Can post-purchase dissonance be lowered with the use of post-purchase
deals?
A study in the hotel industry
Mark Hoekstra
University of Groningen Faculty of Economics and Business
Master Thesis Marketing Management & Marketing Intelligence June 2018 Dirk Huizingastraat 4 9713GM Groningen (06) 33742515 m.hoekstra.29@student.rug.nl Student number: S2515652 1st supervisor: Lisette de Vries
Abstract
Nowadays, the internet is the biggest channel for hotel room distribution. Booking intermediaries like Booking.com signal free-cancellation to lower the risks that a customer might feel during the booking process. This paper aims to analyse whether the number of alternatives on a hotel booking intermediary website is affecting the dissonance level. Furthermore, it is researched if this dissonance level could be lowered by the use of a post-purchase deal. The main relationship that is tested is whether post-post-purchase dissonance is a predictor of cancellation. Results of the study suggest that the previously stated relations are not confirmed. The only relationship found is that of emotion on post-purchase dissonance. It is found that negative emotions are predicting dissonance a more positive way (higher negative emotions lead to higher post-purchase dissonance) than positive emotions predict dissonance in a negative way (higher positive emotions lead to lower post-purchase dissonance). Other conclusions, limitations of the paper and future research are discussed in the main body of this paper.
Keywords: post-purchase deal, post-purchase dissonance, cancellation probability, hotel
Table of content
1. Introduction ... 4
2. Theoretical background ... 7
2.1 Post-purchase dissonance ... 7
2.2 Post-purchase deal ... 7
2.3 Type of post-purchase deal ... 8
2.4 Price-sensitivity ... 10
2.5 Cancellation probability ... 11
2.6 Control variables... 11
2.7 Conceptual model ... 12
3. Methodology ... 13
3.1 Data collection method ... 13
3.2 Population and sampling method... 13
3.3 Operational definitions ... 14 3.4 Experimental design ... 15 3.5 Analytical tools ... 15 4. Results ... 17 4.1 Descriptive statistics ... 17 4.2 Correlation analysis ... 17 4.3 Factor analysis ... 18 4.4 Reliability analysis ... 19 4.5 Assumption testing ... 20 4.6 Results ... 21 5. Discussion... 25 5.1 Hypotheses ... 25
5.2 Theoretical and practical implications ... 27
5.3 Limitations and future research ... 28
6. Conclusion... 30
7. References ... 31
8. Appendix ... 38
8.1 Appendix A ... 38
1. Introduction
It is only a decade ago that the Internet became the biggest channel for hotel room distribution. These innovations within the digital space showed huge opportunities for hotel businesses. However, this also increased the competition drastically, which resulted in the rise of booking intermediaries which compare all different hotels on one single website (Thakran & Verma, 2013). It is especially on these websites that there are many alternative hotels to choose from. Previous studies showed that when a customer has to make a choice among many alternatives, where each alternative shows different advantages and disadvantages in comparison to the others, different levels of post-purchase dissonance are likely to occur (Milliman & Decker, 1990). Post-purchase dissonance is a negative phase within the total customer experience, which is not only cognitive in nature, it also has an emotional component (Sweeney et al., 2000) which should be taken into account. These emotions can, for example, vary from ‘in doubt’ to ‘regret’, ‘indifference’ or ‘distress’. Marketers are struggling to measure this cognitive dissonance when customers are moving from the pre-purchase to the post-purchase phase (Hasan & Nasreen, 2014).
To lower the perceived risks that a customer might experience during the booking phase, booking intermediaries signal their ‘free cancellation’ policy (Chen et al. 2011). Because of these lenient cancellation policies, the probability exist that these intermediaries attract certain ‘deal-seeking’ travellers. These deal-seekers anticipate for price changes over time to search for the best deal. There are known cases of customers making multiple bookings, where in the end only one booking is used (Talluri & Van Ryzin, 2004). The two main characteristics when booking a hotel are price and availability, which are not stable over time and still change after the booking is made. So, for a price-sensitive customer, the booking process is not necessarily completed when a reservation has been made (Schwartz, 2006).
These stricter cancellation policies however, have the prejudicial effect of reducing the total number of bookings (Smith et al., 2015).
To prevent customers from cancelling their hotel reservations, hotels need to ‘lock-in’ a customer directly after the booking is made. When hotels are able to lock-in a customer directly, the probability that a customer will be in doubt about their decision and search for alternatives, should be lower. But how is a hotel able to lock-in a customer directly after the booking is made? Farrell and Klemperer (2007) state that switching costs have the potential to bind customers to sellers if products or services are incompatible. So, when a hotel is able to increase the switching costs, without stricter cancellation policies, and make their hotel less compatible, customers should be more locked-in.
When taking the persuasion techniques from Cialdini (2007) into account, the reciprocity principle is a technique that has the potential to work. It is based on the idea that when you give something away, people feel obligated to give something back. This, for example, happens in stores where people get a cup of tea at the entrance. Within this research the reciprocity principle can be used to increase the perceived switching costs. A new construct is made, the so called ‘post-purchase deal’, which is given to a customer directly after the booking is made. This post-purchase deal can be used in the hotel, like a free ticket to the sauna, but it can also be used outside the hotel, like a free city tour. The main focus of the post-purchase deal is to lower the post-purchase dissonance, which should diminish the need to search for alternatives.
However, not every customer is a so called ‘deal-seeking’ customer. In a study by Dickinger and Mazanec (2008), it was found that recommendations of friends and online reviews were the most important factors that influence the online hotel booking process. For that reason, a real booking intermediary is used, and price-sensitivity is taken into account. If it is possible to identify whether someone is price-sensitive, the best kind of post-purchase deal could be used and offered to the right group.
They found that the intention of a customer to cancel a reservation decreased when the temporal and monetary sunk costs increased. Within the conclusion of the paper, the authors state that future research should focus on other mediators too (e.g., perceived fairness, importance of travel etc.). Within this research, post-purchase dissonance will be used as a mediator, because it is in line with previous findings where a high number of alternatives explained higher dissonance levels. This research contributes to existing literature by measuring if this post-purchase dissonance is affecting the cancellation probability.
Managerial implications of this research are that marketers and organizations are better able to lower the post-purchase dissonance by making use of post-purchase deals. Consequently, by lowering the post-purchase dissonance, the probability of cancelation should be diminished.
The research question of this paper will be:
Can post-purchase dissonance be lowered with the use of post-purchase deals?
The sub questions of this paper will be:
- Which post-purchase deals work best?
- Will a lower post-purchase dissonance result in a lower cancellation probability?
2. Theoretical background
2.1 Post-purchase dissonanceAfter a purchase is made and the consumer completed the process, this consumer will evaluate the purchase where possibly some negative attributes come to mind. Moreover, the consumer might have the feeling that he did not need the product in the first place. These negative and positive attributes can create dissonance in the customer’s mind (Sweeney et al., 2000). “Any conflicting thought in the human mind which arises from the discrepancy between what the consumer believes in and any information which negates that, is referred to as cognitive dissonance” (Hasan & Nasreen, 2014, p. 65). Hasan and Nasreen (2014) also found that the more alternatives available, where a consumer should make a decision among a high number of positive and negative attributes, the greater the dissonance experienced. A well-known study is that of Iyengar and Lepper (2000), where the different amount of gourmet jams and chocolates in supermarkets is studied. They found that in the case of extensive choice, participants reported being more dissatisfied and having more regret about the choices they had made in comparison to those participants who had a limited choice of products.
2.2 Post-purchase deal
Deals are defined as a reward that might generate consumer response as they improve the price-value relationship that the product or service offers to consumers (Schultz et al., 1998). Within this research, these deals are post-purchase deals because they are only presented directly after the purchase. The post-purchase phase is defined as the phase after completing a transaction, it is the phase where consumers confirm their expectation through a post-purchase evaluation process and form their satisfaction level (Kim et al., 2009). The post-purchase deal will be defined as a reward that might generate immediate consumer response which is given directly after completing a transaction.
Hypothesis - Post-purchase deal on post-purchase dissonance
As already stated by Hasan and Nasreen (2014), if an individual makes a decision from many alternatives, varying levels of post-decision dissonance will emerge. Because particularly the hotel booking intermediaries offer a large number of alternatives, it is expected that after the decision for a certain hotel is made, post-purchase dissonance will occur.
found that subjects who received post-transaction reassurances reported lower perceived dissonance scores in comparison to the control group, who did not get any communication after the purchase was made. However, if the store called the person instead of mailing them, to tell them it was the right choice, the dissonance became higher because it could have aroused suspicions. A more recent research from Tseng (2017) was performed with online tourists. The purpose of the study was to enhance regretful tourists by providing post-purchase information (sellers’ ratings) to reduce the post-purchase cognitive dissonance. This effect was confirmed. Furthermore, Festinger (1957) argues that individuals, motivated by the negative affective state of dissonance to engage in discrepancy reduction, can reduce dissonance by changing the original cognitions, adding/subtracting cognitions (e.g., new attitudes, behaviours, or beliefs), or adjusting the importance of the cognitions. More recent work used these outcomes and found that “reduction of dissonance can be achieved by psychologically increasing the attractiveness of the chosen alternative and/or by mentally decreasing the attractiveness of the rejected alternative” (Bawa & Kansal, 2008, p. 42). When a post-purchase deal is given, the attractiveness of the chosen alternative increases. Consequently, it is expected that:
H1: Receiving a post-purchase deal directly after the hotel booking is made, will lower the post-purchase dissonance.
2.3 Type of post-purchase deal
organization and should be used at another company. Yi and Jeon (2003) classified this into two categories: direct vs. indirect rewards. The direct rewards were related to the actual product or service, where the indirect rewards are irrelevant to the product or service.
Next to these different types of deals, the promotion depth is importance. Promotion depth is defined as the value of the promotion, it is the percentage of discount compared to the regular price. It is found that the choice for a certain promotion is higher when high depth promotions are used in percentage-off rates rather than in cents-off promotion (DelVecchio et al., 2007).
Hypothesis – The differences between the type of post-purchase deal
Post-purchase deals can differ in their promotion depth. Alford and Biswas (2002) found that a higher discount resulted in higher value perceptions (4.51 vs. 3.72), lower search intention (5.29 vs. 5.51) and a higher buying intention (3.66 vs. 2.69). However, other research shows that this price discount can be too high. Barat and Paswan (2005) found that a high discount signals that the previous price of the product was too high, resulting in the fact that consumers were less likely to buy the product. Because of the research by Barat and Paswan (2005), it is expected that:
H2: The positive effect of the post-purchase deal on post-purchase dissonance will be greater when a 50% discount deal is used compared to a 100% discount deal.
When using a reference price, the customer knows the value of the deal that is given. Strong evidence exists that even though consumers tend to be sceptical of reference prices given by the company, consumer perceptions of value and savings are positively influenced by such comparative prices even when the prices are undue (Alford & Biswas, 2002). Anchoring prices are a well-known concept within the travel industry. It basically refers to the tendency people have to adjust their price estimates based on the initial reference price (Book et al., 2016).
not. Because within this research a customer is focussed on the booking of a hotel and not on booking the whole holiday, we expect the customer to be more interested in the hotel deal in comparison to the alliance deal. Mainly because at this moment in time, the customer is comparing hotel rooms. When a deal is offered by the hotel, the switching costs of the hotel become higher (Farrell and Klemperer, 2007). Because of this, it is expected that:
H3: The positive effect of the post-purchase deal on post-purchase dissonance will be greater when an ‘own-deal’ is used in comparison to an ‘alliance deal’.
2.4 Price-sensitivity
It should be taken into account that customers differ, some customers care more about prices and deals than others. Highly price-sensitive customers will seek lower prices compared to less price-sensitive customers. Price sensitivity is a variable that measures individual differences and is defined as the way in which buyers react to prices and to price changes (Goldsmith et al., 2005). Specifically, how customers feel about the price for an offer. Income is a very relevant variable in this case, because of the income effect: “other things being equal, consumers are willing to spend more if they have more to spend” (Goldsmith et al., 2010, p. 328). Within this research, the price-sensitivity of a respondent will be measured to find out whether this has an effect on the deal-proneness and the according post-purchase dissonance.
Hypothesis – The moderating and direct effect of price-sensitivity
Basic microeconomic theory suggests that “lower income consumers will obtain greater relative economic benefit from coupon usage compared to their counterparts, which results in the higher likelihood of redeeming coupons” (Noble et al., 2017, p. 554). However, this straightforward idea was not supported by academic marketing research, where either a positive effect of income was shown, or an insignificant effect. However, Noble et al. (2017) researched coupon redemption related to income categories and price sensitivity and found a significant negative effect. Higher price-sensitive consumers, redeemed their coupons at a higher rate compared to the lower price-sensitive categories. Because of these recent findings it is expected that:
2.5 Cancellation probability
Shlifer and Vardi (1975) researched the cancellation probability on airline tickets and found that the cancellation probability varies from 0.3 to 0.8, being higher for reservations made a few months in advance, and 0.3 only two weeks before departure. They define the cancellation probability as the probability that a reservation will be cancelled before take-off. Within this research the cancellation probability is defined as the probability that a reservation will be cancelled before the check-in date of the hotel.
Hypothesis - Post-purchase dissonance on cancellation probability
To find out whether a change in post-purchase dissonance is directly related to a behavioural change, several different researches are taken into account. Donnely and Ivancevich (1970) performed a research among automobile customers who already bought a car, which was not yet delivered. They did a test where one group was positively reinforced after the buying, versus the control group which was not. The salesman reinforced the group by calling them about the features of the car, and that they had made a good decision. The results show that the back-out rates were significantly lower for the experimental group, the group that was positively reinforced. More recent research from Lee (2005) focussed on product returns in combination with purchase dissonance. It was found that consumers who showed forms of post-purchase dissonance were significantly using product returns more instead of a dissonance reduction strategy. More research on this topic, especially in the hotel industry, is needed. Because of the previous findings it is expected that:
H5: A lower post-purchase dissonance will result in a lower cancellation probability.
2.6 Control variables
Bonniot-H1, H2, H3 H4
H5
Figure 1 – Conceptual model
Cabanac et al. (2012) found that cognitive dissonance is also related to specific emotions. For that reason, emotion will be used as a control variable. According to Cabanac (2002), emotions can be considered as a mental state with hedonic content. The control variables of this research will be income, personality and emotion.
2.7 Conceptual model
All previous hypotheses taken into consideration, leads to the following conceptual model (figure 1).
Price-sensitivity
Post-purchase deal Post-purchase
3. Methodology
3.1 Data collection methodWithin this study, the relationship between different variables was tested using an online survey. This type of method made it possible to get much data in a relatively short time (Blumberg, Cooper and Schindler, 2014). The survey was posted on Facebook & LinkedIn. Furthermore, the survey was also directly sent to respondents with a personalized message to increase the response rate. Once completed the survey, the respondent was asked to share the survey with some relatives or friends, increasing the total number of respondents.
3.2 Population and sampling method
The population consists of everyone above the age of 18, since this is the minimum age required to book a hotel in the Netherlands. Because everyone on Facebook & LinkedIn also has to be 18 years or older, this is a perfect tool to reach respondents. It does not matter if the respondent ever booked a hotel before, because also those that never booked a hotel before, are interesting within this research. Hoshino-Browne et al., (2005) found that cultural differences also have an effect on the perceived dissonant levels. To control for this, only Dutch respondents are used.
To come up with a sample, the convenience sampling method and the snowball sampling method were used. The convenience sampling method is less reliable, but it is a cheap and quick way to reach a lot of respondents (Blumberg, Cooper and Schindler, 2014). To tackle the reliability issue, three different channels to reach respondents were used. Friends and relatives were targeted directly or by using Facebook. Work and student related respondents were targeted directly, through Facebook or LinkedIn. Own business-related respondents were targeted directly when they checked-in at our own accommodation (which is a group accommodation located in Sneek, the Netherlands), or those that were targeted with a Facebook post on our company Facebook.
3.3 Operational definitions Post-purchase dissonance
Koller and Salzberger (2007) developed a scale to measure post-purchase dissonance. They evaluated the scale in a research based on the purchase of a package deal. This scale was also used by Tanford and Montgomery (2005), who focussed on the relationship between cognitive dissonance and travel purchase decisions. Because these previous studies are in line with this research, these items will be used. One item from O'Neill and Palmer (2004) was added, which was the item ‘I should have spent more time considering my choice of hotel’, to control for the possibility that people take less time than they would do in real life. The items can be found in Appendix A.
Price sensitivity
The construct price sensitivity is measured using the scale of Goldsmith et al. (2003). The six items regarding the price sensitivity were also measured on a 7 point Likert-scale. According to Croasmun and Ostrom (2011), a 7-point Likert-scale is the most reliable scale to use. The scale within this research will be: strongly disagree, disagree, somewhat disagree, neither disagree nor agree, somewhat agree, agree, strongly agree. The items can be found in Appendix A.
Types of post-purchase deal
To test the different types of post-purchase deals, two attributes were used. The first attribute is the type of deal; it can be a hotel-specific deal (a ticket for the sauna) or a 3rd party deal (a city tour). The other attribute is price; the deal can be free (a free city tour) or it can be a discount (a 50% discount on a city tour). The other option is that no deal is given. Altogether, there were five different options, which means that there were five different scenarios for the survey. The different deals can be found in Appendix A.
Control variables
experiment was performed where participants needed to describe their cognitive dissonance emotions. The respondents came up with 78 different emotions. We chose to use the 16 most common emotions from the paper that were also possible to merge with this research. Emotions that were so uncommon to feel within this type of research, were deleted. For example, emotions like ‘solidarity’ and ‘anxiety’ were deleted. These kinds of emotions would only confuse the respondents, which may have a negative result on the outcomes of the survey. The control variables that were used can be found in Appendix A.
3.4 Experimental design
Within the first part of the survey, the respondent is asked to choose a hotel from the largest booking intermediary website Booking.com. The town that needs to be visited is Paris, the number of people is 2 and the length of stay is 3 nights. This is the most realistic setup where the number of alternatives is high. Furthermore, the aim of the hotel visit is for vacation purposes, not a business trip. All hotels have a free-cancellation policy.
The second part of the survey is different among the five variants. Here, the respondents will get one of the alternatives: no deal, a hotel free deal, a hotel discount deal, a 3rd party free deal or a 3rd party discount deal. This means that a between-subject design is used, where a respondent will get only one out of five scenarios.
The third part is again the same for all respondents and measures the post-purchase dissonance. Do the respondents show feelings of regret, doubt or disappointment? Also the emotions were taken into account and the respondent had to rate to what extend a certain emotion is felt. Hereafter, the probability of cancellation will be measured on a binary scale, asking the question whether the respondent might cancel the booking, when a better alternative arises in the upcoming period.
In the fourth part of the survey the demographic information about the participant is obtained, which is mainly used for the control variables personality and income. The final part of the study is focussing on the variable price-sensitivity. Different questions are used to measure whether a person is price-sensitive or not.
3.5 Analytical tools
Figure 2 – Hayes model 7
whether the items were all measuring the construct that it should measure. When different items were combined into a construct, the Cronbach’s Alpha was calculated to check the internal consistency (Tavakol and Dennick, 2011).
4. Results
4.1 Descriptive statistics225 respondents started the survey, but only 153 respondents finished it. The average age of the respondents was 34 with a minimum age of 18 and a maximum age of 74. These 153 respondents will be used for further analysis. This total consisted of 71 male and 82 female respondents.
According to the IV and DV, also some descriptive statistics can be shown. The IV ‘purchase deal’ consisted of five different possibilities; the four different types of post-purchase deal or no deal. 32 respondents did not get a deal and 121 respondents got one out of the four possible deals. The mediation variable post-purchase dissonance showed a mean of 2.6 out of 7 which means that the feeling of overall dissonance is low. Our DV ‘cancellation probability’ was measured using a binary scale. It shows that 61.4% of the respondents were willing to cancel the reservation when a better alternative became available during the upcoming period. On the other hand, 38.6% was not willing to cancel the reservation.
Data cleaning
Within the dataset, there were 16 respondents who reported that they did not want to give their income. For these respondents the mean is imputed so this data can still be used. Furthermore, 1 respondent was 17 years old and was therefore deleted from the dataset, since this person is legally not allowed to book a hotel in the Netherlands.
4.2 Correlation analysis
First, a correlation analysis is done to check if the items of a construct are correlating. When this is the case, they are probably measuring the same construct. To check if this assumption is true, a factor analysis and a reliability analysis are performed subsequently.
4.3 Factor analysis
To check whether the different items are measuring the same construct, a factor analysis is done. This is done by using a principle component analysis. The moderator variable ‘price sensitivity’ was measured using 6 items on a 7-point Likert scale. All factor loadings should be higher than .5, which is the case for four out of the six items regarding price-sensitivity. The first item ‘In general, the price or cost of buying products is important to me’ and the last item ‘A really great product is worth paying a lot of money for’ do not have factor loadings of at least .5 and were therefore deleted. Bartlett’s test of sphericity was significant (p=0,000), which means that all correlations within the correlation matrix are significant. The Kaiser Meyer Olkin (KMO) is used to check whether enough items are predicted by a factor. The outcome of the test was .876. According to Firend and Abadi (2014), a value above .7 means that the sample is adequate. Also, the communalities are checked, these communalities measure the percentage of variance in a given variable explained by all the extracted factors. They should be higher than .4, which is the case for the price sensitivity items.
The mediator variable ‘post-purchase dissonance’ was measured using 8 items on a 7-point Likert scale. Again, a factor analysis is used where the same outcomes as the price sensitivity construct for the Bartlett’s test of sphericity and the KMO test are reported. The only factor loading that is not above .5 is the item ‘I am not quite sure about my decision’, this item is deleted. Also, the communalities of the items are all above .4, which is good.
The emotions were all separately measured on a 7-point Likert scale. According to Cabanac (2002), the emotions can be separated into positive and negative constructs. Therefore, these emotions were also included in the factor analysis. The factor loadings below .5 were expectation, minor interest, satisfaction, boredom, displeasure and indifference. These items were deleted. The results of the factor analysis are visible in table 1. It is clear that the factor loadings in row 2 are positive emotions and the factor loadings in row 3 are the negative emotions.
Number of factors
‘post-purchase dissonance’, ‘positive emotions and ‘negative emotions’ are used. The outcome of the factor analysis is shown in Appendix B.
1 2 3 4 DISSONANCE2 .741 DISSONANCE3 .757 DISSONANCE4 .780 DISSONANCE5 .819 DISSONANCE6 .807 DISSONANCE7 .811 DISSONANCE8 .769 DISSONANCE9 .670 PS2 .659 PS3 .766 PS4 .704 PS5 .767 JOY .707 PLEASURE .725 ENTHUSIASM .717 EXCITEMENT .768 LUCK .818 CURIOSITY .692 NERVOUSNESS .752 IMPATIENCE .720 CONCERN .729 STRESS .633
Table 1 – Rotated component matrix
Extraction method: principle component analysis with Varimax rotation
4.4 Reliability analysis
positive emotions construct (α = .848) and the negative emotions construct (α = .793) show an appropriate internal consistency.
4.5 Assumption testing
Within this research, the dependent variable is the cancellation probability. It is measured on a binary scale, as the respondent is asked whether he/she would cancel the reservation when a better alternative became available in the upcoming period. The answer that could be given was yes, or no. For this type of dependent variable, we can use a binary logit model. “One of their most common applications is to estimate the effect of a particular variable of interest on a binary outcome when potentially confounding variables are controlled” (Karlson et al., 2012, p. 287). The logit model is used instead of the probit model because of the mathematical convenience. The probabilities are easier to calculate and the estimates of the parameters are also easier to interpret. Because of the common use of survey data, a logit model is preferred in a lot of studies (Horowitz & Savin, 2001). The major assumptions for the binary logit model are according to Tabachnick & Fidell (2012):
Discrete outcome
The outcome must be discrete, which means that the dependent variable should be zero or one. This is the case within this research, where the cancellation probability is measured on a yes/no scale.
Outliers
There should be no outliers. This can be checked by changing the outcomes into z-scores where they have a mean of zero and a standard deviation of one. This way, outliers can be detected and the values that are below -3.29 or above 3.29 should be deleted. This is done for all variables that are used in the model. No outliers were detected as the minimum value is -2.48 and the maximum value is 3.23.
Multicollinearity
Linearity
It is assumed that the dependent variable should have a linear relationship with the predictors. This assumption is checked by looking whether the interaction term of the predictor and its log transformation are significant (Hosmer & Lemeshow, 1989). This is tested by running a binary logistic regression by using the IV * Ln(IV) and the DV. This regression shows no significant interaction effect, which means that the linearity assumption is met.
4.6 Results
The model that should be tested is P(Y=1|PPDeal, PPDissonance, PS), where the probability of cancellation (Y=1) is explained by the variables post-purchase deal, post-purchase dissonance and price-sensitivity. To control for the effects, the variables personality, positive emotions, negative emotions and income are used as control variables. For this type of model, 2 different tests are used. First, the effect on dissonance is explained by using the moderation mediation analysis developed by Hayes. This Hayes model 7 is used to check whether there is an effect on the post-purchase dissonance when a deal is given compared to when no deal is given. The outcomes of the test are reported in table 2. It shows that the main effect of the post-purchase deal is not significantly explaining dissonance. The Hayes model also tests relationships between variables by taking different levels of dissonance into account. This test did not show any significant relationships. The same applies to the moderator variable price sensitivity, which is also not significantly explaining dissonance. The only variables that are significantly explaining dissonance are the control variables positive emotions and negative emotions. The interpretation is done using the odds ratio. This interpretation is done in the paragraph ‘odds ratio’.
cancellation probability. The outcomes of this test are also visible in table 2. The relationship between dissonance and cancellation probability is not significant.
The new variable DissonanceHat can then be used in the Hayes model to check if there is a moderated mediation effect. First of all, de moderator price-sensitivity does not show a significant relationship. Consequently, there is no moderated mediation. This has also been checked in the index of moderated mediation, where it is shown that there is no significant moderator effect. Next to that, the direct effect from X (deal_yes_no) on Y (cancellation) is not significant (p = .0611). Also, the interaction effect (p = .1505) and the effect from the mediator dissonance on cancellation is not significant (p = .7393). Because of these insignificant outcomes, there is no direct effect and no mediation effect within the model.
To check whether the different types of deals resulted in different dissonance levels, a between-groups-ANOVA test was conducted. This ANOVA test showed no significant difference between the different types of deals and the corresponding dissonance levels, F(4,147) = .922, p = .453.
Consequent
M (Dissonance) Y (Cancellation)
Antecedent Coeff. SE p Coeff. SE p
Deal_yes_no -.9378 .7781 .2301 .7600 .4058 .0611
Price sensitivity -.2236 .1637 .1741 - - -
Dissonance - - - -.1257 .3778 .7393
Int. Deal x Price sensitivity .2665 .1843 .1505 - - -
Income -.0183 .0392 .6409 - - - Positive emotions -.4593 .0978 .0000*** - - - Negative emotions .5305 .0682 .0000*** - - - Confident shopper -.0590 .0653 .3677 - - - Negative shopper -.0768 .0887 .3883 Constant 4.7000 1.0652 .0000*** .1968 1.0301 .8485 𝑅"= .4781 F(9, 140) = 14.25, p = 0.0000*** McFadden R2 = .0174 Cox-Snell R2 = .0230 Nagelkerke R2 = .0312 P-value = .1709 Table 2 – Results Hayes Model 7 and 2SLS model
Model performance
Before interpreting the model significance and R2, the residual plots need to be checked. If these plots give unwanted residual patters, it means that the outcomes are biased. The residual plots show us that they are normally distributed and the test for Skewniss and Kurtosis are low. Because of these good outcomes, we can interpret the model fit.
The F test of the Hayes model 7 is significant. Therefore, the model explains the mediator variable dissonance significantly better than the null-model where only the constant is included. Because of the significant model, the R2 can also be interpreted. The R2 informs us of the relative predictive capability of the model. It indicates the gains of the predictive accuracy of our independent variables, compared to when these variables are not known (Lewis-Beck & Skalaban, 1990). In the Hayes model, the predictive capability of dissonance is almost 48% (R2 = .4781), which means that 48% of the variance of post-purchase dissonance is explained by these variables.
The second part of the Hayes model explaining the independent variable cancellation is not significant (p = .1709), meaning that no conclusions can be drawn that this model is better explaining the cancellation probability than the null-model with only the intercept.
Odds ratio
The interpretation of the coefficients is done by using the odds ratio. The odds ratio can be calculated by using the Exp (Beta). The odds ratio shows the odds that an outcome will occur given a particular exposure, compared to the odds of the outcome occurring in the absence of that exposure (Scotia, 2010). It is calculated as the probability that the event will occur divided by the probability that the event will not occur. If the odds ratio is greater than 1 it indicates that as the predictor increases, the probability that the outcome will occur also increases. A value less than 1 indicates that as the predictor increases, the odds of the outcome will decrease (Field, 2009).
In other words, a dissonant feeling will increase more when negative emotions are perceived compared to the decrease of dissonance when positive emotions are perceived.
5. Discussion
Within this chapter, the results of the Hayes model 7 and the 2SLS model will be discussed. Starting with the Hayes model 7, which is used to explain the mediator variable dissonance. Next, the relationship between the mediator variable dissonance and the dependent variable cancellation, which is tested using the 2SLS model, will be discussed. After the discussion of the hypotheses, the control variables, the theoretical and practical implications will be handled. Afterwards, the limitations of the study and future research will be discussed.
5.1 Hypotheses
The first hypothesis was: Receiving a post-purchase deal directly after the hotel booking is
made, will lower the post-purchase dissonance. This hypothesis has not been confirmed. There
can be two possible explanations for this. The first possibility is that people are not sensitive for these kinds of deals. The second explanation: people who did not receive a deal also don’t perceive high dissonant feelings. Dissonance was measured on a 7 point Likert scale where the mean value was 2.6 for the total sample size, meaning that the second explanation is most likely. This means that the conclusion of Hasan and Nasreen (2014) and Milliman & Decker (1990), regarding the relationship between a high level of alternatives and a high dissonant level, is not supported in this case.
Chernev (2003) also gives a possible explanation. In this study, it was found that people with clear preferences of a product or service preferred choosing from a larger assortment. Satisfaction increased for those people when the number of options increased. This is referred to as the opposite of choice overload. The same result was found by Mogilner et al. (2008), who found only a negative relationship between satisfaction and choice for those people that were less familiar with the choice domain. Because the online domain is nowadays the biggest channel for hotel room distribution, the probability exist that people know their preferences and use the corresponding filters on the hotel intermediary booking website when making their decision. Because of these findings, a possible explanation for a low dissonance level is the prior preferences that people have when booking a hotel.
Our second and third hypotheses were The positive effect of the post-purchase deal on
post-purchase dissonance will be greater when a 50% discount deal is used compared to a 100% discount deal and The positive effect of the post-purchase deal on post-purchase dissonance will be greater when an ‘own-deal’ is used in comparison to an ‘alliance deal’.
with a post-purchase deal and dissonance is also not supported. When the first hypothesis would have been significant, the probability of a significant difference between the type of deal would also be higher. However, this is not the case.
The fourth hypothesis is the hypothesis regarding the moderator effect: For
price-sensitive customers, the positive effect of the post-purchase deal on post-purchase dissonance will be greater than for non-price-sensitive customers. This hypothesis is also not supported.
This means that there is no significant difference between the dissonance levels for price-sensitive and non-price-price-sensitive customers when a post-purchase deal is given. An explanation for this could be that price-sensitive customers are already looking for the cheapest hotel and when they found a good hotel, they already have a low dissonant level. When the deal is received directly after their booking, the dissonant level may not decrease any further. Another explanation which is also supported in the literature, is that price-sensitivity is not a predictor of deal proneness. This finding is supported by Blattberg et al. (2010), where they conclude that deal proneness is highly influenced by household resources, time related variables and employment status. They also found that deal proneness was effected by income, however when adjusted for household resources, the effect was diminished. This could be an explanation why the hypothesis is not confirmed.
The final hypothesis was: A lower post-purchase dissonance will result in a lower
cancellation probability. The 2SLS model was used which showed that dissonance was not a
predictor of cancellation probability. The mean of the perceived post-purchase dissonance was low and there was no significant difference between the dissonance level of people that received a deal and those that did not received a deal. On the other hand, the mean cancellation probability was 61.4%, which is quite high. Because of that, we can assume that people don’t need to have a dissonant feeling when they cancel a reservation. Maybe some people are just satisfied with their reservation right now, but when a better alternative pops up, they might still cancel the booking. This can be referred to as ‘delayed dissonance’ where a person does not feel any forms of post-purchase dissonance right after the booking of a hotel, but it might feel some dissonance when he/she sees new alternatives.
Control variables
The control variables personality, income and emotion were also taken into account. The variables personality and income had no significant effect on post-purchase dissonance.
Sweeney et al. (2000) also developed a scale for measuring dissonance, where some emotions were used. Examples of the used scale are ‘I felt hollow, angry, frustrated or in pain’. However, these emotions are far from matching post-purchase dissonance in this research. For that reason, we used the emotions from the research of Bonniot-Cabanac et al. (2012). The emotions joy, pleasure, enthusiasm, excitement, luck and curiosity were bundled into the construct ‘positive emotions’. This variable shows a negative significant effect on dissonance. On the other hand, the variables nervousness, impatience, concern and stress form the construct ‘negative emotions’. This variable is a positive and significant predictor of dissonance. The analysis showed that negative emotions are more influencing dissonance in a positive way than positive emotions influence dissonance in a negative way. This means that when a business wants to lower the perceived dissonance of a customer, more impact can be made when focussing on decreasing the negative emotions instead of increasing the positive ones.
5.2 Theoretical and practical implications
The theoretical contribution to the literature on post-purchase dissonance and hotel room cancellation is threefold. This study shows that (1) researchers cannot just assume that dissonant feelings are felt when the number of alternatives is high, (2) post-purchase dissonance is not a predictor of cancellation, (3) the emotional aspect should be taken into account when measuring dissonance, where it is found that the negative emotions are more influencing dissonance in a positive direction than the positive emotions influence dissonance in a negative direction.
First, Hasan and Nasreen (2014) state that dissonant feeling is felt when a customer has to make a choice among many alternatives. This research shows this is not the case within the hotel industry. There were more than 1600 alternative hotels to choose from and there was no significant difference in dissonance levels when a post-purchase deal was given versus when no deal was given. Therefore, the assumption that many alternatives lead to high dissonance levels is not always correct.
Third, the emotional aspect is already taken into account when measuring dissonance in, for example, the scale developed by Sweeney et al. (2000). However, these emotions are very severe and do not fit within this study. When making a simple choice like booking a hotel room, these emotions are not quite common. Therefore, the emotions from the research of Bonniot-Cabanac et al. (2012) were used, which were more in line with this study. Some of these emotions were indeed influencing post-purchase dissonance. Furthermore, it is found that negative emotions are more influencing dissonance than the positive emotions.
The practical contribution of the paper is mainly based on the emotional aspect. As stated before, negative emotions are more influencing dissonance in a positive direction than positive emotions are influencing dissonance in a negative direction. Organizations and marketers can use this knowledge in lowering post-purchase dissonance by first focusing on lowering the perceived negative emotions. Should a company be able to lower the perceived nervousness or stress, the perceived post-purchase dissonance could be decreased. On the other hand, the positive emotions show less effect. Using this knowledge, organizations should be able to lower the post-purchase dissonance in focusing on the improvement of the negative emotions.
5.3 Limitations and future research
This study has some limitations that should be taken into account for future research. First of all, the study uses a between-design instead of a within-design, which means respondents get only one out of the four possible post-purchase deals, or they do not get a deal. Using the within-design, the respondent will see all possible deals and will evaluate each deal compared to the other. Charness, Gneezy and Kuhn (2012) refer to this as the between-within paradigm where respondents use several criteria in evaluating the post-purchase deal. These criteria may however differ regarding the between or within design.
Another limitation is that the study is fictional. The respondent was asked to choose a hotel as he would also do in a real case scenario, but the possibility the respondent will do it in another way always exists. Therefore, the perceived dissonance after choosing the hotel might differ compared to a real scenario.
deal, but it is good to check if there is a difference between deals. Which deals are working better and for which kind of groups? For example, demographic differences in the sensitivity to different kind of deals could exist. Maybe, older people like the city trip more where younger people are more interested in the sauna tickets. Deals could also be specific for different cities, where people visiting Paris could be more interested in a city trip and people visiting Oslo are more interested in a free sauna ticket.
The second limitation could be tackled by using real data from a booking intermediary website. This way, the dissonance levels reported by the respondents right after the booking of a hotel could be more realistic.
Future research should also focus on the variables that explain dissonance. This study shows that the positive and negative emotions are predictors of dissonance. There are, however, more predictors which need to be indicated so marketers can find ways to lower this dissonance by changing these predictors.
Furthermore, future research should focus on the variables that predict cancellation. This research shows that dissonance is not a predictor of cancellation, but high cancellation rates are still a huge problem for hotel businesses. These predictors should be indicated in order to know the reasons for people cancelling their hotel reservations. These variables can then help in lowering the cancellation probability.
6. Conclusion
Although it is assumed that aspects of post-purchase dissonance are felt when a customer has to make a choice among many alternatives, this is not confirmed within the hotel industry. Moreover, the reciprocity principle of Cialdini (2007), where a post-purchase deal is given to lower the perceived post-purchase dissonance, does not show a significant effect. This means that the research question of this paper ‘Can post-purchase dissonance be lowered with the use of post-purchase deals?’ is answered with no. Post-purchase dissonance is not lowered when a post-purchase deal is given. Furthermore, this research shows no significant difference between the type of deal.
Finally, the dependent variable cancellation is not significantly explained by post-purchase dissonance. This means that even if people do not feel any dissonant feelings after booking a hotel, they might still cancel the reservation when a better alternative arises. They might be satisfied with their reservation right now and do not feel any forms of doubt, but still cancel the reservation for any reason. This means that also our sub question ‘Will a lower post-purchase dissonance result in a lower cancellation probability?’ is answered with no. Post-purchase dissonance does not have an effect on the cancellation probability within the hotel industry.
7. References
Achrol, R. S., & Kotler, P. (1999). Marketing in the network economy. The Journal of
Marketing, 146-163.
Alford, B. L., & Biswas, A. (2002). The effects of discount level, price consciousness and sale proneness on consumers' price perception and behavioral intention. Journal of Business
research, 55(9), 775-783.
Antonio, N., de Almeida, A., & Nunes, L. (2017). Predicting hotel booking cancellations to decrease uncertainty and increase revenue. Tourism & Management Studies, 13(2).
Arnold, R. R. (1984). Improving the Effectiveness of Marketing Communications. Proceedings, Southwest Marketing Association, Houston, Texas, 61-64.
Barat, S., & Paswan, A. K. (2005). Do higher face-value coupons cost more than they are worth in increased sales?. Journal of Product & Brand Management, 14(6), 379-386.
Bawa, A., & Kansal, P. (2008). Cognitive dissonance and the marketing of services: some issues. Journal of Services Research, 8(2), 31.
Blattberg, R., Buesing, T., Peacock, P., & Sen, S. (2010). Identifying the deal prone segment. In Perspectives On Promotion And Database Marketing: The Collected Works of Robert C
Blattberg (pp. 79-87).
Bonniot-Cabanac, M. C., Cabanac, M., Fontanari, J. F., & Perlovsky, L. I. (2012). Instrumentalizing cognitive dissonance emotions. Psychology, 3(12), 1018.
Book, L. A., Tanford, S., & Chen, Y. S. (2016). Understanding the impact of negative and positive traveler reviews: Social influence and price anchoring effects. Journal of Travel
Research, 55(8), 993-1007.
Chen, C. C., Schwartz, Z., & Vargas, P. (2011). The search for the best deal: How hotel cancellation policies affect the search and booking decisions of deal-seeking customers. International Journal of Hospitality Management, 30(1), 129-135.
Chernev, A. (2003). When more is less and less is more: The role of ideal point availability and assortment in consumer choice. Journal of consumer Research, 30(2), 170-183.
Cialdini, R. B (2007). Influence: The psychology of persuasion (pp. 173-174). New York: Collins.
Cox, D. R., and E. J. Snell. 1989. The Analysis of Binary Data, 2nd ed. London: Chapman and Hall.
Croasmun, J. T., & Ostrom, L. (2011). Using Likert-type scales in the social sciences. Journal
of Adult Education, 40(1), 19.
Cummings, W. H., & Venkatesan, M. (1976). Cognitive dissonance and consumer behavior: A review of the evidence. Journal of Marketing Research, 303-308.
DelVecchio, D., Krishnan, H. S., & Smith, D. C. (2007). Cents or percent? The effects of promotion framing on price expectations and choice. Journal of Marketing, 71(3), 158-170.
Dickinger, A., & Mazanec, J. (2008). Consumers’ preferred criteria for hotel online booking. Information and Communication Technologies in Tourism 2008, 244-254.
Donnely Jr, J. H., & Ivancevich, J. M. (1970). Post-purchase reinforcement and back-out behavior. Journal of Marketing Research (pre-1986), 7(3), 399.
Fabrigar, L. R., Wegener, D. T., MacCallum, R. C., & Strahan, E. J. (1999). Evaluating the use of exploratory factor analysis in psychological research. Psychological Methods, 4(3), 272.
Festinger, L. 1957, A Theory of Cognitive Dissonance, Stanford University Press, Stanford, CA.
Field, A. (2009). Discovering statistics using SPSS. Sage publications.
Firend, A. R., & Abadi, M. F. (2014). Impact of service quality, trust and perceived value on customer loyalty in Malaysia services industries. Procedia-Social and Behavioral
Sciences, 164, 298-304.
George, D., & Mallery, P. (2003). SPSS for Windows step by step: A simple guide and reference.
11.0 update (4th ed.). Boston: Allyn & Bacon.
Gino, F. (2008). Do we listen to advice just because we paid for it? The impact of advice cost on its use. Organizational behavior and human decision processes, 107(2), 234-245.
Goldsmith, R. E., Flynn, L. R., & Goldsmith, E. B. (2003). Innovative consumers and market mavens. Journal of Marketing theory and practice, 11(4), 54-65.
Goldsmith, R. E., Kim, D., Flynn, L. R., & Kim, W. M. (2005). Price sensitivity and innovativeness for fashion among Korean consumers. The Journal of social
psychology, 145(5), 501-508.
Goldsmith, R. E., Flynn, L. R., & Kim, D. (2010). Status consumption and price sensitivity. Journal of Marketing Theory and Practice, 18(4), 323-338.
Goldsmith, R. E., & Newell, S. J. (1997). Innovativeness and price sensitivity: managerial, theoretical and methodological issues. Journal of Product & Brand Management, 6(3), 163-174.
Hasan, U., & Nasreen, R. (2014). The Empirical Study of Relationship between Post Purchase Dissonance and Consumer Behaviour. Journal of Marketing Management, 2(2), 65-77.
Hoshino-Browne, E., Zanna, A. S., Spencer, S. J., Zanna, M. P., Kitayama, S., & Lackenbauer, S. (2005). On the cultural guises of cognitive dissonance: The case of Easterners and Westerners. Journal of personality and social psychology, 89(3), 294.
Hunt, S. D. (1970). Post-transaction communications and dissonance reduction. The Journal of
Marketing, 46-51.
Iyengar, S. S., & Lepper, M. R. (2000). When choice is demotivating: Can one desire too much of a good thing? Journal of personality and social psychology, 79(6), 995.
Karlson, K. B., Holm, A., & Breen, R. (2012). Comparing regression coefficients between same-sample nested models using logit and probit: A new method. Sociological
Methodology, 42(1), 286-313.
Kassarjian, H.H. & Cohen, J.B. 1965, ‘Cognitive dissonance and consumer behaviour’, California Management Review, vol. 8, no. 1, pp. 55–64.
Kim, D. J., Ferrin, D. L., & Rao, H. R. (2009). Trust and satisfaction, two stepping stones for successful e-commerce relationships: A longitudinal exploration. Information systems
research, 20(2), 237-257.
Koller, M., & Salzberger, T. (2007). Cognitive dissonance as a relevant construct throughout the decision-making and consumption process-an empirical investigation related to a package tour. Journal of customer behaviour, 6(3), 217-227.
Lewis-Beck, M. S., & Skalaban, A. (1990). The R-squared: Some straight talk. Political
Analysis, 2, 153-171.
Lichtenstein, D. R., Burton, S., & Netemeyer, R. G. (1997). An examination of deal proneness across sales promotion types: a consumer segmentation perspective. Journal of
Lee, D. H. (2015). An alternative explanation of consumer product returns from the postpurchase dissonance and ecological marketing perspectives. Psychology &
Marketing, 32(1), 49-64.
McLachlan, G. J. (1987). On bootstrapping the likelihood ratio test stastistic for the number of components in a normal mixture. Applied statistics, 318-324.
Mehrotra, R., & Ruttley, J. (2006). Revenue management(second ed.). Washington, DC, USA: American Hotel & Lodging Association (AHLA).
Milliman, R. E., & Decker, P. J. (1990). The use of post-purchase communication to reduce dissonance and improve direct marketing effectiveness. The Journal of Business
Communication (1973), 27(2), 159-170.
Mogilner, C., Rudnick, T., & Iyengar, S. S. (2008). The mere categorization effect: How the presence of categories increases choosers' perceptions of assortment variety and outcome satisfaction. Journal of Consumer Research, 35(2), 202-215.
Nagelkerke, N. J. D. 1991. A note on the general definition of the coefficient of determination. Biometrika, 78:3, 691-692.
Noble, S. M., Lee, K. B., Zaretzki, R., & Autry, C. (2017). Coupon clipping by impoverished consumers: Linking demographics, basket size, and coupon redemption rates. International
Journal of Research in Marketing, 34(2), 553-571.
Noone, B. M., & Lee, C. H. (2010). Hotel overbooking: The effect of overcompensation on customers’ reactions to denied service. Journal of Hospitality & Tourism Research, 35(3), 334– 357.
Osborne, J. W. (2015). What is rotating in exploratory factor analysis. Practical Assessment,
Research & Evaluation, 20(2), 1-7.
Palmer, A., & Bejou, D. (1995). Tourism destination marketing alliances. Annals of tourism
Park, J. Y., & Jang, S. S. (2014). Sunk costs and travel cancellation: Focusing on temporal cost. Tourism Management, 40, 425-435.
Rammstedt, B., & John, O. P. (2007). Measuring personality in one minute or less: A 10-item short version of the Big Five Inventory in English and German. Journal of research in
Personality, 41(1), 203-212.
Reibstein, D. J., & Traver, P. A. (1982). Factors affecting coupon redemption rates. Journal of
Marketing, 46, 102–113.
Schultz, W. (1998). Predictive reward signal of dopamine neurons. Journal of
neurophysiology, 80(1), 1-27.
Schwartz, Z. (2006). Advanced booking and revenue management: Room rates and the consumers’ strategic zones. International Journal of Hospitality Management, 25(3), 447-462.
Shlifer, E., & Vardi, Y. (1975). An airline overbooking policy. Transportation Science, 9(2), 101-114.
Shoemaker, R. W., & Tibrewala, V. (1985). Relating coupon redemption rates to past purchasing of the brand. Journal of Advertising Research, 25, 40–47.
Smith, S. J., Parsa, H. G., Bujisic, M., & van der Rest, J.-P. (2015). Hotel cancelation policies, distributive and procedural fairness, and consumer patronage: A study of the lodging industry. Journal of Travel & Tourism Marketing, 32(7), 886–906.
Stockman, L. H. (1957). The influence of consumer deals on urban household purchases of butter, margarine, vegetable shortening, and salad and cooking oils in metropolitan Chicago (Doctoral dissertation, Northwestern University).
Sweeney, J. C., Hausknecht, D., & Soutar, G. N. (2000). Cognitive dissonance after purchase: A multidimensional scale. Psychology and Marketing, 17(5), 369-385.
Talluri, K. T., & Van Ryzin, G. (2004). The theory and practice of revenue management. Boston, MA, USA: Kluwer Academic Publishers
Tanford, S., & Montgomery, R. (2015). The effects of social influence and cognitive dissonance on travel purchase decisions. Journal of Travel Research, 54(5), 596-610.
Tavakol, M., & Dennick, R. (2011). Making sense of Cronbach's alpha. International journal
of medical education, 2, 53.
Thakran, K., & Verma, R. (2013). The emergence of hybrid online distribution channels in travel, tourism and hospitality. Cornell Hospitality Quarterly, 54(3), 240-247.
Tseng, A. (2017). Why do online tourists need sellers' ratings? Exploration of the factors affecting regretful tourist e-satisfaction. Tourism Management, 59, 413-424.
Webster Jr, F. E. (1965). The" deal-prone" consumer. Journal of Marketing Research, 186-189.
Xie, J., & Gerstner, E. (2007). Service escape: Profiting from customer cancellations. Marketing Science, 26(1), 18-30.
Yi, Y., & Jeon, H. (2003). Effects of loyalty programs on value perception, program loyalty, and brand loyalty. Journal of the academy of marketing science, 31(3), 229-240.
Zakhary, A., Atiya, A. F., El-Shishiny, H., & Gayar, N. E. (2011). Forecasting hotel arrivals and occupancy using Monte Carlo simulation. Journal of Revenue and Pricing
8. Appendix
8.1 Appendix ASurvey
The respondent had to choose a hotel when using the following link:
Next, the following items needed to be answered on a 7 point Likert-scale. Post-purchase dissonance (mediator)
1. I am not quite sure about my decision
2. I am annoyed that I will miss things from other hotels now 3. When thinking of the decision, I feel uncomfortable 4. Perhaps I should have spent the money on something else 5. I do not know whether this was the right choice
6. Now, after the booking, I feel uneasy
7. I am wondering if I have made the right choice 8. I would like to undo my decision
9. I should have spent more time considering my choice of hotel
Emotions (mediator)
Do you feel any of these emotions after booking the hotel?
10. Joy 11. Pleasure 12. Satisfaction 13. Enthusiasm 14. Excitement 15. Luck 16. Indifference 17. Minor interest 18. Boredom 19. Expectation 20. Concern 21. Stress 22. Nervousness 23. Impatience 24. Displeasure 25. Curiosity Cancellation (DV)
26. Is there a probability that you will cancel the reservation when a better alternative will come up in the upcoming period?
Income (CV)
Personality (CV)
I see myself as a…
1. Confident shopper 2. Positive shopper 3. Inspiring shopper 4. Confused shopper 5. Nervous shopper 6. Negative shopper 7. Guilt ridden shopper Price sensitivity (moderator)
8. In general, the price or cost of buying products is important to me
9. I know that a new kind of product is likely to be more expensive than older ones, but that doesn’t matter to me
10. I don’t mind paying more to try out a new product
11. I am less willing to buy a product if I think that it will be high in price 12. I don’t mind spending a lot of money to buy a product
8.2 Appendix B
Construct Eigenvalue % of variance Cumulative %
1 7.655 34.796 34.796 2 3.121 14.186 48.982 3 2.097 9.534 58.516 4 1.284 5.835 64.351 5 .877 3.988 68.338 6 .838 3.810 72.148
Results of factor analysis for ‘price-sensitivity’, ‘post-purchase dissonance’ and ‘emotions’ constructs
