University of Groningen
The interaction effect
of purchase history and
average product rating
on purchase intention
in a situation of brand
familiarity and
unfamiliarity
A study in the mobile phone industry
Abstract
Table of content
1.
Introduction1.1 Social proof page 4
1.2 Brand familiarity page 5
1.3 Mobile phone market page 5
1.4 In this research page 6
2. Conceptual background
2.1 Herd behavior through purchase history and average product rating page 8
2.1.1 Main effects page 8
2.1.2 Interaction effects page 9
2.2 Brand familiarity page 9
2.2.1 Main effect page 9
2.2.2 Interaction effect page 10
2.3 Covariates page 10
2.4 Conceptual model page 12
3. Methodology
3.1 Type of research page 13
3.2 Method page 13
3.3 Procedure page 13
3.4 Participants page 16
3.5 Operationalisation of variables page 17
3.6 Dependent variables page 19
3.7 Manipulation check page 19
3.8 Covariates page 20
3.9 Control variables page 21
3.10 Plan of analysis page 21
4. Results
4.1 Product attitude and purchase intention page 23
4.3 Regression analysis page 26
4.4 Outcomes of proposed hypothesis page 28
5. Discussion, limitations and further research recommendations
5.1 Discussion of the results page 30
5.2 Limitations and further research recommendations page 31 6. Appendix
Appendix A. Factor analysis and Cronbach’s alpha of constructs page 33 Appendix B. VIF values to check for multicollinearity page 34
1. Introduction
1.1 Social proof
The emerging online economy provides consumers with easy access to numerous choices of products (Chen, 2008). Moreover, the internet made some significant changes to the way we search for information and how we adopt this information. Previously, when consumers needed information, they looked at third-party certification or sought advice from friends (King et al., 2014). Nowadays, electronic interaction via the internet allows consumers to socially interact with one another, exchange product-related information, and make informed purchase decisions via computer-mediated conversations (Hoffman & Novak, 1996; Blazevic et al., 2013).
Most online retail websites provide a great deal of information about other online users’ choices and product popularity (Liu et al., 2016). This information includes several social aspects and signals to attract customers (Busalim & Hussin, 2016). One increasingly important source of signals is social proof, where consumers rely on the collaboratively shared information and experiences of other to infer a course of action (Neelameghan & Jain, 1999; Rao et al., 2001; Reinstein & Snyder, 2005). Social proof is an important cue to behave in a certain manner and is widely used in online retailing according to Cisco Internet Business Solutions Group (2013). This research will focus on two common online marketing instruments of social proof: average product ratings (Chen, 2008) and purchase history of the product (Huang & Chen, 2006; Chen, 2008).
Hoffer (1955) was one of the first researcher to highlight the principle of social proof: “When people are free to do as they do, they usually imitate each other”. A famous example of this principle in the literature is the Asch conformity experiment (Asch, 1956). In this experiment people made a decision which was similar with the group decision of the other members of the group, even though it was very clear that this answer was wrong. This construct is known as herding behavior and can be seen as a form of social proof, and has been commonly used by marketers to induce consumer purchase intentions, for years (Bearden & Etzel, 1982).
In a consumption setting, herd behavior is defined as a change in consumer product evaluations, purchase intentions, or purchase behavior resulting from exposure to the evaluations, intentions, or purchase behaviors of referent others (Asch, 1956). Therefore, people frequently select popular brands because they believe popularity indicates better quality (Chen, 2008). An online marketing instrument to reflect popularity is the purchase history of the product (i.e. the total amount of sold products) or the average product rating.
the comparison among similar products (Filieri, 2015). A study conducted by Cisco Internet Business Solutions Group in 2013 among 5000 shoppers showed that online ratings and reviews on retailer websites are considered as the main source of information when purchase decisions are made (52%). Followed by advice from friends and family members (49%) and advice from store employees (12%).
Previous studies have investigated herd behavior in digital auctions (Dholakia, et al., 2002; Stafford et al., 2006), in the internet bookstore (Chen, 2008), in daily-deal sites (Li & Wu, 2012) and software downloading, and bid numbers and download counts have been used by consumers to indicate quality and suitability (Hanson & Putler, 1996). In digital auctions, buyers tend to bid for listings that others have already bid for, while ignoring similar or more attractive listings that people did not bid on (Dholakia et al., 2002).
Studies on online word of mouth have demonstrated that average product ratings have significant influences on consumers' purchasing decisions (Senecal and Nantel, 2004; Chevalier and Mayzlin, 2006).
According to a study about online book purchasing by Chen (2008) sales volume and star ratings serve as heuristic cues for making purchase decisions. A heuristic is defined as a mental generalization of knowledge based on previous experience that provides a shortcut in processing information (Fiske & Taylor, 1984). Specifically, high sales volumes or star ratings can influence consumers online product choices. Also Ye et al. (2013) mention that specifically higher purchase history is associated with higher sales.
1.2 Brand familiarity
The findings of Chen (2008) implies that this might be an effective tool for known brands, since these brands are more likely to acquire high sales volume or star ratings. Therefore it is interesting to make a distinguish between familiar brands (i.e. whether the consumer is familiar with a brand or knows the brand) and unfamiliar brands (Purnawirawan et al., 2015). Moreover, familiar or known brands have strong memory networks, while unknown brands have no (or little) existing cognitive structures (Berger et al., 2009; Connors et al., 2011). Besides, according to Ye et al. (2013) familiarity of the brand influences customer’s decision making. Nevertheless, Ye et al. (2013) highlight that more research is needed in the field of brand familiarity to give reliable outcomes on the actual effects of familiarity with the products on customer’s decision making.
1.3 Mobile phone market
choices, but also due to uncertainty regarding the set of options (Greenleaf & Lehmann, 1995). In a market where consumers are faced with plentiful information, people often imitate others rather than make better decisions responded to conditions (Chen, 2008).
Previous studies examining on signals of social proof, have shown that they are most useful and effective for products that a shopper has yet to experience (Kirmani & Rao, 2000). Experience goods are products for which product quality is relatively difficult to ascertain before consumption (e.g. massages). Search goods are products for which product quality is relatively easier to ascertain before consumption (e.g. clothing, accessories) (Nelson 1974; Hong et al., 2012). Mobile phones are classified in the mid of the experience good and search goods (Tsang & Hsu, 2009). According to Li and Wu (2014) there is a stronger effect of herding for experience goods than for search goods because the additional signals from prior sales are more valuable for evaluating experience goods whose value is harder to ascertain. Mobile phones are classified in the mid of the experience good and search goods (Tsang & Hsu, 2009) and therefore it is interesting to see the strength of the different variables on the purchase intention of mobile phones.
In the mobile phone market retailers make use of different online marketing instruments of social proof. Some retailers only display the average product rating on their websites (e.g. Belsimpel.nl and Mobiel.nl). Other retailers use popularity sales rankings in combination with average product rating to display the product on their website (e.g. Beslist.nl). While other retailers do not show any additional information based on customer evaluations or previous transactions (e.g. Telfort.nl). Since retailers seem to have different approaches to the use of different online marketing instruments of social proof it is interesting to perform more research in this matter. Moreover, different online retailers are offering the same products and can only stand out in price and service. Therefore it is important for retailers to influence consumers in the decision making during the buying process.
Furthermore, in the mobile phone market there are almost 100 different mobile phone brands (phonegg.com). Since people use all accessible information when the brand is unfamiliar (Todd & Benbasat, 2000), some brands which are very well known (e.g. Apple and Samsung) might need different approaches than unfamiliar brands (e.g. Wileyfox, and Xiaomi).
With the emergence of the internet, it is important to understand the potential of online marketing instruments in influencing consumer product choices and to exploit the numerous opportunities it creates. Based on the theory and the practical insights we want to answer the following research question:
1.4 In this research
In this thesis we will conduct an experiment in which participants are confronted with the average product ratings in a situation in which they are familiar with the brand or unfamiliar. Also the influence of availability of the purchase history will be researched in a situation in which consumers are familiar or unfamiliar with the brand. Since resources are limited for this research, and previous research has highlighted that the effects are bigger, this paper limits itself by focusing only on high purchase history of the product and high average product ratings. To the best of my knowledge this is the first study that focuses on average product ratings and purchase history of the product while considering the different effects for brand familiarity and unfamiliarity.
2. Conceptual Background
In this section relevant literature will be discussed, relevant relationship will be highlighted and hypotheses will be constructed. Starting with purchase history of the product, followed by average product rating and ending with knowledge about brand familiarity.
2.1 Herd behavior through purchase history and average product rating 2.1.1 Main effects
The availability of purchase history can generate a herding effect: “A situation where everyone is doing what everyone is doing” (Banerjee, 1992).
As mentioned, consumers frequently use others’ evaluations about a product as an indicator of quality (Park & Lessig, 1977). Specifically, when consumers do not have enough information, they are more likely to refer to others’ opinions or behaviors because it is safer to follow others’ decision (e.g., Asch, 1956; Allen, 1965; Bearden & Etzel, 1982; Banerjee, 1992;). This matter has also been studied in online shopping environments. Researchers have found that product sales volume positively influences consumers online choices regarding that product (Huang & Chen, 2006; Hui-Ying et al., 2010). The reason for this is that consumers perceive other users’ final choice as more reliable information than their own private information (Dholakia et al., 2002; Chen et al., 2011). Moreover, purchase history signals the quality of the item to the buyer as it represents choices made by previous customers (Ye et al., 2013).
Given that people use behavior of others as reference, this leads to the following hypothesis:
H1: The presence of a high purchase history positively affects purchase intention
Showing high star numbers of average customer reviews can influence online buying behavior and bring about an informational cascade (Chen & Lin, 2008). Informational cascades refer to the situation ‘‘when it is optimal for an individual, having observed the actions of those ahead of him, to follow the behavior of the preceding individual without regard to his own information’’ (Bikhchandani et al., 1992). The reason for this might be that favorable average product ratings are viewed as signals of quality (Moon et al., 2010). Furthermore, average product ratings can influence product sales by changing consumer valuation of the products (Chevalier & Mayzlin, 2006).
As a result, purchase intention will increase in accordance with the presence of a high average product rating which leads to the following hypothesis.
2.1.2 Interaction effect
Generally, we assume that the interaction effect between a high purchase history amount and high average product rating will be higher than the individual effects of both. As mentioned, currently none of the mobile phone retailers are using this combination on their websites.
One reason why retailers do not frequently display a combination of product rating and purchase history could be find in the research by You et al. (2015). Their findings present that displaying average product ratings is more effective for products where consumer cannot view the previous consumption of the product. Because, if consumers see a product being used, they may buy the product to conform with others under certain conditions (Schmidt and Spreng 1996), which reduces the benefits of information obtained through the average product rating (You et al., 2015).
For these reasons we assume that the individual added effects of displaying the purchase history or the average product rating is higher than the interaction effect between the purchase history and average product rating.
H3: The aggregate effect when both high purchase history and high average product rating are
present on purchase intention is lower than the added individual effects of high purchase history and high average product rating.
2.2 Brand familiarity 2.2.1 Main effect
which leads to the following hypothesis.
H4: If a brand is familiar the purchase intention will increase
2.2.2 Interaction effect
As consumers have obtained a certain knowledge or expertise regarding familiar brands and have no or little knowledge and expertise regarding an unfamiliar brand, the informative effect of online reviews is stronger for unknown than for well-known brands (Vermeulen & Seegers 2009; Zou et al., 2011). When consumers are familiar with a brand, they are capable of making their own judgment without relying on online reviews. In complex and unfamiliar situations, people attempt to utilize all accessible information (Todd & Benbasat, 2000).
Both purchase history and average product ratings can be seen as heuristic cues (Chen, 2008). When people are less informed, heuristic cues influence people’s behavior (Wood, 2000). Furthermore, consumers may use heuristics, such purchase history of the product and average products ratings, to minimize the effort of decision-making (Lee & Geistfeld, 1998). Therefore, it is expected that the change in purchase intention, when purchase history is present, is significantly greater for consumers when the brand is unfamiliar than when the brand is familiar, in comparison to a situation where purchase history is not present.
H5: The presence of a high purchase history will result in a relatively more positive effect on
purchase intention when the brand is unfamiliar than when the brand is familiar.
Since average products ratings can also be seen as heuristics cues, the same expectation will account for average product ratings.
H6: The presence of a high average rating will result in a relatively more positive effect on
purchase intention when the brand is unfamiliar than when the brand is familiar. 2.3 Covariates
H7: The extent of general attitude towards purchase history will positively influence purchase
intention.
H8: The extent of general attitude towards purchase history will positively influence the relation
between purchase history and purchase intention.
H9: The extent of general attitude towards average product ratings will positively influence
purchase intention.
H10: The extent of general attitude towards average product ratings will positively influence the
relation between purchase history and purchase intention.
The variable product involvement is identified as covariate since consumers with higher product involvement are more likely to perceive attribute differences, to place higher importance on the product, and to possess greater commitment in their brand choices (Howard and Sheth 1969). Further, higher product involvement motivates consumers to search for more information and spend a greater amount of time making an optimal decision (Clarke and Belk 1978). Moreover, cues (such as purchase history and average product rating) are more influential when involvement is low (Chaiken, 1980; Petty et al., 1983; Petty & Cacioppo, 1986). Therefore the results might differ substantial between participants who are highly involved and participants who are low involved with mobile phones. This leads to the following hypothesis.
H11: The extent of product involvement will positively influence purchase intention. H12: The presence of a high purchase history will result in a relatively more positive effect on
purchase intention when involvement is low than when involvement is high.
H13: The presence of a high average product rating will result in a relatively more positive
2.4 Conceptual model
3. Methodology
In this section the methodology of the empirical research to test the hypothesis will be discussed. Starting with an introduction of the type of research, then the method, procedure and participants will be discussed. Following with the operationalization of the variables and the manipulation check.
3.1 Type of research
To determine the effects of the variables, a causal research in the form of an experiment was performed. This allows us to have a very high level of control over the variables and to manipulate the variables in such a way that they are useful for this research (Krik, 2013). An online questionnaire was created using professional survey design software called Qualtrics and was composed of closed-ended questions that were measured using a 7-point Likert scale ranging from strongly disagree (1) to strongly agree (7) or by the use of a slider from totally disagree (0) to totally agree (100).
3.2 Method
To validate the research model, the effect of availability of purchase history, average product rating, brand familiarity and the two-way interactions effects on purchase intention was studied in a 2 (high purchase history of product: present vs. not present) x 2 (high average product rating: present vs. not present) x 2 (brand familiarity; familiar vs. unfamiliar) factorial between-subject design as is suggested by (Qiu et al., 2012). This allows us to explore possible individual effects but also different interaction effects of the three independent variables on the dependent variable: purchase intention, as stated in the hypotheses. This design produced eight conditions (see figure 2).
3.3 Procedure
For the manipulation of high purchase history an extra rectangular block was created above the phone. A Previous research has shown that a rectangle block is an effective way to make the stimuli clearly visible in an online setting (Qiu et al., 2012). The design of the text and block were the same as in the original lay out as shown in figure 3. In the no purchase history present condition, this block was removed again while maintaining the same lay out in the rest of the image. The number 8438 was used to represent high purchase history. Based on common sense I assume that this amount of customers is representing the term high purchase history well. To make sure that customers actually perceived this number as high a control question was added to check this condition.
created next to each other.
To avoid being affected by other factors, related features of the phones and lay out of the web page were kept identical including, price, performance and availability. Other parts such as mobile network providers were faded to avoid being of any influence.
The survey was programmed that participants had to watch the stimuli for at least 10 seconds. After the participants had seen the manipulation, questions regarding product attitude and purchase intentions were asked. Hereafter, questions about the manipulation check were asked to make sure the manipulation was correct. Following, questions regarding the covariates were asked. To increase internal validity some questions regarding the manipulation check were asked. Following up with questions regarding covariates. Finally, general questions were asked regarding age, ethnicity and gender. The questionnaire lasted approximately 7 minutes.
3.4 Participants
of in total 199 people, 92 (46.2%) male and 107 (53.8%) woman with an average age of 25 (Mage= 25.17, SD=8.44). Of which, 175 people were Dutch and 24 people from other countries. All participants were equally distributed between the eight different conditions, but since 17,4% of the participants did not fulfill the criteria some differences between the number of people in each group occurred. An overview of the statistics per condition can be found in table 2.
Condition Number of participants
Gender Age
1 N=21 Male= 43%, Female= 57% Mage= 25.19, SD=9,11 2 N=25 Male= 44%, Female= 56% Mage= 23.52 SD=4.25 3 N=23 Male= 43%, Female= 57% Mage= 23.04, SD=2.71 4 N=15 Male= 40%, Female= 60% Mage= 27,53, SD=11.19 5 N=36 Male= 47%, Female= 53% Mage= 27.77 SD=11,04 6 N=25 Male= 32%, Female= 68% Mage= 22.68, SD=2.91 7 N=28 Male= 54%, Female= 46% Mage= 27.11, SD=12.32 8 N=26 Male= 62%, Female= 38% Mage= 24.08, SD=5.62
3.5 Operationalisation of variables
Variable Source Items Cronbach’s Alpha Purchase intention Bickart & Schindler, 2001; Garbarino & Strahilevitz, 2004; Chu et al., 2005
-How likely is it that you will buy this product?"
-
Product attitude
Batra & Ray, 1986 -How do you feel about this mobile phone? 1. Bad - Good 2. Dislike - Like 3. Unpleasant - Pleasant 4. Unfavorable - Favorable .90 General attitude towards ratings
Park et al., 2007 -When I buy a product online, I always look at the rating that is presented on the website -When I buy a product online, the ratings presented on the website are helpful for my decision making .84 General attitude towards purchase history
Park et al., 2007 -When I buy a product online, I always look at the amount of previous sales that is presented on the website, if available
-When I buy a product online, the amount of previous sales presented on the website are helpful for my decision making
-If I do not pay attention to the previous amount of sales presented on the website when I buy a product online, I worry about my decision.
.78
Product involveme nt
Bruner & Hensel, 1994
-How intensively do you use mobile phones -How much are you involved with mobile phones
-How much do you feel like you are a mobile phone expert
-How intensively are you interested in mobile phones, relative to other people?
3.6 Dependent variables
The level of purchase intention was measured by a factor which asks the participants to evaluate how likely it is that they will buy a product online on a 7pt Likert scale: one being extremely unlikely, and seven being extremely likely. According to prior research, a single item has been used to measure consumers’ purchase intention in online settings (Bickart & Schindler, 2001; Garbarino & Strahilevitz, 2004; Chu et al., 2005).
Product attitude was measured with four items on a 7pt Likert scale adapted from Batra and Ray (1986). To check the reliability of the measures, a factor analysis was performed and the Cronbach’s alpha for each item was calculated. All items together had a Cronbach’s alpha above the benchmark of 0.70 (Nunnaly & Bernstein, 1994). The true values are listed in Appendix A.
3.7 Manipulation check
To validate that the stimuli was presented correctly to the participants questions about the manipulation were asked to the participants of the manipulated conditions. The first question was: Did you notice that the total number of mobile phones sold was shown in the previous picture? And: Did you notice that the overall rating of the mobile phone was shown in the previous picture? This was measured using a dichotomous scale. Noticeable is that relatively a lot of participants have not consciously noticed the stimuli (see table 3). Therefore I’ve decided to move the participants from conditions with stimuli to groups representing conditions without that specific stimuli. For example, if a participant of condition 5 indicates that he or she has not seen the product rating, which was present in condition 5, this participant will now represent the condition in which rating was not present, meaning this specific participant will be moved to condition 7. This resulted in bigger differences between groups sizes and therefore limits the generalizability of this research. The new descriptives per condition can be found in table 4.
Stimuli Number of participants that have been faced with the stimuli
Condition Number of participants
Gender Age
1 N=13 Male= 46%, Female= 54% Mage= 22.31, SD=2.25 2 N=18 Male= 39%, Female= 61% Mage= 23.72, SD=4.88 3 N=18 Male= 50%, Female= 50% Mage= 23.78, SD=6.00 4 N=19 Male= 53%, Female= 47% Mage= 26.47, SD=10.15 5 N=16 Male= 47%, Female= 53% Mage= 24.06, SD=6.52 6 N=24 Male= 29%, Female= 71% Mage= 22.42, SD=2.73 7 N=60 Male= 47%, Female= 53% Mage= 28.12, SD=12.01 8 N=30 Male= 57%, Female= 43% Mage= 24.20, SD=5.28
3.8 Covariates
To control for possible confounding effects, participants were asked to fill in questions about their general attitude towards product ratings, general attitude towards purchase history, product involvement, and personality. This was done to check if different personality characteristics and preferences do not influence the outcome of the survey.
To measure the general attitude towards product ratings and general attitude towards purchase history the construct of Park et al. (2007) was used. In this study general attitude was measured based on 5 items on a 7pt Likert scale. One of these items, for either general attitude towards ratings and general attitude towards purchase history, was deleted upfront, since the Dutch translation of this item is almost similar to another item in the scale. Therefore the questions:
0.7 when two items were deleted. For the variable general attitude towards purchase history Cronbach’s alpha was above 0.7 when one item was deleted. The true values are listed in Appendix A. In order to measure the extent of product involvement the scale from Bruner and Hensel (1994) was adopted. This scale was consists of four items and was measured using a slider from none at all (0) to A great deal (100). The Cronbach’s alpha was calculated and was above 0.7. The true values are listed in Appendix A.
3.9 Control variables
This research The control variables age was measured by giving the opportunity to fill in a number. The control variable gender was measured with a dichotomous scale. Brand loyalty was measured by a single item: “Do you intend to buy the same brand as your current phone brand again?”. The participants could answer on a 7pt Likert scale ranging from definitely not to
definitely yes.
3.10 Plan of analysis
To identify the results of the experiment we will perform an ANOVA analysis and a regression analysis (Cohen, 1968). We began the analysis by performing a three-way ANOVA. Many experimental designs that perform ANOVA analysis use sample size of 30 or above per condition (e.g. Senecal & Nantel, 2004). Nevertheless it is also not uncommon to find group sample size of less than 20 per condition (e.g. Roskos-Ewoldsen et al., 2002; Ziegler et al., 2004; Ziegler & Diehl, 2011). According to Balkin and Sheperis (2011) small sample sizes reduce the power to detect a significant effect, but also decrease the chance of finding significance when there is none. Since there are only limited amount of participants within each group after the reconstruction of the dataset we have to take these limitations into account.
loss of power of the effects.
4. Results
First the results regarding the ANOVA tests will be presented, following by the results of the regressions analysis and lastly an overview of outcomes of the hypothesis.
4.1 Product attitude and purchase intention
The results of the ANOVA analysis reveal that the three independent variables explain 2% of the variance in purchase intention (R²=.02) and 9% of product attitude (R²=.09). For this reason any further analyses will be performed with product attitude as dependent variable.
4.2 ANOVA
The main effects of rating-present and purchase history on product attitude successfully reached significance as can been seen in table 6. Participants in the rating-present conditions ( M=4.56,
SD=1.19) show a higher mean value than the no rating-present conditions (M=4.01, SD=1.25). The results also show significance for the main effect of purchase history-present. For the purchase history-present conditions participants show a higher mean value ( M=4.56, SD=1.05) than in the no purchase history-present conditions ( M=3.99, SD=1.32), meaning that both individual effects of rating-present and purchase history-present support current literature. There was no statistically significant two-way interaction between rating-present, and purchase history-present on product attitude (see table 6). Participants who saw ratings and purchase history, had a higher product attitude ( M=4.81, SD=1.05) than the individual outcomes in which only rating was present (M=4.34, SD=1.28) or only purchase history was present (M=4.36,
SD=1.02) as can be seen in figure 8.
Effect on Product attitude Coefficient Alpha
Rating-present F(1.197)=8.93 .00
Purchase history-present F(1.197)=9.60 .00
Rating-present * purchase history F(1.195)=.01 .91
Brand familiarity F(1.197)=.24 .63
Brand familiarity * rating-present F(1.195)=1.84 .18 Brand familiarity * purchase history-present F(1.195)=0.21 .65
4.3 Regression analysis
Furthermore, the variables: the general attitude towards purchase history and ratings and product involvement were included as covariates. We performed a multiple regression analysis in order to examine the degree of multicollinearity, to test the hypothesis and to analyze the control variables.
familiarity (BF) PH * PR -.03 .19 -.08 -.06 .25 PH * BF .03 .17 .00 .04 .27 PR * BF .41 .35 .44 .53 .44 Product involvement (PI) .57** .59* PI * PH -.44 -.47 PI * PR -.22 -.38 General attitude towards PH .05 -.04 Att. PH * PH .09 .20 General attitude towards PR -.07 -.03 Att. PR * PR .53 .58 Brand loyalty -.10 Age -.02 Gender -.25 F(6,192)= 3.33*, R²=.09, R²a=.07 F(9,168)= 2.77*, R²=.13, R²a=.08 F(8,189)= 2.70*, R²=.10, R²a=.07 F(8,190)= 2.78*, R²=.11, R²a=.07 F(16,159)= 2.11*, R²=.18, R²a=.09 *=p<0.01,**=p<0.05, ***=p<0.10
added as covariate. In comparison to model 1, the R² only increased minimal to 10%. For model 4, only attitude towards purchase history was added to the base model. This increased R² slightly to 11%. Model 5 included all independent variables, covariates and control variables, the full model explains 18% of the variance.
In model 2 sales-present is moving towards to significant (β=.59, p=.12 and covariate product involvement is significant (β=.57, p=.02). These results mean that consumers value the present of purchase history, but are also influenced by their personal product involvement. In model 4 purchase history-present is again close to significant (β=.53, p=.08), showing the importance of purchase history even further. In the full model, model 5, the influence of covariate product involvement is noticeable again (β=.59, p=.03). Meaning that participants were positively influenced by their personal product involvement. The independent variables average product rating and brand familiarity, and the covariates general attitude towards purchase history and general attitude towards average product rating did not significant contribute to the multiple regression model.
4.4 Outcomes of proposed hypothesis
Hypothesis 1 to 6 were tested by the use of an ANOVA test to see whether the present of the independent variables purchase history, average product rating and brand familiarity did influence the dependent variable product attitude. Hypothesis 7 to 13 were tested by performing a regression analysis to see whether the extent of general attitude towards purchase history, general attitude towards average product ratings and product involvement did influence the product attitude. The outcomes of the hypothesis can be found in table 11.
Hypothesis Accepted/
Rejected
H1: The presence of a high purchase history positively affects purchase intention Accepted H2: The presence of a high average product rating positively affects purchase
intention
Accepted
H3: The aggregate effect when both high purchase history and high average
product rating are present on purchase intention is lower than the added individual effects of high purchase history and high average product rating.
Rejected
H5: The presence of a high purchase history will result in a relatively more positive effect on purchase intention when the brand is unfamiliar than when the brand is familiar.
Rejected
H6: The presence of a high average rating will result in a relatively more positive effect on purchase intention when the brand is unfamiliar than when the brand is familiar.
Rejected
H7: The extent of general attitude towards purchase history will positively influence purchase
Rejected
H8: The extent of general attitude towards purchase history will positively influence the relation between purchase history and purchase intention.
Rejected
H9:The extent of general attitude towards average product ratings will positively influence purchase intention.
Rejected
H10: The extent of general attitude towards average product ratings will
positively influence the relation between purchase history and purchase intention.
Rejected
H11: The extent of product involvement will positively influence purchase
intention.
Accepted
H12: The presence of a high purchase history will result in a relatively more positive effect on purchase intention when involvement is low than when involvement is high.
Rejected
H13: The presence of a high average product rating will result in a relatively more positive effect on purchase intention when involvement is low than when involvement is high.
5. Discussion
5.1 Discussion of the results
Average product ratings were already extensively researched by several authors. However, research regarding the present of purchase history and regarding brand familiarity was still very limited. The objective of this research was to give managers more insight in useful online marketing tools to highlight products in their online retail store for both brand familiar and brand unfamiliar products.
It was proposed that average product rating and purchase history would increase product attitude. The assumptions were tested by an experiment in which participants were randomly assigned to one of eight conditions. The results of this study show positive main effects of the present of both purchase history of the product and the present of average product rating on product attitude. These effects confirm the expected positive relationship between those variables. This result indicates that buyers do take purchase history and average product rating into account when developing a product attitude. These findings support previous literature about purchase history of the product (Huang & Chen, 2006; Hui-Ying et al., 2010; Ye et al., 2013) and average product rating (e.g. Chevalier & Mayzlin, 2006).
Furthermore, it was proposed that the interaction effect between purchase history and average product rating was smaller than the added individual effects of both. The relationship of the proposed interaction effect between the variables was not found to be significant. Meaning that the added value of using both purchase history and product rating is not significant better than using one of the two online marketing instruments. Although the mean scores of the interaction effect between purchase history and average product rating tent to move towards the the opposition of the proposed hypothesis. Meaning that the present of both purchase history of the product and average product rating might result in a higher product attitude.
familiar. Both of these assumptions are rejected because no significant relation was found. Surprisingly the findings move toward to opposite to what is predicted. Meaning that the effects of purchase history and average product rating were bigger when the brand was familiar.
Product involvement seemed to be an important predictor of purchase intention. People with a higher product involvement had a significant higher product attitude. The other covariates general attitude towards purchase history and general attitude towards average product ratings were not of any significant influence on product attitude.
Concluding, we can say that online marketers should use the online marketing instruments purchase history of the product and average product rating to increase a customer's’ product attitude towards a product. Marketers should use these instruments for all brands since no significant difference between familiar brands and unfamiliar brands were found. This is especially true in circumstances where consumers are highly involved in the product.
5.2 Limitations and further research recommendations
Because limited studies exist on online herding behavior and brand familiarity, numerous possible research possibilities exist. First, this study only examined one kind of product, namely mobile phones. Future studies could consider to investigate other products to get a better understand of the value of social proof in the online marketplace.
The results might be affected because the research was done in the mobile phone market. In industries, which are characterised by an evolving technology and a high level of product innovation, consumers are less likely to rely on prior knowledge and more likely to rely on externally retrieved information (Punj & Staelin 1983; Hulland & Kleinmuntz 1994). Therefore the differences of product attitude brand familiarity and brand unfamiliarity might have been very small. Because the participants might have ignored all prior knowledge about the familiar brand and only relied on the retrieved information.
Within this research the brand HTC was used to represent a familiar brand. Of the participants only 5,5 % uses a HTC currently. It would be interestingly to see if the effects are different if research is done on a brand which is more commonly used. I would recommend to do research with the use of either Apple (52,3%) or Samsung (21,1%), since these brands seem to be the most popular brand among young people.
Furthermore, this study limits itself to only high average product ratings. Reichheld and Sasser (1990) indicated that positive information can increase revenue by attracting new customers. Meanwhile, negative information reduces the credibility of corporate advertising (Solomon, 1998). Therefore it might also be interesting to focus on negative average product ratings.
As well for purchase history of the product this study limits itself to only high purchase history of the product. Present research has only focused on high purchase history which has proven itself to be associated with higher sales (Ye et al., 2013). Further studies could examine the effect of low purchase history, this can be useful for firms with low brand familiarity since those are often less frequently sold (Chen, 2008).
Furthermore this study was conducted in the Netherlands and mainly contained Dutch participants. The results thus may show different effects consumers in other cultures. Therefore further studies may focus on cultural differences in the value of social proof and the effects of brand familiarity.
Additionally, the average age of the participants was 25 and therefore this study mainly represents the behavior of young people. Further studies could examine the effects for older people.
6. Appendix A
Appendix A: Factor analysis and Cronbach’s alpha of constructs
Construct Cronbach’s alpha
Items Factor
Loadings
Product attitude .90 How do you feel about this mobile phone? 1. Bad - Good 2. Dislike - Like 3. Unpleasant - Pleasant 4. Unfavorable - Favorable .78 .83 .79 .71 General attitude towards rating
.84 1.When I buy a product online, I always look at the overall product rating that is presented on the website
2. When I buy a product online, the overall product rating presented on the website are helpful for my decision making .86 .86 General attitude towards purchase history
.78 1. When I buy a product online,I always look at the amount of previous sales that is presented on the website, if available 2. When I buy a product online, the amount of previous sales presented on the website are helpful for my decision making
3. If I do not pay attention to the previous amount of sales presented on the website when I buy a product online, I worry about my decision
.75
.76
.58
Product involvement .78 1. How intensively do you use mobile phones?
2. How much are you interested in mobile phones?
3. How much do you see yourself as a mobile phone expert?
4. How intensively are you interested in mobile phones, relative to other people?
Appendix B
Appendix B: VIF values to check for multicollinearity
Variable VIF Variable VIF
7. References
Aaker, D.A. (1991). Managing Brand Equity. New York: The Free Press.
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Newbury Park: Sage.
Allen, V. L. (1965). Situational Factors In Conformity. Advances in Experimental Social Psychology, 133-175.
Asch, S. E. (1956). Studies of independence and conformity: A majority of one against a unanimous majority. Psychological Monographs, 70–79.
Banerjee, A. V. (1992). A Simple Model of Herd Behavior. The Quarterly Journal of Economics, 107(3), 797-817.
Batra, R., & Ray, M.L. . 1986) Affective responses mediating acceptance of advertising. Journal of Consumer Research, 13(2), 234–249.
Bearden, W. O., & Etzel, M. J. (1982). Reference group influence on product and brand purchase decisions. Journal of Consumer Research, 9(2), 183–194.
Benedicktus, R. L., Brady, M. K., Darke, P. R., & Voorhees, C. M. (2010). Conveying
Trustworthiness to Online Consumers: Reactions to Consensus, Physical Store Presence, Brand Familiarity, and Generalized Suspicion. Journal of Retailing, 86(4), 322-335.
Berger, J.A., Sorensen, A.T., & Rasmussen, S. (2009). Positive Effects of Negative Publicity: When Negative Reviews Increase Sales. Marketing Science, 29(5), 815–827.
Bickart, B., & Schindler, R. M. (2001). Internet forums as influential sources of consumer information. Journal of Interactive Marketing, 15(3), 31–40.
Journal of Business Research, 15(3), 251–262.
Blazevic, V., Hammedi, W., Garnefeld, I., Rust, R. T., Keiningham, T., Andreassen, T.W., & Carl, W. (2013). Beyond traditional word-of-mouth. Journal of Service Management, 24(3), 294-313.
Bruner, G. C., & Hensel, P. J. (1992). Marketing scales handbook. Chicago, IL: American Marketing Association.
Busalim, A. H., & Hussin, A. R. (2016). Understanding social commerce: A systematic literature review and directions for further research. International Journal of Information Management, 36(6), 1075-1088.
Chaiken, S. (1980). Heuristic versus systematic information processing and the use of source versus message cues in persuasion. Journal of Personality and Social Psychology, 39(5), 752-766.
Chen, Y. (2008). Herd behavior in purchasing books online. Computers in Human Behavior, 24(5), 1977-1992.
Chen, Y., Wang, Q., & Xie, J. (2011). Online Social Interactions: A Natural Experiment on Word of Mouth Versus Observational Learning. Journal of Marketing Research, 48(2), 238–254. Chevalier, J.A., & Mayzlin, D. (2006). The effect of word of mouth on sales: online book
reviews. Journal of Marketing Research, 43(3), 345–354.
Chu, W., Choi, B., & Song, M. R. (2005). The Role of on-line retailer brand and infomediary reputation in increasing consumer purchase intention. International Journal of Electronic Commerce, 9(3), 115–127
Cisco Internet Business Solutions Group (2013), Catch and Keep Digital Shoppers,
http://www.cisco.com/c/dam/en_us/about/ac79/docs/retail/Catch-and-Keep-the-Digital-Shopper_
PoV.pdf.
Clarke, K., & Belk, R.W. (1978). The Effects of Product Involvement and Task Definition on Anticipated Consumer Effort. In Advances in Consumer Research, 5, 313-318.
Connors, L.,. Mudambi, S.M., & Schuff, D. (2011). Is It the Review or the Reviewer? A
Multi-method Approach to Determine the Antecedents of Online Review Helpfulness. In System Sciences (HICSS), 44th Hawaii International Conference on: IEEE, 1-10.
Corbitt, B., Thanasankit, T. & Yi, H. (2003). Trust and e-commerce: a study of consumer perception. Electronic Commerce Research and Applications, 2(3), 203-215.
Dholakia, U. M., Basuroy, S., & Soltysinski, K. (2002). Auction or agent (or both)? A study of moderators of the herding bias in digital auction. International Journal of Research in Marketing, 19, 115–130.
Filieri, R. (2015). What makes online reviews helpful? A diagnosticity-adoption framework to explain informational and normative influences in e-WOM. Journal of Business Research, 68(8). 1261-1270.
Fiske, S., & Taylor, S. (1984). Social cognition. Reading, MA: Addison-Wesley.
Garbarino, E., & Strahilevitz, M. (2004). Gender differences in the perceived risk of buying online and the effects of receiving a site recommendation. Journal of Business Research, 57(7), 768–775.
Greenleaf, E. A.,& Lehmann,D. R. (1995). Reasons for substantial delay in consumer decision making. Journal of Consumer Research, 22, 186–199.
Hanson,W. A., & Putler, D. S. (1996). Hits and misses: Herd behavior and online product popularity. Marketing Letters, 7, 297–305.
Hoffer, E. (1955). The Passionate State of Mind. New York: Harper & Row.
Hoffman, D. L., & Novak, T. P. (1996). Marketing in Hypermedia Computer-Mediated Environments: Conceptual Foundations. Journal of Marketing, 60(3), 50.
Hong, Y., Chen, P.Y., & Hitt, L.M. (2012). Measuring Product Type With Dynamics of Online Product Review Variance. Proceedings of the 33rd International Conference on Information Systems (ICIS), Orlando, Florida.
23(5), 413-428.
Hui-Ying, L., Qiang, Y., & Sharma, G. (2010). Herding behavior in C2C e-commerce: Empirical investigation in China. 2010 International Conference on Management Science & Engineering 17th Annual Conference Proceedings.
Hulland, J.S., & Kleinmuntz, D.N. (1994). Factors Influencing the Use of Internal Summary Evaluations Versus External Information in Choice. Journal of Behavioral Decision Making, 7 (2), 79–102.
Jacoby, J., Speller, D. E., & Berning, C. K. (1974). Brand Choice Behavior as a Function of Information Load: Replication and Extension. Journal of Consumer Research, 1(1), 33.
Keller, K.L. (1993). Conceptualizing, Measuring, and Managing Customer Based Brand Equity. Journal of Marketing, 57(1), 1–22.
King, R.A., Racherla, P. & Bush, V.D. (2014). What We Know and Don't Know About Online Word-of-Mouth: A Review and Synthesis of the Literature. Journal of Interactive Marketing, 28 ,167–183.
Kirmani, A., & Rao, A.R. (2000). No pain, no gain: A critical review of the literature on signaling unobservable product quality. Journal of Marketing, 64(2), 66–79.
Kirk, R. E. (2013). Experimental design: procedures for the behavioral sciences. Thousand Oaks: Sage Publications.
Kutner, M. H., Nachtsheim, C. J., Neter, J. (2004). Applied Linear Regression Models (4th ed.). McGraw-Hill Irwin.
Iacobucci, D., Posavac, S. S., Kardes, F. R., Schneider, Matthew J., & Popovich, D. L. (2015). Toward a more nuanced understanding of the statistical properties of a median split. Journal of Consumer Psychology (to appear).
Lee, J., & Geistfeld, L. V. (1998). Enhancing consumer choice: Are we making appropriate recommendations? The Journal of Consumer Affairs, 32(2), 227–251.
Workshop on information systems and economics. December 15–16, Orlando, FL.
Liu. Q., Huang, S. & Zhang, L. (2016) The influence of information cascades on online purchase behaviors of search and experience products. electronic Commerce Research. 16-4. 553-580. Liu, Q., & Zhang, L. (2014). Information cascades in online reading: An empirical investigation of panel data. Library Hi Tech, 32(4), 687–705.
Maheswaran, D., Mackie, D. M., & Chaiken, S. (1992). Brand Name as a Heuristic Cue. Journal of Consumer Psychology, 1(4), 317-336.
Maity, M., Dass, M., & Malhotra, N. K. (2014). The Antecedents and Moderators of Offline Information Search: A Meta-Analysis. Journal of Retailing, 90(2), 233-254.
Moon, S., Bergey, P. & Iacobucci, D. (2010). Dynamic effects of movie ratings on movie revenues and viewer satisfaction. Journal of Marketing, 74(1), 108-121.
Neelamegham, R., & Jain, D. (1999). Consumer choice process for experience goods: An econometric model and analysis. Journal of Marketing Research, 36(3), 373–386.
Nelson, P. (1974). Advertising as information. The Journal of Political Economy,82(4), 729-754. Nunally, J.C., & Bernstein, I.H. (1994). Psychometric Theory (3rd Edition). New York:
McGraw-Hill.
O’Brien, R. M. (2007). A Caution Regarding Rules of Thumb for Variance Inflation Factors". Quality & Quantity. 41 (5), 673.
Park, D.H., Lee, J., & Han, I. (2007). The Effect of On-Line Consumer Reviews on Consumer Purchasing Intention: The Moderating Role of Involvement. International Journal of Electronic Commerce. 11(4), 125-148.
Park, C. W., & Lessig, V. P. (1977). Students and housewives: differences in susceptibility to reference group influence. Journal of Consumer Research, 4(2), 102–110.
Petty, R.E., Cacioppo, J.T., & Schumann, D. (1983). Central and peripheral routes to advertising effectiveness: The moderating role of involvement. Journal of Consumer Research, 10(2), 135-146.
Punj, G.N., & Staelin, R. (1983). A Model of Consumer Information Search Behavior for New Automobiles. Journal of Consumer Research, 9(4), 366–380.
Purnawirawan, N., Eisend, M., Pelsmacker, P. D., & Dens, N. (2015). A Meta-analytic Investigation of the Role of Valence in Online Reviews. Journal of Interactive Marketing, 31, 17-27
Qiu, L. ,Pang, J., & Lim, K. (2012). Effects of conflicting aggregated rating on eWOM review credibility and diagnosticity: The moderating role of review valence. Decision Support Systems. 54(1), 631-643.
Rao, H., Greve, H. R., & Davis, G. F. (2001). Fools Gold: Social Proof in the Initiation and Abandonment of Coverage by Wall Street Analysts. Administrative Science Quarterly, 46(3), 502.
Reinstein, D.A., & Snyder, C.M. (2005). The influence of expert reviews on consumer demand for experience goods: A case study of movie critics. Journal of Industrial Economics, 53(1), 27–51.
Roskos-Ewoldsen, D.R., Bichsel, J. & Hoffman, K. (2002). The influence of accessibility of source likability on persuasion. Journal of Experimental Social Psychology, 38(2), 137-143. Schmidt, J.B., & Spreng, R.A. (1996). A Proposed Model of External Consumer Information Search. Journal of the Academy of Marketing Science, 24 (3), 246–256.
Senecal, S., & Nantel, J. (2004). The influence of online product recommendations on consumers' online choices. Journal of Retailing, 80(2), 159–169.
Stafford, M. R., Kilburn, A. J., & Stern, B. B. (2006). The effects of reserve prices on bidding behaviour in online auctions. International Journal of Internet Marketing and Advertising, 3(3), 240–253.
Todd, P., & Benbasat, I. (2000). Inducing compensatory information processing through decision aids that facilitate effort reduction: An experimental assessment. Journal of Behavioral Decision Making,13(1), 91–106.
Tsang, S.S., & Hsu, W.C. (2009). Transforming experience good into search good: how virtual experience may change the internet advertising market?. Proceedings of the 9th International Conference on Electronic Business in Minneapolis, MN, USA, 437-445.
Yates, D.S., Moore, D.S., & Starnes, D.S. (2008). The Practice of Statistics, 3rd Ed. Freeman. Ye, Q., Cheng Z, J., & Fang, B. (2012). Learning from other buyers: The effect of purchase history records in online marketplaces. Decision Support Systems.
You, Y., Vadakkepatt, G. G., & Joshi, A. M. (2015). A Meta-Analysis of Electronic Word-of-Mouth Elasticity. Journal of Marketing, 79(2), 19-39.
Vermeulen, I.E., & Seegers, D. (2009). Tried and tested: The impact of online hotel reviews on consumer consideration. Tourism Management, 30(1), 123–127.
Wood, L. (2000). Brands and brand equity: definition and management. Management Decision, 38(9), 662-669.
Ziegler, R., & Diehl, M. (2011). Mood and multiple source characteristics: mood congruency of source consensus status and source trustworthiness as determinants of message scrutiny.
Personality and Social Psychology Bulletin, 37(8), 1016-1030.
Ziegler, R., Diehl, M., Zigon, R., & Fett, T. (2004). Source consistency, distinctiveness, and consensus: the three dimensions of the Kelly ANOVA model in persuasion. Personality and Social Psychology Bulletin, 30(3), 352-364.
Zou, P., Yu, P.B., & Hao, Y. (2011). Does the Valence of Online Consumer Reviews Matter for Consumer Decision Making? The Moderating Role of Consumer Expertise. Journal of
