially close.
The hypothesized relationship between temporal, spatial and social distance and consumer evaluations was not mediated by hypothetical distance. Figure 1 illustrates the model with temporal distance as the independent variable and perceived probability as the mediator with consumer evaluations as the dependent variable. The same simple mediation model applies to the other variables. Temporal distance was not significantly related to hypothetical distance (β = .09, p = .57) and consumer evaluations (β = .02, p = .87). The relationship between hypothetical distance (the mediator) and consumer evaluations was also not significant (β = .03, p = .67). The indirect effect was (.09)(.03) = .0027. As Hayes (2013) recommends, the significance of this indirect effect was measured using bootstrap procedures. Unstandardized indirect effects were computed for 10,000 bootstrapped samples, and if 0 does not fall in the 95% confidence interval, it would indicate significance. The 95%
confidence interval ranged from -.01 to .05, so it can be concluded that the indirect effect was not significant. Another way of testing the significance of the indirect effect is a Sobel test. A Sobel test checks whether the reduction in the effect of the independent variable is statistically significant when the mediator is added to the model (Baron & Kenny, 1986). A Sobel test was conducted and resulted in a p-value of .83, confirming the lack of significance.
Spatial distance was not significantly related to hypothetical distance (β = -.20, p = .23) and consumer evaluations (β = .20, p = .15). The relationship between hypothetical distance and consumer evaluations was also not significant (β = .03, p = .55). The indirect
effect was (-.20)(.03) = -.006. Unstandardized indirect effects were computed for 10,000 bootstrapped samples, and the 95% confidence interval ranged from -.07 to .02, so it can be concluded that the indirect effect was not significant. A Sobel test confirmed this finding, resulting in a p-value of .66.
The relationship between social distance and hypothetical distance (β = .05, p = .75) and consumer evaluations (β = -.03, p = .80) was also not significant. Hypothetical distance and consumer evaluations were not significantly related (β = .03, p = .62). The indirect effect was (.05)(.03) = .0015. Unstandardized indirect effects were computed for 10,000
bootstrapped samples, and the 95% confidence interval ranged from -.02 to .05, so it can be concluded that the indirect effect was not significant because 0 falls in this interval. This lack of significance was confirmed by a Sobel test, which resulted in a p-value of .89.
The relationship between temporal, spatial and social distance and purchase intentions was also not mediated by hypothetical distance. Temporal distance was not significantly related to hypothetical distance (β = .09, p = .57) and purchase intentions (β = .18, p = .31). The relationship between hypothetical distance and purchase intentions was also not
significant (β = -.03, p = .67). The indirect effect was (.09)(-.03) = -.0027. Unstandardized indirect effects were computed for 10,000 bootstrapped samples, and the 95% confidence interval ranged from -.06 to .02, so it can be concluded that the indirect effect was not significant. A Sobel test confirmed this result, resulting in a p-value of .84.
Spatial distance was not significantly related to hypothetical distance (β = -.20, p = .23) and purchase intentions (β = .17, p = .34). Hypothetical distance and purchase intentions were not significantly related (β = -.02, p = .75). The indirect effect was (-.20)(-.02) = .004. Unstandardized indirect effects were computed for 10,000 bootstrapped samples, and the 95% confidence interval ranged from -.03 to .07, so it can be concluded that the indirect effect was
not significant. A Sobel test was undertaken to confirm the lack of significance of the indirect effect, and it supported the earlier findings with a p-value of .81.
Social distance was not significantly related to hypothetical distance (β = .05, p = .75) and purchase intentions (β = .15, p = .42). The relationship between hypothetical distance and purchase intentions was also not significant (β = -.03, p = .68). The indirect effect was (.05) (-.03) = -.0015. Unstandardized indirect effects were computed for 10,000 bootstrapped samples, and the 95% confidence interval ranged from -.05 to .02, so it can be concluded that the indirect effect was not significant. This was confirmed by a Sobel test, which resulted in a p-value of .90.
It can be concluded that the hypothesized impact of temporal, spatial and social distance on consumer evaluations and purchase intentions was not mediated by hypothetical distance. In other words, hypothetical distance does not function as an ‘intermediary variable’ in the (hypothesized) relationship between temporal, spatial and social distance and consumer evaluations and purchase intentions.
Adding the different control variables (importance of eWOM, having been to Sydney/Amsterdam, having been to a physiotherapist) both separately and together had no significant effect on the relationship between temporal, social and spatial distance and
consumer evaluations and purchase intentions. In conclusion, no support was found for either hypothesis, even when the different control variables were added.
Discussion
Do people think events that happened a long time ago, in a faraway place or events that are experienced by people not similar to themselves think those events are less likely to occur to themselves compared to events that happened recently, nearby or if they were experienced by
similar others? And if people think something they read online is less likely to occur to them, does this information have less impact on their behavior as consumers than when they think they could experience something similar? These are the questions that were attempted to answer in this study.
Previous research found that people think unlikely events occur at greater temporal and spatial distance (Wakslak, 2012) and that people think unlikely events are more likely to be experienced by other people rather than themselves (Chambers, Windschitl & Suls, 2003; Kruger & Burrus, 2004). Because evidence had been found that the links between temporal, spatial and social distance work both ways (Stephan, Liberman & Trope, 2010), and the study by Wakslak (2012) already showed one direction of influence between temporal, spatial and hypothetical distance, it was hypothesized that temporal, spatial and social distance are positively related to hypothetical distance. One area of research where the relationship between temporal, spatial and social distance and hypothetical distance can be of interest is electronic word of mouth. This is because events described in eWOM narratives can have occurred a long time ago or recently, in a faraway place or nearby, or can be written by someone perceived as similar or dissimilar to the reader. If temporal, spatial and social distance lead to hypothetical distance, it means people think eWOM narratives describing events that occurred a long time ago, in a faraway place and were experienced by dissimilar others think such an event is less likely to occur to themselves. An event that is perceived as unlikely should have less impact on someone’s consumer behavior (in this study consumer evaluations and purchase intentions) than an event that is perceived as likely to occur. Therefore, it was hypothesized was that the effect of temporal, spatial and social distance on consumer evaluations and purchase intentions was mediated by hypothetical distance.
No support was found for the hypothesis that temporal, spatial and social distance are positively related to hypothetical distance. The lack of support for the relationship between
social distance and hypothetical distance can possibly be explained by the fact that the studies that found a relationship in the opposite direction operationalized social distance in another way. Chambers, Windschitl & Suls (2003) and Kruger & Burrus (2004) operationalized social distance as the distinction between the self and others, while in this study interpersonal
(dis)similarity was used. The latter case is always concerned with someone other than the self (the reader of the message in this instance), while this is not so in the first case. Although the manipulation check supported its effectiveness, it is possible that people who read the review paid little attention to the name of the author and what kind of person this was. Although Wakslak (2012) concluded that even when social distance was operationalized in another way it should still be positively related to hypothetical distance, the two different
operationalizations differ to such an extent that it could be cause of a lack of significance. As for temporal and spatial distance, it is possible that the manipulations in the experiment were not effective. It was mentioned just once that the reviewer went to the physiotherapist in question last week or last year, and while changing the date the message was posted meant to increase temporal distance as well, this was not very prominent. Although the choice for Sydney was made because it is a modern city and people would not make assumptions about the quality of service, this manipulation could have been counter-effective. Most people (92%) in the sample had never been to Sydney, but because it is a modern city that nearly everyone knows something about and English is its native language, people might have felt that Sydney was closer to them than it actually was. This could have resulted in a lower perceived spatial distance and could possibly explain the lack of
significance of the relationship between spatial distance and hypothetical distance.
The mediation hypothesis was also not supported. A possible explanation for this is that people might have had strong feelings about the physiotherapist even before they read the review. Some people might have thought the physiotherapist was too expensive or elitist
because of the claims made on the web page. If people already had strong feelings about the physiotherapist before reading the review, the review would probably have less effect than when these prior feelings were not present. Because everybody was exposed to the same information on the website and people in all conditions thus formed these prior feelings, the possibly reduced impact of the review for everyone could explain why there were no
significant results. Another factor that could have played a role is the fact that the participants were shown only one review. It is thinkable that people tend to base their decisions on
multiple reviews, as this would allow people to compare the different messages and get a better picture of what the product/service is really like. Some people might have thought that the rude treatment the reviewer received at the physiotherapist was his own fault and had nothing to do with the physiotherapist. If more reviews were shown where people all said they were being treated rudely, it would probably have more impact on their evaluations and purchase intentions. Importance of eWOM can be ruled out as an explanation, as most of the sample (more than 82%) claimed online reviews were either somewhat important or very important to them, and adding this variable as a control did not change the results.
As most consumers use online reviews in their consumption decision process, understanding the factors that surround it is highly important. If a company shows only old reviews on its website and, although they are positive, people think the events described are unlikely to occur to them because they happened a long time ago, a company does not profit as much from the eWOM compared to when they update their website with newer reviews. This is just one example of why understanding the relationship between construal level theory and online word of mouth is important for companies operating online. Though the
hypotheses were not supported in this study, it is too early to claim that they must be false. Multiple studies in the past established the reverse directions of influence, and if some of the limitations of this experiment are addressed, the results could be different.
Future research could test these hypotheses in a setting other than physiotherapy. Although people tend to use reviews more for services than for products (Murray, 1991), a different setting where participants are not faced with making a quick decision could provide different results. Another study could manipulate social distance in a different way, such as by using either formal or colloquial language (Brown, 1987). This way, the amount of social distance is more equal between participants in the same condition and depends less on the subjective perceptions they have about the reviewer. Future research could also test the hypotheses using more than one review (a method used by Lee, Park & Han, 2008). For example, an experiment could show ten reviews, of which eight are negative and two positive. Because most people use more than one review to form their evaluations and purchase
intentions, this would be an interesting change to the experiment. Finally, an experiment where people are asked both before and after the review what their evaluations and intentions are could also prove to result in different findings, as then it will be clear whether the results were caused by the website or the review.
As more and more companies find themselves online, it is important to understand all the aspects of online business. Most consumers use some kind of online reviews when buying new products/services, so understanding the different factors influencing the effectiveness of online word of mouth is both interesting for researchers and crucial for companies that want to succeed in today’s digital marketplace.
References
Addis, D. R., & Schacter, D. L. (2008). Constructive episodic simulation: Temporal distance and detail of past and future events modulate hippocampal engagement. Hippocampus, 18(2), 227-237.
Arndt, J. (1967). Role of product-related conversations in the diffusion of a new product. Journal of Marketing
Research, , 291-295.
Bar-Anan, Y., Liberman, N., Trope, Y., & Algom, D. (2007). Automatic processing of psychological distance: Evidence from a stroop task. Journal of Experimental Psychology: General, 136(4), 610.
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social
Brucks, M. (1985). The effects of product class knowledge on information search behavior. Journal of
Consumer Research, , 1-16.
Bruner, G. C., Hensel, P. J., & James, K. E. (2001). Marketing scales handbook.
Buckner, R. L., & Carroll, D. C. (2007). Self-projection and the brain. Trends in Cognitive Sciences, 11(2), 49-57.
Buttle, F. A. (1998). Word of mouth: Understanding and managing referral marketing. Journal of Strategic
Marketing, 6(3), 241-254.
Chambers, J. R., Windschitl, P. D., & Suls, J. (2003). Egocentrism, event frequency, and comparative optimism: When what happens frequently is “more likely to happen to me”. Personality and Social Psychology
Bulletin, 29(11), 1343-1356.
Chevalier, J. A., & Mayzlin, D. (2003). The Effect of Word of Mouth on Sales: Online Book Reviews, Chu, S., & Kim, Y. (2011). Determinants of consumer engagement in electronic word-of-mouth (eWOM) in
social networking sites. International Journal of Advertising, 30(1), 47-75.
East, R., Hammond, K., & Lomax, W. (2008). Measuring the impact of positive and negative word of mouth on brand purchase probability. International Journal of Research in Marketing, 25(3), 215-224.
Eyal, T., Sagristano, M. D., Trope, Y., Liberman, N., & Chaiken, S. (2009). When values matter: Expressing values in behavioral intentions for the near vs. distant future. Journal of Experimental Social Psychology,
45(1), 35-43.
Feldman, S. (1966). Motivational aspects of attitudinal elements and their place in cognitive interaction.
Cognitive Consistency: Motivational Antecedents and Behavioral Consequences, , 75-108.
Fujita, K., Henderson, M. D., Eng, J., Trope, Y., & Liberman, N. (2006). Spatial distance and mental construal of social events. Psychological Science, 17(4), 278-282.
Hayes, A. F. (2008). Introduction to mediation, moderation, and conditional process analysis: A
regression-based approach Guilford Press.
Hennig-Thurau, T., Gwinner, K. P., Walsh, G., & Gremler, D. D. (2004). Electronic word-of-mouth via
consumer-opinion platforms: What motivates consumers to articulate themselves on the internet? Journal
of Interactive Marketing, 18(1), 38-52.
Kahneman, D. (2011). Thinking, fast and slow Macmillan.
Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica:
Journal of the Econometric Society, , 263-291.
Kanouse, D. E., & Hanson Jr, L. R. (1987). Negativity in evaluations. Preparation of this Paper Grew Out of a
Workshop on Attribution Theory Held at University of California, Los Angeles, Aug 1969.
Kruger, J., & Burrus, J. (2004). Egocentrism and focalism in unrealistic optimism (and pessimism). Journal of
Experimental Social Psychology, 40(3), 332-340.
Lee, J., Park, D., & Han, I. (2008). The effect of negative online consumer reviews on product attitude: An information processing view. Electronic Commerce Research and Applications, 7(3), 341-352. Liberman, N., Sagristano, M. D., & Trope, Y. (2002). The effect of temporal distance on level of mental
construal. Journal of Experimental Social Psychology, 38(6), 523-534.
Liberman, N., & Trope, Y. (1998). The role of feasibility and desirability considerations in near and distant future decisions: A test of temporal construal theory. Journal of Personality and Social Psychology,
75(1), 5.
Liberman, N., & Trope, Y. (2008). The psychology of transcending the here and now. Science, 322(5905), 1201-1205.
Liviatan, I., Trope, Y., & Liberman, N. (2008). Interpersonal similarity as a social distance dimension: Implications for perception of others’ actions. Journal of Experimental Social Psychology, 44(5), 1256-1269.
Mizerski, R. W. (1982). An attribution explanation of the disproportionate influence of unfavorable information.
Journal of Consumer Research, , 301-310.
Murray, K. B. (1991). A test of services marketing theory: Consumer information acquisition activities. The
Journal of Marketing, , 10-25.
O'keefe, D. J. (2002). Persuasion: Theory and research SAGE Publications, Incorporated.
Park, C., & Lee, T. M. (2009). Information direction, website reputation and eWOM effect: A moderating role of product type. Journal of Business Research, 62(1), 61-67.
Richins, M. L. (1983). Negative word-of-mouth by dissatisfied consumers: A pilot study. The Journal of
Marketing, , 68-78.
Richins, M. L., & Root-Shaffer, T. (1988). The role of evolvement and opinion leadership in consumer word-of-mouth: An implicit model made explicit. Advances in Consumer Research, 15(1), 32-36.
Schacter, D. L., Addis, D. R., & Buckner, R. L. (2007). Remembering the past to imagine the future: The prospective brain. Nature Reviews Neuroscience, 8(9), 657-661.
the web. Journal of Interactive Marketing, 21(4), 76-94.
Skowronski, J. J., & Carlston, D. E. (1989). Negativity and extremity biases in impression formation: A review of explanations. Psychological Bulletin, 105(1), 131.
Stephan, E., Liberman, N., & Trope, Y. (2010). Politeness and psychological distance: A construal level perspective. Journal of Personality and Social Psychology, 98(2), 268.
Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Journal of Experimental Psychology,
18(6), 643.
Sweeney, J. C., Soutar, G. N., & Mazzarol, T. (2008). Factors influencing word of mouth effectiveness: Receiver perspectives. European Journal of Marketing, 42(3/4), 344-364.
Thompson, N. (2003). More companies pay heed to their'word of mouse'reputation. New York Times, 23 Trope, Y., & Liberman, N. (2010). Construal-level theory of psychological distance. Psychological Review,
117(2), 440.
Wakslak, C. J. (2012). The where and when of likely and unlikely events. Organizational Behavior and Human
Decision Processes, 117(1), 150-157.
Wakslak, C. J., Trope, Y., Liberman, N., & Alony, R. (2006). Seeing the forest when entry is unlikely: Probability and the mental representation of events. Journal of Experimental Psychology: General,
135(4), 641.
Weinberger, M. G., & Dillon, W. R. (1980). The effects of unfavorable product rating information. Advances in
Consumer Research, 7(1), 528-532.
Wilson, T. D., Wheatley, T., Meyers, J. M., Gilbert, D. T., & Axsom, D. (2000). Focalism: A source of durability bias in affective forecasting. Journal of Personality and Social Psychology, 78(5), 821. Yi, Y. (1990). Cognitive and affective priming effects of the context for print advertisements. Journal of
Advertising, 19(2), 40-48.
Zajonc, R. B. (1968). Attitudinal effects of mere exposure. Journal of Personality and Social Psychology,
Appendix A
The website of the physiotherapist shown to participants. People in the low spatial distance condition saw Amsterdam, while people in the high spatial distance condition saw Sydney.
Physiotherapy
If you are in pain, injured or require help with any muscle, joint
or nerve problem, the good news is we can help. The first thing to
know about us is that we are different – genuinely different – and
in a really good way.
We combine some of Sydney’s/Amsterdam’s best, most experienced,
hands-on physiotherapists with excellent massage therapists and use them
in combination to get the best result for you. This means you will have
100% one-to-one attention from a great practitioner for the whole of your
treatment session – every time!
Our treatment includes:
• A clear diagnosis
• An understanding of why the problem occurred
• A plan on how to fix the problem
• Hands-on physiotherapy
• Additional input from our expert massage therapists
• Access to our strong network of sports physicians, GPs and orthopaedic
surgeons
Say goodbye to poor service
At our Sydney/Amsterdam clinic, you are not left on a machine while your
therapist sees someone else. You don’t get short, rushed consultations
because your therapist has overbooked. And you are never be seen by a
Appendix B
The review shown to participants. Participants in the low social distance condition saw Mark de Vries (26) and participants in the low temporal distance condition saw last week and March 2nd, 2014 and the high distance conditions were shown the other option. The review layout was based on the Dutch website www.zorgkaartnederland.nl where you can find reviews about doctors, dentists etc. and the review was based on reviews about
physiotherapists made on that website.
Score: 5,5/10
Posted on March 2nd, 2014/2013, by Mark de Vries (26)/Bogdan Kozlowski (46)
Competent but rude and arrogant staff
I went here last week/last year, and it was a bad experience overall. The counter assistant was rude to me and acted like she did not want to be there. The physiotherapist was competent, but also very arrogant and derogatory. He seemed disinterested when I talked to him, like I was an unwanted patient. He paid little attention to my suggestions about the problem. Luckily my problem was solved in the end, but I probably would not go here again.
