Framing Effects in Consumers’ Decision-Making Process on Attitudes and
Purchase Intention
Master’s Thesis
Graduate School of Communication Master’s Program Communication Science Persuasive Communication
Student: Anna Thenner Student Number: 12242225 Supervisor: Prof. Dr. Ed Peelen
Abstract
This study gains insight on a new method to present options in an online environment with the interactive element of customizing a product in order to reduce people’s choice overload when making a purchase decision. Therefore, a non-interactive frame, which represents the current style of presenting options as a full menu, is compared to more interactive adding and deleting frames. For this quantitative research, participants are asked to choose their preferred pizza in an online experiment. With a questionnaire, their attitudes towards the decision-making process and the final decision are evaluated as well as their purchase intention. Besides, the number of options they choose was measured. Only the frames, not the attitudes influence the number of options participants choose. People in the adding frame choose clearly less options compared to people in the other frames. The purchase intention is equally high across the frames. Positive attitudes are found to positively influence the purchase intention. People in the interactive frames have more positive attitudes towards the making process than people in the non-interactive frame. Additionally, consumer decision-making styles are evaluated to discover possible influences on consumer’s attitudes. But consumer decision-making styles do not influence consumer’s attitudes across the frames.
Framing Effects in Consumers’ Decision-Making Process on Attitudes and Purchase Intention Consumers are confronted with an endless amount of products and services online. Additionally, to the huge number of retailers online, a lot of retailers have an overwhelming sum of products to choose from. This results in consumers that are confused by choice overload or online shopping anxiety, which can negatively influence the quality of decisions and discourage some people to shop online (Nagar & Gandotra, 2016). Chernev, Böckenholt and Goodman (2015) identified difficulty of the decision task, complexity of the choice set, consumer’s preference uncertainty and consumer’s decision goal as factors that influence, if a higher number of options causes choice overload. Besides, they name two main strings of consequences of choice overload: Consequences on the subjective state and consequences of the behavioral outcome. What has not been investigated more extensively, were customers’ attitudes after dealing with choice overload. Furthermore, choice overload occurs more often in situations of average to high similarity (Yun & Duff, 2017) and with a higher number of options (Park & Jang, 2013), meaning that the more similar the products offered by an online store are, the more vulnerable people are to choice overload. That is why, e.g., restaurants that offer pizza online may confuse people with too many options to choose from, as they are all similar. But not all consumers may be influenced by a huge amount of choices the same way. Sproles and Kendall (1986) developed a consumer style inventory (CSI), according to which consumers can be categorized in two main groups, hedonic and utilitarian consumers. While the former enjoys the process of shopping including interactivity, the latter focuses on the quality and value of a purchase and tends to fewer interactive means (Barwitz & Maas, 2018; Gehrt, Rajan, Shainesh, Czerwinski & O’Biren, 2012). The mode in which a website is presented influences consumer’s attitudes (van Noort, Voorvelf & Reijmersdal, 2012) and these attitudes influence consumer’s behavioral intentions (Ajzen, 1991).
The aim of this study is to investigate the effects of different frames on the two different consumer decision making styles regarding to their attitudes towards the
decision-making process and their final choice, as well as their preference for a final option and the resulting purchase intention. Especially in an online environment, marketers should strive to provide services that facilitate consumers’ decision-making process, for example by providing them with decision support systems that allow them to elicit personal preferences (Pilli & Mazzon, 2016). While interactivity has been investigated by making the distinction between personal, semi-personal or impersonal contact to a retailer (Barwitz & Maas, 2018) this study makes the distinction of a non-interactive marketing method to an interactive marketing method based on the possibility to customize a presented item, as this is one construct of interactivity (Voorveld, Neijens & Smit, 2011). One possibility to customize a product or service is to add or subtract items to a base model, which is a method based on the framing literature. In the framing literature, the effects of adding and subtracting frames on different customer types have extensively been analyzed and show consistent findings (Park, Jun & Macinnis, 2000; Biswas & Grau, 2008; Jin, He & Song, 2012; Xiaoyu, Qiao, Jifei & Yi, 2015). In general, customers end up with a higher amount of items in a subtracting frame than in an adding frame. Possible explanations are the anchoring and adjustment model (Wilson, Houston, Etling & Brekke, 1996) and the loss-aversion model (Biswas & Grau, 2008). Connecting this theory to the CSI, the following research question occurs: To what extent does Interactive content marketing positively influence consumers’ Attitudes and
consequentially consumers’ Purchase intention on a website and is this effect moderated by Consumer decision-making styles?
New in the present study is that the two frames are compared to a non-interactive mode that serves as a control condition. This kind of framing has not yet been used to realize interactivity. But especially for smaller companies who are constricted to a small budget and tools to improve their offers online, adding or subtracting frames are a simple method to improve a website with less expenses or expertise. To them, the findings of this study are relevant in order to adapt the display of their products or services based on the dominant
decision-making style of their customers. Additionally, this study contributes to the research on consumers’ decision-making process by combining framing literature with literature on choice overload in consideration of different consumer decision-making styles.
Theoretical framework The role of interactivity in content marketing
The focus of the present study lies on different levels of interactivity on websites that can be retrieved via PC, laptop, tablet or smartphone. Interactivity and the ability to
personalize services are factors to establish distinction from other e-stores (Ganesh, Reynolds, Luckett & Pomirleanu, 2010). Doern and Fey (2006) identify eight key drivers that create value in e-commerce whereof two are “ease of search” and “ease of use” which may vary among consumer types combined with the presentation style of a website. One way to make the search for products easier is to list keywords (Singh, 2002), while traditional approaches provide consumers with more detailed information. Furthermore, Khare, Khare, Mukherjee and Goyal (2016) list services provided by online shops that enhance customers’ use of online shops, such as enjoyment, ease of use and convenience. Effects of these services might vary between consumer types and influence consumers’ attitudes. Furthermore, interactive features are not always perceived as such. The extent to which consumers perceive a website as
interactive is not always in line with the actual interactive features on a website (Voorveld et al., 2011). Interactivity is categorized in the dimensions of two-way communication,
synchronicity and control. More interactive features on a website do not mean that consumers perceive the website as more interactive as a website with less interactive features. But there are six functions that are perceived as contributing to a website’s interactivity. Two of them are the option to customize products and the capability to customize information on the website, which fall in the “control dimension”. In the present study, the choice can be customized in the interactive content marketing methods to ensure interactivity.
Framing Effects
The interactive content marketing method in this study is framed in two different ways in order to compare it to a non-interactive content marketing method. In the literature, there are two dominant theories to explain probable effects that may occur due to the applied framing in the present study. These theories are presented in the following paragraphs.
The principle of the anchoring and adjustment model is, that people tend to decide based on their initial starting or anchor point without making sufficient adjustments (Wilson et al.,1996). When being presented with a product package and the option to add or delete items to it, people’s final decision does not differ a lot from the initial starting point. These anchoring processes occur unintentionally and non-consciously (Wilson et al., 1996). Even arbitrary numbers can anchor people’s judgements without any logical reason for even
considering these numbers as an answer to the presented questions (Wilson et al., 1996). That shows how influencing a relatively simple option frame can be on people’s decision.
Another explanation for option framing effects is the loss-aversion model (Biswas & Grau, 2008) which is derived from the Prospect Theory (Tversky & Kahneman, 1981). A value of a loss of a feature in a given option feels worth more than the value of a feature added to a given option. Consumers always choose more items, when they have to delete items from a fully loaded model than when they have to add items to a base model (Biswas & Grau, 2008; Levin, Schreiber, Lauriola & Gaeth, 2002) across different price levels and product categories (Park, Jun & Macinnis, 2000). Jin et al. (2012) did a study on the effect of option framing on tourists’ choice of package-tour services. Here, participants could up- or downgrade their packages. Also, participants could choose if they wanted to begin from the premium, higher priced package or the basic, lower priced package. The upgrading frame is preferred, as consumers feel more in control and it sparks positive feelings. The outcome is similar to other studies on option framing. Customers end up with a significantly higher price in the downgrading option frame, where they start at a premium package, than in the
upgrading option frame, where the anchor is a basic package. Furthermore, customers in the downgrading condition change less items than the customers in the upgrading condition. Moreover, comparing the effects of option framing on peripheral versus core services, option framing works better on the former than on the latter (Jin et al., 2012). Similar results were found by Xiaoyu et al. (2015) with the addition of importance of specific options for a consumer. Less important options are more likely to be selected in the subtracting option framing mode, while more important options are more likely to be chosen in the adding option framing mode. Overall, according to Xiaoyu et al. (2015) less important options are assumed to magnify the effect of option framing. While previous studies investigated on the influence of monetary aspects, cognitive constraints, commitment to the purchase and involvement on framing effects, the present study focuses on the influence of consumer decision-making styles on attitudes towards a decision-making process and the resulting purchase intention.
Purchase Intention. Both marketing methods, the interactive and the non-interactive method, offer trust and assurance as they help customers with the search of products and therefore assist them in deciding. The easier it is for a consumer to decide, the higher may be the purchase intention. Woo and Kim (2019) define purchase intention as “predicted or planned actions in the future, which is the likelihood of predisposition to turning beliefs and attitude toward a product into action” (p. 324). Van Noort et al. (2012) found that higher perceived levels of interactivity lead to a more intense flow experience which resulted in more positive affective, cognitive and behavioral responses. Although, the effect on cognitive responses was a little smaller than for affective or behavioral responses and occurred only for product-related thoughts and not for website-related thoughts. Consequentially, the following hypotheses occur:
H1a: People in the Subtracting frame will report a higher Purchase intention than people in the Adding Frame or people in the Non-interactive Frame.
H1b: People in the Subtracting frame will end up with a higher Number of toppings than people in the Adding frame or people in the Non-interactive frame.
Attitudes. When mentioning attitudes, the “summary evaluation of a psychological object” (Ajzen, 2001, p.28) is meant. There are supposed to be attribute dimensions, such as good-bad, harmful-beneficial, pleasant-unpleasant and likeable-dislikeable (Ajzen, 2001). A large number of studies is based on the theory of planned behavior (Ajzen, 1991) and proof that attitudes predict people’s behavioral intentions, but its reasons will not be discussed further as it would exceed this study’s scope. One underlying construct of the interactivity effects of websites is consumer’s flow experience (van Noort et al. 2012), which might influence consumers’ attitudes. Consumers’ flow experience is a multidimensional construct, whereof one dimension is perceived control. While one consumer feels to be more in control in the interactive conditions, another consumer may feel more in control in the non-interactive condition, which may lead to different attitudes across the marketing methods, depending on the consumer type. Devaraj, Fan and Kohli (2002) assessed attitudes with satisfaction with an e-commerce channel and found that ease of use and perceived usefulness are indicators for a customer’s attitude towards a channel. Besides, people’s feelings about the decision-making process vary (Iyengar & Lepper, 2000). People, who have a lot of options to choose from, enjoy the process more, but are also frustrated and find it more difficult. Only one of the studies on framing effects, that will be discussed further, looked into people’s reflections on a decision-making process and the final decision. Positive or negative reflections may vary between consumer’s decision-making styles and influence their attitudes. This leads to the following hypotheses:
H2a: The more positive one’s Attitudes towards the final decision and the decision-making process are, the higher is one’s reported Purchase intention.
H2b: The more positive one’s Attitudes towards the final decision and the decision-making process are, the higher is the Number of toppings chosen by a participant.
The concept of Consumer Decision-Making Styles
There are several approaches to categorize consumer’s decision-making styles. The present study has a closer look on the customer types, as Lysonski and Durvasula (2013) did, who created the two groups of utilitarian and hedonic customers, based on the consumer style inventory (CSI) by Sproles and Kendall (1986). The CSI is defined as “a mental orientation characterizing a consumer’s approach to making choices. It has cognitive and affective characteristics (for example, quality consciousness and fashion consciousness). In essence it is a basic consumer personality, analogous to the concept of personality in psychology” (Sproles & Kendall, 1986, p. 268). The related psychographics for utilitarian customers are information attainment, assortment seeking and time pressure. For the hedonic customer type, the related psychographics are shopping enjoyment, shopping escapism and the motivation to confirm or opinion seeking (Barwitz & Maas, 2018). Gehrt et al. (2012) identified two main shopping styles for customers online, the “quality at any price” and the
“reputation/recreation” customer, highly resembling the utilitarian and the hedonic customer style, as the former is characterized with a focus on quality and value of a purchase, whereas the latter gains enjoyment from the process of shopping. Another approach to categorize consumer styles is the cognitive-experiential self-theory (CEST) (Epstein, Pacini, Denes-Raj & Heier, 1996). According to this theory, there are two independent processing systems that operate in different rules, the experiential and the rational system. In this classification the
experiential system resembles the hedonic decision-making style and the cognitive system resembles the utilitarian decision-making style. Biswas (2009) analyzed not only the influence of option frames on consumers, but also their interaction with consumer’s processing modes and found that effects of option frames are influenced by consumer’s decision-making modes. People choose a higher number of items in a delete frame than in an add frame and this effect is even stronger for people making their decisions in an experiential than in a rational mode. The different, previously described frames in this study are assumed to be more or less appealing to either one of the two decision-making styles, which leads to more positive or negative attitudes towards the decision-making process and further influences consumer’s purchase intention. For a visual representation of the concept see figure 1. The following hypotheses arise:
H3a: People in the Adding frame will have more positive Attitudes towards the decision-making process than people in the Subtracting frame or people in the Non-interactive frame and this effect will be stronger the more hedonic one’s Decision-making style is.
H3b: People in the Non-interactive frame will have more positive Attitudes towards the decision-making process than people in the Interactive frames and this effect will be stronger the more utilitarian one’s Decision-making style is.
H4: Consumers in the Non-interactive frame will have more positive Attitudes towards their final decision and the decision-making process and therefore a higher Purchase intention than consumers in the Subtracting frame or the Adding frame and this effect will be stronger for people with a more utilitarian Decision-making style than for people with a more hedonic Decision-making style.
Figure 1: The conceptual model of the effects of different frames on Purchase Intention
mediated by Attitudes and moderated by Consumer Decision-Making Style
Method Experimental design and framework
The experimental design for this quantitative research has three conditions (Marketing Method: non-interactive vs. adding vs. subtracting), a mediator (Attitudes from 1-7) and two moderators (Decision-making style: Utilitarian and Hedonic). The main factor in the current research design is an experimental factor and a between-subjects variable, namely the Online content marketing method. The levels of the main factor are nominal (non-interactive vs. adding vs. subtracting). The mediator are consumer’s Attitudes ranging from more negative (1) to more positive (7). The moderators are the continuous variables more Utilitarian and more Hedonic Decision-making style. The Decision-making style cannot be manipulated and is therefore a quasi-experimental factor.
Product chosen for this study. The product that is used for the experiment is pizza, inspired by a study by Levin et al. (2002). Reasons to use pizza are, that independently from people’s age, gender or income, participants similarly relate to the situation of choosing a pizza. Additionally, ingredients for a pizza can easily be mixed and matched to ensure realistic options in all conditions.
Conditions. There are three experimental conditions that participants are randomly assigned to, one interactive and two interactive content marketing conditions. The non-interactive content marketing condition consists of a menu for pizza. In the first non-interactive condition participants are given the base of a pizza, the dough, and have the possibility to add up to six ingredients. In the second interactive condition participants have the possibility to subtract ingredients from a fully loaded pizza. The available ingredients in all conditions are the same. The pizzas to choose from in the non-interactive condition are composed by different combinations of the same ingredients that can be added or subtracted in the interactive conditions. For the instructions of the experiment see Appendix B.
Measurement of the mediator. Consumer’s attitudes towards their decision for a specific pizza are measured with two scales. First, the decision outcome attitudes (Decision-Attitudes) are measured with a three-item scale derived from Jin et al. (2012). The second scale measures the attitudes towards the decision-making process (Process-Attitudes) and consists of seven items derived from several scales to cover the features that are mentioned above to be characteristics for the evaluation of interactivity by previous studies (Jin et al., 2012; Bigné-Alcañiz, Ruiz-Mafé, Aldás-Manzano & Sanz-Blas, 2008; van Noort et al., 2012). For all measurements, participants are asked to indicate how much they agree with a
statement on a seven-point answering-scale ranging from 1 (strongly disagree) to 7 (strongly agree).
Measurement of the moderators. For the moderating variable, customer decision-making style, participants are asked questions about their purchasing behavior before they are
exposed to the manipulation to make sure to measure their initial decision-making style which has not been influenced by the stimulus material. As Lysonski and Durvasula (2013)
categorized the eight decision-making styles by Sproles and Kendall (1986) into two main groups (utilitarian and hedonic), the most extreme shopping style out of each of these two categories is chosen. For the utilitarian consumer the questions of the “confused by overchoice consumer” are adapted for the present study. For the hedonic consumers the questions for the “recreational – hedonistic consumer” are adapted. In previous studies, these items also had the highest factor loadings (Prakash, Singh & Yadav, 2018) or the highest mean score (Lysonski & Durvasula, 2013) for the consumer decision-making styles. For all the questions see Appendix A.
Dependent Variable. The dependent variable in this study is Purchase intention of a selected product. People’s reported Purchase intention is measured with four items (“I would consider purchasing the pizza I chose”) on a seven-point scale derived from a scale by Barber, Kuo, Bishop and Goodman Jr (2012) (see Appendix A). Besides that, consumer’s Purchase intention is measured by the number of toppings they select.
Procedure. After a pretest, the online experiment was active from November 26th, 2019 until December 8th, 2019. It was distributed via social media to reach people in the researcher’s personal environment. Participants could take part via smartphone, tablet or laptop. After agreeing to the privacy statement, participants could proceed with the survey.
Sample. The total number of participants is 156 (N = 156). Participant’s age ranges from 18 to 79 years (M = 27.69, SD = 9.933). 32.1% of the participants are male (N = 50), 67.3% are female (N = 105) and one person prefers not to say (N = 1). Most participant’s highest degree with 35.3% is a bachelor’s degree (N = 55), followed by 30.8% of participants with a Master’s degree (N = 48) and 24.4% with a High school degree (N = 38). Participant’s current country of residence is predominantly Austria (N = 86), which is indicated by 55.1% of the participants. 19.9% of the participants indicate that the Netherlands are their current
country of residence (N = 31). 19.9% indicate another European country as their current country of residence (N = 31) and the remaining eight participants reside in other parts of the world (N = 8). Participants are equally distributed across the conditions. 34.6% are in the non-interactive condition (N = 54), 32.7% are in each case in the adding-frame and in the
subtracting frame (N = 51).
Data Analysis Data Cleaning and Preparation
First, pretest-data (N = 18) and unfinished cases (N = 50) were excluded from the dataset. Then, missing data was identified. One item concerning consumer decision-making styles has been reverse coded on a seven-point Likert scale.
Construction of variables
A new variable has been computed to identify to which condition each participant has been randomly assigned to. Next, dummy variables were constructed for the adding and the subtracting-frame as well as an overall dummy for the interactive frames to be able to
compare it to the non-interactive frame in following regression and PROCESS analyses. As a result of combining the interactive frames and creating one dummy variable “Interactive”, that has been used in the PROCESS analyses, eventual differences occur or are lost in the Table 1 Constructed variables Variable M SD Eigenvalue factor % of Variance Purchase Intention 5.18 1.55 .92 3.21 80.1 Attitudes Process 4.73 1.38 .91 5.52 55.2 Attitudes Decision 4.96 1.55 .90 1.60 16.0 Utilitarian Consumer 5.87 1.12 .78 1.88 62.7 Hedonic Consumer 4.61 1.19 .64 1.93 48.2
PROCESS analyses. PROCESS analyses have additionally been conducted to only compare the two interactive frames (see Appendix C, table 11-14). Next, a new variable has been computed to identify how many toppings people chose (M = 3.57, SD = 0.96). New variables have also been created for interaction effects in hypotheses H3a and b.
To perform analyses including people’s Purchase intention, Process-Attitudes, Decision-Attitudes and Consumer’s Utilitarian as well as Hedonic Decision-making style, new variables were constructed. For each variable, a factor analysis with Direct Oblimin rotation was conducted. After a reliability analysis was conducted, the mean score out of the related items was calculated. The minimum score for each variable is 1 and the maximum score is 7. People with lower scores have a smaller Purchase intention, more negative Attitudes or are less of a Utilitarian or Hedonic shopper than people with a higher score. Table 1 gives an overview of the new variables.
As assumed, the factor analysis for “Attitudes” revealed two factors, one factor concerning the “Attitudes towards the decision-making process” and one factor regarding the “Attitudes towards the final decision”. Two new variables were constructed. For the new variable “Utilitarian decision-making style” the initial Cronbach’s alpha of .65 could be substantially improved by removing one item. The new variable was constructed out of two items. To check for normal distribution, skewness and kurtosis were inspected for the new variables. Overall, the values were acceptable as can be seen in table 2.
Table 2
Skewness and Kurtosis for Condition, Number of Toppings in condition 1, 2 & 3, Attitudes towards the decision-making process, Attitudes towards the final decision, Purchase intention, Utilitarian consumer style and Hedonic consumer style
Variables Descriptive Statistics
N Skewness Kurtosis Condition 156 .04 -1.52 Nr. Toppings Con1 54 .02 -.43 Nr. Toppings Con2 51 .47 -.58 Nr. Toppings Con3 51 -.06 -.95 Attitudes Process 156 -.58 -.16 Attitudes Decision 156 -.77 -.07 Purchase Intention 156 -1.01 .42
Results
To test Hypothesis H1a and H1b a one-way ANOVA was conducted. The independent variable is the Condition (adding frame vs. subtracting frame vs. non-interactive frame) and the dependent variable is the Purchase intention (1 = lower - 7 = higher Purchase intention) for hypothesis H1a and the Number of toppings (2-6) for hypothesis H1b.
For hypothesis H1a the effect of the Condition on the reported Purchase intention was not significant and the effect size was small, F (2, 153) = 0.43, p = .650, 2 = .006. Post-hoc Bonferroni tests showed that the Purchase intention is slightly lower in the non-interactive condition than in the adding frame condition (Mdiff = -0.22, SD = 0.32) as well as in the subtracting frame condition (Mdiff = -0.28, SD = 0.32) and it is equally high in the adding frame condition and the subtracting frame condition (Mdiff = -0.05, SD = 0.32). That means that hypothesis H1a is not supported. There is no difference in the effects of Condition on reported Purchase intention.
To test hypothesis H1b the effect of Condition on Number of toppings chosen was significant, but the effect size was small, F (2, 153) = 8.37, p = .001, 2 = .099. Post-hoc Bonferroni tests showed that the Number of toppings is lower in the adding frame than in the non-interactive frame (Mdiff = -0.64, SD = 0.18) as well as in the subtracting frame (Mdiff = -0.65, SD = 0.18). There is almost no difference in the Number of toppings between people in the subtracting and the non-interactive frame (Mdiff = 0.01, SD = 0.18) That means that
hypothesis H1b is partly supported. People in the subtracting frame choose significantly more toppings than people in the adding frame, but the same amount as people in the
non-interactive frame.
Next, two simple regressions are conducted to test the direct effect of Attitudes towards decision-making process on Purchase intention. To test hypothesis H2a two simple regressions have been conducted with Process-Attitudes and Decision-Attitudes as
Utilitarian consumer 156 -1.14 1.76
independent variables and reported Purchase intention as dependent variable. The first regression model is statistically significant, F (1, 154) = 91.55, p < .001, R2 = .373. Process-Attitudes do have a significant positive effect on reported Purchase intention, b = .71, b* = .61, t = 9.57, p < .001, 95% CI [.57, 86]. The second regression model is also significant, F (1, 154) = 330.40, p < .001, R2 = .682. Decision-Attitudes positively predict the reported Purchase intention, b = 0.86, t = 18.18, p < .001, 95% CI [0.77, 0.95]. A one-unit increase of Decision-Attitudes makes people more likely to purchase the selected pizza. This effect is statistically strong, b* = .83. Hypothesis H2a is supported. The more positive someone’s Process-Attitudes and Decision-Attitudes are, the higher is this person’s Purchase intention.
To test hypothesis H2b two simple regression analysis have been conducted with Process-Attitudes and Decision-Attitudes as independent variables and Number of toppings as dependent variable. The first regression model is not statistically significant, F (1, 154) = 1.00, p = .320, R2 = .006. Process-Attitudes do not significantly predict the Number of toppings a participant chooses, b = -0.06, b* = -.08, t = -1.00, p = .320, 95% CI [–.17, .06]. The second regression model is not significant either, F (1, 154) = 1.09, p = .298, R2 = .007. Decision-Attitudes do not predict the Number of toppings, b = .05, b* = .08, t = 1.04, p = .298, 95% CI [-.05, .15]. This means that hypothesis H2b is not supported. Process-Attitudes and Decision-Attitudes do not predict the Number of toppings people put on their pizza.
Next, the effects of Condition (adding vs. subtracting vs. non-interactive) on Attitudes have been tested in two separate One-way ANOVAs. The model for the effect of Condition on Process-Attitudes is significant, F (2, 153) = 15.10, p < .001, 2 = .165. Post-hoc
Bonferroni tests showed that people in the non-interactive frame had significantly more negative Process-Attitudes than people in the adding (Mdiff = -1.21, SD = 0.25) and the subtracting frame (Mdiff = -1.14, SD = 0.25). The model for the effect of Condition on Decision-Attitudes is not significant, F (2, 153) = .22, p = 801, 2 = .003. Further results are displayed in table 3 in Appendix C.
To test hypothesis H3a two multiple regression analyses with a dummy variable for the adding frame have been conducted with Condition and the Hedonic Decision-making style as the independent variables and the Process-Attitudes as dependent variable. The first
multiple regression was conducted with excluding the non-interactive condition in order to compare the two interactive frames. The second multiple regression was conducted with excluding the subtracting condition in order to compare the adding frame to the non-interactive condition.
Table 4
Multiple Regression Analysis with Attitudes towards the decision-making process as dependent variable, comparing adding and subtracting frame
Variable B SE B* t p Constant 4.817 .779 6.183 .000 Adding -.431 1.000 -.187 -.431 .667 Hedonic .061 .163 .059 .374 .709 Adding*Hedonic .115 .212 .239 .543 .589 a. F (3, 98) = .66, p = .587, R2 = .02
The first regression model is not statistically significant and contains no significant effects, as displayed in table 4. The second regression model comparing the adding frame to the non-interactive frame is statistically significant. There is, however, no significant effect of Condition on Process-Attitudes. The results are displayed in table 5. This means that
Hypothesis H3a is not supported. People in the adding frame only have more positive Process-Attitudes than people in the non-interactive frame, but not than people in the subtracting frame. However, these effects are not stronger the more hedonic somebody’s Decision-making style is.
Table 5
Multiple Regression Analysis with Attitudes towards the decision-making process as dependent variable, comparing adding and non-interactive frame
Variable B SE B* t p Constant 2.797 .668 4.189 .000 Adding 1.589 0.973 -.547 1.632 .106 Hedonic .250 .137 .217 1.817 .072 Adding*Hedonic -.073 .206 -.121 -.357 .772 a. F (3, 101) = 8.917, p < .001, R2 = .21
To test hypothesis H3b a multiple regression has been conducted. The Condition and the Utilitarian decision-making style served as independent variables and Decision-Attitudes as the dependent variable. A dummy variable was used for the non-interactive frame (non-interactive = 1, else = 0) to compare it to the (non-interactive frames. The regression model is statistically not significant and explains 2.9% of the variance for Decision-Attitudes. There are no significant results regarding the effects of Condition and Utilitarian decision-making style as well as the interaction effect on Decision-Attitudes. Hypothesis H3b is not supported. People in the non-interactive frame do not have more positive Decision-Attitudes than people in the interactive frames. All the results are displayed in table 6.
Table 6
Multiple Regression Analysis with Attitudes towards the final decision as dependent variable, comparing the non-interactive frame to the interactive frames combined
Variable B SE B* t p Constant 4.402 .854 5.156 .000 Non-interactive -1.497 1.346 -.461 -1.112 .268 Utilitarian .103 .142 .074 .723 .471 Non-interactive*Utilitarian .236 .226 .434 1.042 .299 a. F (3, 152) = 1.51, p = .214, R2 = .03
Next, hypothesis H4, which explains the overall model, has been tested using the SPSS macro PROCESS v3.4 by Andrew F. Hayes (2013). To test the overall model, several analyses have been conducted using model 9 of the PROCESS macro. With these analyses of the moderated mediation all possible effects of Conditions on Attitudes moderated by
Decision-making style and further on Purchase intention are revealed. For the first analyses to exactly test hypothesis H4, the interactive frames were combined and compared to the non-interactive frame (X) using a dummy variable. The effects were separately tested for both Attitudes measured (towards decision-making process and towards final decision) as mediators (M) and also for reported Purchase intention as well as for Number of toppings a person chose as dependent variables (Y). In each case, the Utilitarian (W) and the Hedonic
(Z) Decision-making style were the moderators of the effect of Condition on Attitude. Next, to additionally test other possible effects, the interactive frames (adding and subtracting) have been compared to each other with the same moderators, mediators and outcome variables, by using a dummy variable for the subtracting frame.
Table 7 and 8 describe the mediating effect of Process-Attitudes when analyzing the effect of Condition (interactive vs. non-interactive) on reported Purchase Intention (table 7) and Number of toppings (table 8). Table 9 and 10 explain the mediating effect of Decision-Attitudes when analyzing the effect of Condition (interactive vs. non-interactive) on reported Purchase intention (table 9) and Number of toppings (table 10). Only in table 7 there are two significant results predicting the Purchase intention, implying a mediation. There is a
suppression effect. The significant effect of Interactive frame on Purchase intention only occurs when adding the mediator variable Process-Attitudes. Previous analyses showed that people in the interactive frame have significantly more positive Process-Attitudes than people in the non-interactive frame. This effect is not visible in table 7 as there are the moderators added. Hence, there is a mediating effect of Process-Attitudes on the relationship between Condition and Purchase intention, but this effect is not moderated by Utilitarian or Hedonic Decision-making style. The other analyses do not show mediating effects, which means that Process-Attitudes do not mediate the relation between Condition on Number of toppings, neither do Decision-Attitudes influence any relationship between Condition and Purchase intention or Number of toppings. Tables 8 to 10 are in Appendix C. It therefore follows that hypothesis H4 is not supported. People in the non-interactive frame do not have more positive Decision-Attitudes and have more negative Process-Attitudes, than people in the interactive frames and these effects are not stronger the more utilitarian their Decision-making style is. Furthermore, their Purchase intention is not higher consequentially.
The same analyses have been conducted to analyze the mediating effect of Process-Attitudes on the effect of Condition (adding vs. subtracting) on Purchase intention (table 11) and Number of toppings (table 12) as well as of Decision-Attitudes on the effect of Condition (adding vs. subtracting) on Purchase intention (table 13) and Number of toppings (table 14). There are no significant mediating or moderating effects. Tables 11 to 14 explaining the mediating effect of Attitudes on Purchase intention and Number of toppings are in Appendix C.
Discussion
This study provides some interesting results and new perspectives on the interactivity and framing literature. While different frames influence the number of toppings a person chooses, attitudes towards the final decision and the decision-making process influence the purchase intention. Besides, an interactive frame leads to more positive attitudes towards the decision-making process. The shopping style does not influence people’s attitudes.
An online experiment was distributed to a broad sample to reach every possible customer of an online shop. The task in the experiment was to order pizza in one of three conditions, that participants have been randomly assigned to, inspired by a study by Levin et al. (2002). An evaluation of the questionnaire reveals that participants’ reported purchase Table 7
PROCESS analysis to test the mediating effect of Attitudes towards decision-making process on Purchase Intention with Conditions (Interactive vs. Non-interactive) as independent variable and Decision-making styles as moderators
Variable Coefficient SE t p Moderation Constant 2.440 .996 2.450 .015 Interactive 1.597 1.290 1.238 .218 Utilitarian .067 .144 .467 .641 Interactive*Utilitarian .031 .186 .166 .869 Hedonic .242 .132 1.831 .069 Interactive*Hedonic -.128 174 -.735 .463 Mediation Constant 1.804 .357 5.050 .001 Interactive -.703 .230 -3.061 .003 Attitude process .811 .079 10.222 .001
intention is overall rather high, while attitudes are rather neutral to slightly positive. By categorizing consumer decision-making styles in two groups, as done by Lysonski and Durvasula (2013), the present sample consists of rather utilitarian than hedonic shoppers. Table 15 gives an overview of the hypotheses being supported or not.
People in the subtracting frame do not report a higher purchase intention than people in the adding or the non-interactive frame. Which means that the participants are all similarly likely to buy their chosen product. Previous studies in the framing literature measured the quantity of chosen products to identify differences among framing options, as done for hypothesis H1b. As this study additionally compares a non-interactive frame to the adding and subtracting frames, it is regarded interesting to see if people’s likelihood to consider buying a pizza differs. The result is surprisingly insignificant. Literature on choice overload suggests that too many choices influence people’s likelihood to purchase a chosen product (Nagar & Gandotra, 2016; Iyengar & Lepper, 2000). Furthermore, the options that were Table 15
Hypothesis overview
H1a People in the Subtracting frame will report a higher Purchase intention than
people in the Adding frame or people in the Non-interactive frame.
Not supported
H1b People in the Subtracting frame will end up with a higher Number of toppings
than people in the Adding frame or people in the Non-interactive frame.
Partly supported
H2a The more positive one’s Attitudes towards the final decision and the
decision-making process are, the higher is one’s reported Purchase intention. Supported
H2b
The more positive one’s Attitudes towards the final decision and the decision-making process are, the higher is the Number of toppings chosen by a
participant.
Not supported
H3a
People in the Adding frame will have more positive Attitudes towards the decision-making process than people in the Subtracting frame or people in the Non-interactive frame and this effect will be stronger the more hedonic one’s Decision-making style is.
Not supported
H3b
People in the Non-interactive frame will have more positive Attitudes towards their final decision than people in the Interactive frames and this effect will be stronger the more utilitarian one’s Decision-making style is.
Not supported
H4
Consumers in the Non-interactive frame will have more positive Attitudes towards their final decision and the decision-making process and therefore a higher Purchase intention than consumers in the Subtracting frame or the Adding frame and this effect will be stronger for people with a more utilitarian Decision-making style than for people with a more hedonic Decision-making style.
Not supported
available to be selected in the non-interactive condition were highly similar to each other, which is another indicator for choice overload (Yun & Duff, 2017) and led to the assumption that people would be less likely to intend purchasing a product. The result of hypothesis H1a can be explained by a conclusion by Iyengar and Lepper (2000). Even though fewer options lead to a higher purchasing behavior, more options also mean more enjoyment as well as frustration. Besides, people are still able to decide with a high information load. It only gets harder to choose one product (Korhonen, Malo, Pajala, Ravaja, Somervuori & Wallenius, 2018). Once a product is chosen, people might as well buy it, regardless how hard the decision was.
Hypothesis H1b is partly supported and is in line with the framing literature. People in the subtracting frame end up with a higher number of toppings than people in the adding frame. Compared to the non-interactive frame, the difference to the subtracting frame is negligible. The significant difference between adding and subtracting frame is attributable to the anchoring and adjustment model (Wilson et al., 1996) as well as to the loss-aversion model (Biswas & Grau, 2008). Following these models, a higher number of toppings was expected in the subtracting frame. Indeed, participants end up with over half a topping more in the subtracting frame than in the adding frame, which is a lot keeping in mind that
participants could choose on a range from two to six toppings. The difference may be that high because people do not consider this decision as very important, which has an additional magnifying effect on the difference between adding and subtracting frame (Xiaoyu et al., 2015). The fact that the non-interactive condition did resemble the subtracting frame more than the adding frame might be due to the overview of all possible options. This overview possibly makes participants wanting an option with more toppings as a pizza with five toppings sounds more exciting than a pizza with two toppings. But that is an assumption, actual causes for the indifference between the subtracting frame and the non-interactive frame may be interesting for future research using a different product or product category.
Attitudes and their effect were analyzed, due to their proven prediction of people’s behavioral intentions, according to the theory of planned behavior (Ajzen, 1991). Attitudes towards the final decision were analyzed on a pre-existing scale by Jin et al. (2012). The scale to analyze process-attitudes was constructed by using existing items used in previous research (Jin et al., 2012; Bigné-Alcañiz et al., 2008; van Noort et al., 2012). Overall, process-attitudes and decision-attitudes influence participants’ reported purchase intention. People who have more positive attitudes towards their chosen pizza and the decision-making process intend more likely to buy the pizza.
To analyze hypothesis H2b the effect of attitudes on the number of toppings a person chose was evaluated. Attitudes do not influence the number of toppings somebody chooses. This means that solely the frames influence how many toppings people put on their pizza.
Furthermore, it was expected for hypothesis H3a that people in the adding frame have more positive attitudes towards the decision-making process than people in the other
conditions and this effect was expected to be even stronger the more hedonic a person’s decision-making style was. People in the adding frame do have more positive attitudes towards the decision-making process than people in the non-interactive frame, but not than people in the subtracting frame. These effects are not stronger the more hedonic people’s decision-making style is. This result is in line with theory. Ease of use (Doern & Fey, 2006), ease of search (Doern & Fey, 2006) and perceived control (van Noort et al., 2012) are indicators for more positive process-attitudes and are more likely to occur for people in the interactive frames. A significant difference between attitudes in the interactive frames was expected, because previous research showed that people sympathize more with the adding frame (Jin et al. 2012). A reason why especially people in the non-interactive frame have more negative process-attitudes than people in the interactive frames may be the possibility to choose from more than 22 options, namely 57 in the non-interactive frame. More than 22
options lead to more perceived regret (Park & Jang, 2013), which was assumed to lead to more negative attitudes due to choice overload.
For hypothesis H3b, the effect of the conditions and the influence of consumer decision-making styles was analyzed with decision-attitudes as the outcome. No significant results have been found to support this hypothesis when comparing all frames. People in the non-interactive frame do not have significantly more positive decision-attitudes than people in the interactive frames. This effect does not show any significant difference between the two decision-making styles. Decision-attitudes were expected to be more positive in the non-interactive frame. After going through the hassle of selecting one pizza out of an enormous amount of variations, instead of simply clicking on the toppings they wanted, people were assumed to have strong positive Decision-Attitudes. This effect was expected to be stronger for people in the non-interactive frame characterized as utilitarian consumers. People with this consumer decision-making style are more quality and price conscious (Gehrt et al., 2012) and quickly looking for information and a bigger assortment (Barwitz & Maas, 2018), while hedonic consumers enjoy the process of shopping, looking for interaction (Barwitz & Maas, 2018) and recreation (Gehrt et al., 2012). Therefore, the non-interactive frame was giving all the information on the pizza combinations of toppings and doughs on one view, which should have pleased people in the non-interactive frame more the more utilitarian their decision-making style was. Whereas, the process of adding or deleting toppings and choosing a dough in an extra step should have more appealed to people the more hedonic their shopping style was, but these assumptions do not apply to the present study. A reason for that may be the product, people were selecting. People are equally as happy to have found their preferred pizza and are similarly looking forward to it, regardless their decision-making style. Managerial recommendations
For smaller companies these results show, that they can add value to their website by implementing a cost-efficient interactive element in form of a subtracting frame. This frame
positively influences consumer’s attitudes and leads to a slightly higher number of toppings consumers are choosing. The adding frame, in turn, should not be used as an interactive element, if the goal is to make customers select more options on a website.
Conclusion
The following paragraph will concisely answer the research question “To what extent does Interactive content marketing positively influence consumers’ Attitudes and
consequentially consumers’ Purchase intention on a website and is this effect moderated by Consumer decision-making styles?”. After comparing an adding, a subtracting and a non-interactive frame, the conclusion can be drawn that interactivity in the form of customizing a product leads to more positive attitudes towards the decision-making process, but not towards the final decision and does not influence people’s purchase intention. In each frame, people are similarly likely to purchase their selected pizza. Cleverly executed, in the form of a subtracting frame, interactivity leads people to select slightly more toppings for their pizza than a full menu. An adding frame, in turn, leads people to put less toppings on their pizza. For marketers this is interesting if they strive to avoid choice overload by presenting a full menu or if they want to stand out with their website by adding an interactive element to it. Furthermore, people in the interactive frames have more positive attitudes towards the
decision-making process than people confronted with the full menu. And attitudes towards the decision-making process as well as towards the final decision, lead to a higher purchase intention. Therefore, online food ordering platforms are well advised to set up the decision-making process by presenting a full pizza, of which people will barely remove items from. Customers enjoy this interactive selection process more and end up with more toppings on their pizza. Adding an interactive element for the decision-making process in form of a subtracting frame is an effective way to gain happier customers and to sell more at once, especially for companies on a smaller budget. The influence of consumer decision-making style does not play a determining role in the effect of frames on attitudes. Future studies may
analyze their effect on the direct effect of framing on the number of options a person chooses. Besides, marketers can be reassured. They do not have to consider their customer’s decision-making style when choosing the right frame for their website, because people’s attitudes do not differ across the customers’ decision-making styles.
Limitations and future research
To ensure internal validity, the conditions were exactly the same with the only difference in the instructions to add or delete toppings or to choose one out of 57 options. First, this high number ensures that people are confronted with choice overload. Second, this is the amount of combinations people can create with six different toppings and the only instruction to either select or keep two to six toppings in the interactive frames. That means that participants were able to select one out of 57 possible combinations in each condition. The ecological validity, therefore, falls short a little. Usually, people are not confronted with these many options for pizza, that are at the same time as similar to each other. Furthermore, apart from the instruction that they are going to order pizza from a restaurant they liked, participants did not get in touch with other elements that reminded them of a typical online pizza delivery service, like a name of the restaurant, an appealing website layout or images. The focus in this study laid solely on the effects of the framing method to choose the preferred pizza. Adding a photo for each ingredient or seeing the finally selected pizza may lead to more positive attitudes and strengthen the mediation effect of attitudes on purchase intention or the number of toppings someone chooses. For future research it may be interesting to add these elements.
To avoid that participants choose a pizza with more toppings just to get more for the same price, a price was added to each pizza. Each ingredient costed one euro. The monetary aspect was not further investigated in this study, as previous research did not find any remarkable influence of that aspect (Biswas & Grau, 2008; Park et al. 2000) and the price differences are the same across the conditions. However, these studies did not have a control
group like the non-interactive group in this study. Maybe the monetary aspect influenced this group as they saw the price for each combination right away and did not have to calculate it like people in the other two groups. Future research may consider analyzing the monetary effect in combination with analyzing consumer’s cognitive constraints, which may influence consumer decision according to the CEST (Epstein et al., 1996).
One last limitation at this point is that people’s frequency of online food orders has not been measured. People who order food online regularly may have more positive attitudes towards the decision-making process than people who barely order food online when they are in an interactive condition, as these conditions are a way for online-stores to differ from other online-stores (Ganesh et al. 2010) and may therefore be a welcome change for frequent users of online food ordering websites. Or the other way around. As frequent users are used to a long menu, they might not be interested in more interactive methods to order their food. Another interesting aspect for future studies.
References
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human
Decision Processes, 50(2), 179–211. https://doi.org/10.1016/0749-5978(91)90020-T
Ajzen, I. (2001). Nature and operation of attitudes. Annual Review of Psychology, 52, 27–58. Barber, N., Kuo, P., Bishop, M., & Goodman, R. (2012). Measuring psychographics to assess
purchase intention and willingness to pay. Journal of Consumer Marketing, 29(4), 280–292. https://doi.org/10.1108/07363761211237353
Barwitz, N., & Maas, P. (2018). Understanding the Omnichannel Customer Journey: Determinants of Interaction Choice. Journal of Interactive Marketing, 43, 116–133. https://doi.org/10.1016/j.intmar.2018.02.001
Bigné-Alcañiz, E., Ruiz-Mafé, C., Aldás-Manzano, J., & Sanz-Blas, S. (2008). Influence of online shopping information dependency and innovativeness on internet shopping
adoption. Online Information Review, 32(5), 648–667. https://doi.org/10.1108/14684520810914025
Biswas, D. (2009). The effects of option framing on consumer choices: Making decisions in rational versus experiential processing modes. Journal of Consumer Behaviour, 8(5), 284–299. https://doi.org/10.1002/cb.288
Biswas, D., & Grau, S. (2008). Consumer choices under product option framing: Loss aversion principles or sensitivity to price differentials? Psychology and
Marketing, 25(5), 399–415. https://doi.org/10.1002/mar.20217
Chernev, A., Böckenholt, U., & Goodman, J. (2015). [Review of Choice overload: A
conceptual review and meta-analysis]. Journal of Consumer Psychology, 25(2), 333–
358. https://doi.org/10.1016/j.jcps.2014.08.002
Devaraj, S., Fan, M., & Kohli, R. (2002). Antecedents of B2C Channel Satisfaction and Preference: Validating e-Commerce Metrics. Information Systems Research, 13(3), 316–333. doi: 10.1287/isre.13.3.316.77
Doern, R., & Fey, C. (2006). E-commerce developments and strategies for value creation: The case of Russia. Journal of World Business, 41(4), 315–327.
https://doi.org/10.1016/j.jwb.2006.08.001
Epstein, S., Pacini, R., Denes-Raj, V., & Heier, H. (1996). Individual Differences in Intuitive– Experiential and Analytical–Rational Thinking Styles. Journal of Personality and
Social Psychology, 71(2), 390–405. https://doi.org/10.1037/0022-3514.71.2.390
Ganesh, J., Reynolds, K., Luckett, M., & Pomirleanu, N. (2010). Online Shopper Motivations, and e-Store Attributes: An Examination of Online Patronage Behavior and Shopper Typologies. Journal of Retailing, 86(1), 106–115.
https://doi.org/10.1016/j.jretai.2010.01.003
Gehrt, K., Rajan, M., Shainesh, G., Czerwinski, D., & O’Brien, M. (2012). Emergence of online shopping in India: shopping orientation segments. International Journal of
Retail & Distribution Management, 40(10), 742–758.
https://doi.org/10.1108/09590551211263164
Hayes, A. F. (2013). Introduction to Mediation, Moderation, and Conditional Process Analysis. New York: Guilford Publications.
Iyengar, S., & Lepper, M. (2000). When Choice is Demotivating: Can One Desire Too Much of a Good Thing? Journal of Personality and Social Psychology, 79(6), 995–1006. https://doi.org/10.1037/0022-3514.79.6.995
Jin, L., He, Y., & Song, H. (2012). Service customization: To upgrade or to downgrade? An investigation of how option framing affects tourists’ choice of package-tour services. Tourism Management, 33(2), 266–275. https://doi.org/10.1016/j.tourman.2011.03.005 Khare, A., Khare, A., Mukherjee, S., & Goyal, T. (2016). Do consumer shopping styles
influence consumer attitudes toward services offered by shopping websites? Journal
of International Consumer Marketing, 28(1), 28–41.
https://doi.org/10.1080/08961530.2015.1072075
Korhonen, P., Malo, P., Pajala, T., Ravaja, N., Somervuori, O., & Wallenius, J. (2018). Context matters: The impact of product type, emotional attachment and information overload on choice quality. European Journal of Operational Research, 264(1), 270– 279. https://doi.org/10.1016/j.ejor.2017.06.060
Levin, I., Schreiber, J., Lauriola, M., & Gaeth, G. (2002). A Tale of Two Pizzas: Building Up from a Basic Product Versus Scaling Down from a Fully-Loaded Product. Marketing
Letters, 13(4), 335–344. https://doi.org/10.1023/A:1020370516638
Lysonski, S., & Durvasula, S. (2013). Consumer decision making styles in retailing: evolution of mindsets and psychological impacts. Journal of Consumer Marketing, 30(1), 75– 87. https://doi.org/10.1108/07363761311290858
Nagar, K., & Gandotra, P. (2016). Exploring Choice Overload, Internet Shopping Anxiety, Variety Seeking and Online Shopping Adoption Relationship: Evidence from Online
Fashion Stores. Global Business Review, 17(4), 851–869. https://doi.org/10.1177/0972150916645682
Park, J.-Y., & Jang, S. (2013). Confused by too many choices? Choice overload in tourism. Tourism Management, 35, 1–12.
https://doi.org/10.1016/j.tourman.2012.05.004
Park, C., Jun, S., & Macinnis, D. (2000). Choosing What I Want Versus Rejecting What I Do Not Want: An Application of Decision Framing to Product Option Choice
Decisions. Journal of Marketing Research, 37(2), 187–202. https://doi.org/10.1509/jmkr.37.2.187.18731
Pilli, L., & Mazzon, J. (2016). Information overload, choice deferral, and moderating role of need for cognition: Empirical evidence. Revista de Administração (São Paulo), 51(1), 36–55. https://doi.org/10.5700/rausp1222
Prakash, G., Singh, P., & Yadav, R. (2018). Application of consumer style inventory (CSI) to predict young Indian consumer’s intention to purchase organic food products. Food
Quality and Preference, 68, 90–97. https://doi.org/10.1016/j.foodqual.2018.01.015
Singh, M. (2002). E-services and their role in B2C e-commerce. Managing Service Quality:
An International Journal, 12(6), 434–446.
https://doi.org/10.1108/09604520210451911
Sproles, G., & Kendall, E. (1986). A Methodology for Profiling Consumers’ Decision‐ Making Styles. Journal of Consumer Affairs, 20(2), 267–279.
https://doi.org/10.1111/j.1745-6606.1986.tb00382.x
Tversky, A., & Kahneman, D. (1981). The framing of decisions and the psychology of choice. Science (New York, N.Y.), 211(4481), 453–458.
https://doi.org/10.1126/science.7455683
van Noort, G., Voorveld, H., & van Reijmersdal, E. (2012). Interactivity in Brand Web Sites: Cognitive, Affective, and Behavioral Responses Explained by Consumers’ Online
Flow Experience. Journal of Interactive Marketing, 26(4), 223–234. https://doi.org/10.1016/j.intmar.2011.11.002
Voorveld, H., Neijens, P., & Smit, E. (2011). The Relat ion Between Actual and Perceived Interactivity. Journal of Advertising, 40(2), 77–92. https://doi.org/10.2753/JOA0091-3367400206
Wilson, T., Houston, C., Etling, K., & Brekke, N. (1996). A New Look at Anchoring Effects: Basic Anchoring and Its Antecedents. Journal of Experimental Psychology:
General, 125(4), 387–402. https://doi.org/10.1037/0096-3445.125.4.387
Woo, E., & Kim, Y. (2019). Consumer attitudes and buying behavior for green food products.
British Food Journal, 121(2), 320–332. https://doi.org/10.1108/BFJ-01-2018-0027
Xiaoyu, W., Qiao, L., Jifei, W., & Yi, L. (2015). The effect of option framing on consumers decision-making efficiency: Evidence from online service customization. African
Journal of Business Management, 9(6), 298–310.
https://doi.org/10.5897/AJBM2014.7647
Yun, J., & Duff, B. (2017). Is utilizing themes an effective scheme? Choice overload and categorization effects within an extensive online choice environment. Computers in
Appendices Appendix A Questionnaire
Imagine, you are in the situation to order food online. How would you usually approach this purchase? Please answer the following statements on the answering scale from 1 (strongly disagree) to 7 (strongly agree).
1. Ordering food online is an enjoyable activity of my life (hedonic) 2. Getting very good quality is very important to me (utilitarian) 3. In general, I usually try to buy the best overall quality (utilitarian)
4. Sometimes it is hard to choose which restaurant to order from (utilitarian) 5. Ordering food online is not a pleasant activity to me* (hedonic)
6. To get variety, I order food at different restaurants (hedonic) 7. It is fun to order something new and exciting (hedonic) *reversecoded
Experiment
Now, that you have chosen your preferred pizza, think of the style in which the options for available pizza have been presented to you. Please answer the following statements on the scale from 1 (strongly disagree) to 7 (strongly agree).
1. I am very satisfied with the style the available options were presented (Jin, He & Song, 2012)*
2. This decision task is an enjoyable experience to me (Jin, He & Song, 2012)* 3. I don’t feel any discomforts in the decision process (Jin, He & Song, 2012)* 4. I think it was easy to find the pizza that I want (Bigné, 2008)*
5. I feel I had a great deal of control over the decision I made (van Noort, 2012)*
6. I think this way of presenting available options for pizza enables me to accomplish my ordering more quickly (Bigné, 1012)*
7. I think this way of presenting available options for pizza helps me to make better purchase decisions (Bigné, 2012)*
Thinking of your final choice of pizza, please answer the following statements on the scale from 1 (strongly disagree) to 7 (strongly agree).
1. I am very satisfied with the pizza I chose (Jin, He & Song, 2012) * 2. My chosen pizza will best meet my needs (Jin, He & Song, 2012)* 3. My chosen pizza has a good value (Jin, He & Song, 2012)*
4. I would consider purchasing the pizza I selected (Barber, Kuo, Bishop & Goodman Jr, 2012)**
5. I intend to try the pizza I selected (Barber, Kuo, Bishop & Goodman Jr, 2012)** 6. I plan on buying the pizza I selected in the future (Barber, Kuo, Bishop &
Goodman Jr, 2012)**
7. I am interested in tasting the pizza I selected (Barber, Kuo, Bishop & Goodman Jr, 2012)**
*measuring Attitudes towards the final decision **measuring the purchase intention
When choosing a pizza, could you actively select each toping individually? Yes/no
Almost done! The following questions are asked to get some demographic insights. What is your age (in years)?
__________
What is your gender? - Male
- Female
- Other: ________ - Prefer not to say
In which country do you currently live?
What is the highest degree or level of education you have completed? - Compulsory school
- High school - Apprenticeship
- Trade school - Bachelor’s degree - Master’s degree - Ph.D. or higher - Other:____ - Prefer not to say
Appendix B Experiment
Condition 1 (non-interactive condition)
Imagine you are about to order pizza online. Please have a look at the menu and choose your favorite pizza. For each pizza you can choose the dough you prefer. Choose your dough and pizza by simply ticking the box next to it.
Neapolitan (soft mixture with a high and soft crust)
Calabrian (thin crisp mixture with a low and crumbly crust)
Ingredients:
1. Tomato sauce, Mozzarella 2. Tomato sauce, Mushrooms 3. Tomato sauce, Salami 4. Tomato sauce, Pepperoni 5. Tomato sauce, Black olives 6. Mozzarella, Mushrooms 7. Mozzarella, Salami 8. Mozzarella, Pepperoni 9. Mozzarella, Black olives 10. Mushrooms, Salami 11. Mushrooms, Pepperoni 12. Mushrooms, Black olives 13. Salami, Pepperoni
14. Salami, Black olives 15. Pepperoni, Black olives
17. Tomato sauce, Mozzarella, Salami 18. Tomato sauce, Mozzarella, Pepperoni 19. Tomato sauce, Mozzarella, Black olives 20. Tomato sauce, Mushrooms, Salami 21. Tomato sauce, Mushrooms, Pepperoni 22. Tomato sauce, Mushrooms, Black olives 23. Tomato sauce, Salami, Pepperoni
24. Tomato sauce, Salami, Black olives 25. Tomato sauce, Pepperoni, Black olives 26. Mozzarella, Mushrooms, Salami 27. Mozzarella, Mushrooms, Pepperoni 28. Mozzarella, Mushrooms, Black olives 29. Mozzarella, Salami, Pepperoni
30. Mozzarella, Salami, Black olives 31. Mozzarella, Pepperoni, Black olives 32. Mushrooms, Salami, Pepperoni 33. Mushrooms, Salami, Black olives 34. Mushrooms, Pepperoni, Black olives 35. Salami, Pepperoni, Black olives
36. Tomato sauce, Mozzarella, Mushrooms, Salami 37. Tomato sauce, Mozzarella, Mushrooms, Pepperoni 38. Tomato sauce, Mozzarella, Mushrooms, Black olives 39. Tomato sauce, Mozzarella, Salami, Pepperoni
40. Tomato sauce, Mozzarella, Salami, Black olives 41. Tomato sauce, Mozzarella, Mushrooms, Black olives 42. Tomato sauce, Mushrooms, Salami, Pepperoni 43. Tomato sauce, Mushrooms, Salami, Black olives 44. Tomato sauce, Salami, Pepperoni, Black olives 45. Tomato sauce, Mushrooms, Pepperoni, Black olives 46. Mozzarella, Mushrooms, Salami, Pepperoni
47. Mozzarella, Mushrooms, Salami, Black olives 48. Mozzarella, Salami, Pepperoni, Black olives 49. Mozzarella, Mushrooms, Pepperoni, Black olives 50. Mushrooms, Salami, Pepperoni, Black olives
51. Tomato sauce, Mozzarella, Mushrooms, Salami, Pepperoni 52. Tomato sauce, Mozzarella, Mushrooms, Salami, Black olives 53. Tomato sauce, Mozzarella, Salami, Pepperoni, Black olives 54. Tomato sauce, Mozzarella, Mushrooms, Pepperoni, Black olives 55. Tomato sauce, Mushrooms, Salami, Pepperoni, Black olives 56. Mozzarella, Mushrooms, Salami, Pepperoni, Black olives
57. Tomato sauce, Mozzarella, Mushrooms, Salami, Pepperoni, Black olives
Condition 2 (adding frame) Step 1
Imagine you are about to order pizza online. For your pizza you can first choose the dough you prefer and then the ingredients you wish to have on your pizza.
Please choose the dough you prefer by ticking the box next to it. There is no price difference between the types of dough. Both options cost 5,00€.
Neapolitan (soft mixture with a high and soft crust)
Calabrian (thin crisp mixture with a low and crumbly crust)
Step 2
You have chosen the base of your pizza, the dough. Now, you can customize your pizza by adding up to all the ingredients on the list. Please, choose at least two ingredients by ticking the box next to it. Each ingredient you add will add 1 € to your basic Pizza dough of 5,00€.
Tomato sauce (+1 €) Mozzarella (+1 €) Mushrooms (+1 €) Salami (+1 €) Pepperoni (+1 €) Black Olives (+1 €)
Condition 3 (subtracting frame) Step 1
Imagine you are about to order pizza online. For your pizza you can first choose the dough you prefer and then the ingredients you wish to have on your pizza.
Please choose the dough you prefer by ticking the box next to it. Neapolitan (soft mixture with a high and soft crust)
Calabrian (thin crisp mixture with a low and crumbly crust)
Step 2
You have chosen the base of your pizza, the dough. Your pizza is currently fully loaded with six ingredients. Now you can customize your pizza by removing up to four of the ingredients on the list. You can remove ingredients by ticking the box next to it. Each ingredient you remove will subtract 1 € from your fully loaded pizza of 11,00€.
Tomato sauce (-1 €) Mozzarella (-1 €) Mushrooms (-1 €) Salami (-1 €) Pepperoni (-1 €) Black Olives (-1 €) Appendix C Tables Table 3
Comparison of the effects of Condition on Attitudes towards the final choice
Condition 1 Condition 2 Mdiff SD p
Non-interactive Adding -.08 .30 1.000
Non-interactive Subtracting -.20 .30 1.000
Adding Subtracting -.12 .31 1.000
Table 8
PROCESS analysis to test the mediating effect of Attitudes towards decision-making process on Number of toppings with Conditions (Interactive vs. Non-interactive) as independent variable and Decision-making styles as moderators
Variable Coefficient SE t p Moderation Constant 2.440 .996 2.450 .015 Interactive 1.597 1.290 1.238 .218 Utilitarian .067 .144 .467 .641 Interactive*Utilitarian .031 .186 .166 .869 Hedonic .242 .132 1.831 .069 Interactive*Hedonic -.128 174 -.735 .463 Mediation Constant 3.833 .275 13.950 .001
