The effect of shopping-unrelated phone usage on consumer behavior
Marloes StraatmanUniversity of Groningen
Faculty of Business & Economics MSc Marketing June 2020 Van Sijsenstraat 6A 9724NN Groningen 0639760876 m.straatman@student.rug.nl Student number: 2877090
TABLE OF CONTENTS
1. Introduction………... 2 2. Theoretical background………. 5 3. Method………... 12 4. Results……… 17 5. Discussion ………. 28 6. Conclusion………... 31 7. References………. 32 8. Appendix ……….. 37ABSTRACT
Across the globe, mobile phone usage has rapidly increased over the last few years. Research indicated that many consumers also use their phone when they are shopping, for shopping unrelated mobile phone activities. Despite the frequency of shopping-unrelated phone usage and the growing mobile dependency nowadays, there is little knowledge about how this affects our consumer behavior. The research provides more clarification on how shopping-unrelated mobile phone usage influences the shopping behavior of consumers through an experiment. The findings of this experiment show that shopping-unrelated mobile phone usage increases the probability to make unplanned purchases among consumers. This effect is mediated by the increase in cognitive load of the consumers. In addition to this, we found that different shopping-unrelated mobile phone activities have different effects on the cognitive load and amount of unplanned purchases made. The practical implications of this research are especially interesting for managers who are considering incorporating mobile phones into their consumer-based strategy.
1. INTRODUCTION
Mobile phone usage has rapidly increased over the last few years (Deloitte NL, 2019). People use their phone all day long; while watching TV, when they are out for dinner, while crossing the road, but also when they are shopping (Deloitte US, 2017). Results from the Global Mobile Consumer Survey of Deloitte from 2017 pointed out that 92% of the people owning a phone admitted to use their smartphone during their shopping trip. During the shopping trip people use their phone for various functions like browsing the internet for product information, online reviews, to compare prices or to look for discount codes. These kinds of activities are all related to the shopping task. However, it appears that almost half of the in-store phone usage (46%) is unrelated to the shopping task, such as engaging in private conversations, checking social media, listening to music, emailing or browsing the internet (Martin, 2016). Despite the frequency of shopping-unrelated phone usage and the growing mobile dependency nowadays, there is little knowledge about how this affects our consumer behavior.
While many researchers have examined the effect of mobile omnichannel marketing instruments (e.g. shopping-related phone usage) on buying behavior, Bellini & Aiolfi (2017) were the first to consider the effect of shopping-unrelated phone usage on the buying behavior of consumers. They found in their field experiment that consumers using mobile technology in a shopping-unrelated manner made fewer unplanned purchases. Their explanation for this was that these customers miss out on in-store marketing instruments, because they are distracted by their phones and therefore are less tempted to make unplanned purchases.
There are several reasons possible that could have caused these contradicting outcomes. First of all, both studies, concerning the number of unplanned purchases, were field studies. This causes that the experiments have a lower internal validity, which stands for the ability to measure what the experiment sets out to measure (Grimes & Schulz, 2002). In both experiments they measured the number of unplanned purchases made and the type of phone usage by letting the customers fill in a questionnaire after the shopping trip. They analyzed these variables and came up with certain results. However, Bellini & Aiolfi (2013) did not control for other external factors that may have influenced their consumer behavior during the shopping trip. For example, earlier research has revealed that the number of aisles visited (Inman et al.,2009), the amount of time spent in-store (Inman et al, 2009), use of a shopping list (Thomas & Garland, 2009) and the method of the payment (Soman, 2003) all influence the probability of making unplanned purchases. These factors will beyond a doubt have differed between all subjects, but are not taken into consideration in their field research.
Furthermore, in both experiments they looked at shopping-unrelated mobile phone usage in general. While there are a range of different activities, which may have very different effects. Scandria et al. (2019), also mentions this in her limitations for future research. Maybe one kind of activity, for example checking social media, results in a higher cognitive load than for example answering a phone call.
Therefore, the goal of this paper is to offer new insights on this topic by performing a lab experiment. This enables us to shut out other external factors that have an effect on impulse buying and allows us to focus solely on the effect of the different kinds of shopping-unrelated mobile phone usage on the probability to make unplanned purchases. The main research question of this paper is
“How does shopping-unrelated phone usage affect the probability to make unplanned purchases?”
The main focus of this paper will be on the different impact the different shopping-unrelated phone usage tasks have on the amount of unplanned purchases made. This study will also measure if the different phone tasks have different effects on the cognitive load of the subjects and if it prevents them from noticing marketing stimuli. By measuring both of these mediators and controlling for external factors, the ultimate goal of this study is to solve the contradiction in the literature for once and for all. Also, the previous research that has been done on the subject is extended by looking at the different types of shopping-unrelated mobile phone usage that may have different effects.
2. THEORETICAL BACKGROUND
In this chapter the theoretical background of the mediators, dependent and independent variables of our conceptual model will be discussed. Based on the existing literature and theories that are available concerning the variables of this research, hypotheses will be made.
2.1 Unplanned purchases
Unplanned purchasing includes products for which the decision to purchase was made in-store and not beforehand (Iyer,1989). It is a very common phenomenon in shopping. Earlier research has shown that over 50% of the purchases in supermarkets was unplanned (POPAI,1995; Stilley et al, 2010). The data of this research also showed that even though over 50% of the purchases were unplanned, the average planned spend laid only slightly lower than the average total spend. This indicates that shoppers have a mental budget for the trip that includes room to make unplanned purchases (Stilley, et al., 2010). So, these purchases were not part of the primary formed shopping plan.
The main effect that is examined in this research is the effect of shopping-unrelated phone usage on the probability to make unplanned purchases. As mentioned in the previous chapter, there is a contradiction in the literature about this effect. One research argues that shopping-unrelated mobile phone usage has a positive effect on the amount of unplanned purchases made. They found that shopping-unrelated mobile phone usage is a considerable distraction for consumers, and causes them to have subsequently poor adherence to their shopping plan, which results in more unplanned purchases (Scandria et al., 2019). This theory is further explained by the cognitive load theory, which will be extensively explained later on in this chapter. Other research found opposite results. They argue that shopping-unrelated mobile phone usage decreases the amount of unplanned purchases made, because it prevents them from noticing marketing stimuli (Bellini & Aiolfi, 2013). When comparing the two studies, it stands out that the study of Scandria et al., (2019) is better in terms of quality, since she used control variables to make sure that the right effect is measured and not influenced by other factors. Also, there is already a lot of knowledge about the distracting nature of mobile phones in terms of cognitive load. For instance, the research of Ward et al., (2016) where they found that the mere presence of somebody’s own smartphone already occupies the limited-capacity cognitive resources. And the research of Caramia et al., (2017) where they examined that effect of smartphone use on walking. Walking is an automatic motor task, but does require some cognitive resources and attention (Hausdorff et al., 2005). Since shopping is logically a more cognitive demanding task than walking, smartphone use is expected to have a significant effect on cognitive load and performance here as well.
For these reasons the following hypothesis is established:
H1: Shopping-unrelated phone usage during the shopping trip has a positive effect on the probability to make unplanned purchases.
2.2 Noticing marketing stimuli
extra displays, communications, extra activities, etc.) to lose its effectiveness and therefore the amount of unplanned purchases may decrease (Bellini & Aiolfi, 2017).
H2a: Shopping-unrelated mobile phone usage has a negative effect on the amount of marketing stimuli noticed.
H2b: The amount of marketing stimuli noticed has a positive effect on the probability to make unplanned purchases.
2.3 Cognitive load theory
The cognitive load theory origins from the late 1980s and is about the limited capacity of the working memory of humans (Sweller, 1988). Working memory capacity is the short-term memory that actively processes incoming information and where small amounts of information (Miller, 1956) are stored for a short duration (Dosher, 2003), as opposed to long-term memory. Since the working memory capacity is limited, it is hard for individuals to work on multiple tasks at the same time. This causes that the efficiency of the tasks performed will decrease when the individual engages in multiple tasks that are cognitive demanding (Sweller, 1988). When a consumer has a shopping plan, but is distracted by their mobile phone during the shopping trip by tasks that are not related to the primary shopping task, the cognitive load increases and consumers will handle the shopping task less efficiently. When cognitive resources are depleted, individuals will be more likely to make unplanned purchases (Vohs & Faber, 2007). Therefore, the following hypothesis is formed.
H3a: Shopping-unrelated mobile phone usage has a positive effect on cognitive load. H3b: Cognitive load has a positive effect on the probability to make unplanned purchases.
H4: Different shopping unrelated mobile phone tasks have different effects on the probability to make unplanned purchases
2.4 Shopping-mobile phone usage
Mobile phone usage dramatically increased over the last few years and people use their phone all day long, also while at the same time engaging in other daily tasks (Deloitte US, 2017; Deloitte NL, 2019). A lot of research has been done on the distracting nature of mobile phones while multitasking; for example the effect of smartphone use while driving. All studies indicated that the distraction of a smartphone is at the expense of performance, such as inattentional blindness and delayed reaction times (Strayer & Johnston, 2001; Caird et al., 2008; World Health Organization, 2011). Moreover, the distraction of a mobile phone seems to also affect the performance of tasks that cost less cognitive resources, like walking, in a negative way (Caramia et al., 2017). And even the mere presence of one’s own smartphone occupies cognitive resources (Ward et al., 2016). This all indicates that this shift to an ever-connected world, where the mobile phone is always present or in use, has definitely an impact on our behavior and performance. And that is exactly the reason why it is interesting that little research has been done before on the effect of shopping-unrelated mobile phone usage while shopping on consumer behavior.
The different types of activities that are examined in this study are checking social media (social), engaging in private conversations (convo), playing a game (game), listening to music (music) and e-mailing (e-mail). These activities are based on field research results of Scandria et al., (2019) and Bellini & Aiolfi (2013), where they inventoried what kind of shopping-unrelated mobile phone tasks where performed while grocery shopping.
representative of how a private conversation while shopping would occur and high cognitive load texting is more similar to e-mailing.
2.4.1 Listening to music
Listening to music was not part of the research of (Caramia et al., 2017) and there has not been done other relevant research about this activity and the cognitive effort it costs. There has been done a lot of research among the effects of background music in stores on customer behavior and satisfaction (Garlin & Owen, 2006; Yalch & Spangenberg, 2000). But this is not relevant in this study, since the treatment is only about the interruption of choosing a new song, without music playing on the background. Therefore, the impact that this activity is expected to have on the cognitive load and ability to notice marketing stimuli is estimated through common sense and logical reasoning. The task does not require the consumer to be very actively involved with their smartphone, except for changing the song when necessary. Therefore, it is expected to have the lowest impact of all activities on the cognitive load of the amount of marketing stimuli noticed.
2.4.2 Engaging in private conversations
Making a phone call during the shopping trip is a very different kind of distraction. In contrast to other activities it is not hindering the sight, but is only a distracting sound. However, research among phone distractions during a car ride have proven that it’s still quite a distraction, and hinders people to focus on their primary task. (McKnight & McKnight, 1993). In the experiment the phone call will be simulated by a recording of a friend telling a story over the phone. Simulated conversations have been used in research before and provide a suitable proxy for real-world mobile phone conversations (Drews et al. 2008). Because making a call is not hindering the sight and because it appeared to have a low impact on the cognitive load in earlier research (Caramia et al., 2017) this activity is expected to have one of the lowest impacts on cognitive load and the amount of marketing stimuli noticed.
2.4.3 Social media usage
2018; Tariq et al., 2012). In the research of Caramia et al., (2017) scrolling through social media is found to have a medium impact on cognitive load. For these reasons, social media usage is also expected that have an average / medium impact on the cognitive load and amount of marketing stimuli noticed.
2.4.4 Playing a game
Since the global adoption of 3G mobile technology the smartphone developed into a serious gaming platform (Kim, 2013). For the experiment the game app Wordfeud was used, because it’s a very well-known and popular game, which demands some cognitive resources. Also, it is an interactive game that sends push notifications, so it is likely that people are distracted by it when they get a notification during the shopping trip. It is expected that playing a game will cost a higher amount of cognitive resources, which is also confirmed in the findings of Caramia et al., (2017). Therefore, it is expected to have a relatively high impact on cognitive load and the amount of marketing stimuli noticed.
2.4.5 E-mailing
As mentioned before, emailing is more similar to high cognitive load chat texting, since it is a more formal way of communicating and requires more attention than just casually texting over Whatsapp for example. Caramia et al., (2017) found that high cognitive chat texting costed the most cognitive resources. Therefore, it is expected that reacting to an e-mail while shopping will have the highest impact on cognitive load and amount of marketing stimuli noticed.
As mentioned before, the order in which these shopping-unrelated mobile phone activities are discussed represent the expected impact they will have on cognitive load and the amount of marketing stimuli noticed, in ascending order. That being said, the following hypothesis is formed about how the different activities will affect the probability to make unplanned purchases.
H5: Respondents that engage in activities that have a higher impact on cognitive load and amount of marketing stimuli noticed, as stated in the order mentioned above, will have a higher probability to make unplanned purchases.
Figure 1: Conceptual model
3. METHOD
The hypothesis was tested by conducting an experiment, by means of a survey in Qualtrics that simulated a shopping trip in a Dutch Supermarket (Appendix 8.1). There were six different treatments that Qualtrics randomly assigned to the subjects, this ensures that every treatment has an equal number of respondents. In order to reach as many participants as possible, the survey was shared on multiple Social Media platforms, like Facebook, Instagram and LinkedIn. The survey was also shared in multiple big Facebook groups. This helped to make the subject pool more representative for the entire population, since it provided more subject diversity in age and level of education. Sharing the survey to only my personal network, would have caused the subject pool to be mostly students, so this was a good solution.
3.1 Procedure and treatments
A shopping trip through a Dutch supermarket was simulated by using anecdotes and videos of all different stages during the shopping trip. The subjects were shown a shopping list upfront, giving them a goal on what kinds of products to buy. During the shopping trip the subject was given a series of choices related to whether they wanted to buy a certain product or not. They were asked to answer what the probability was (on a scale of 0 - 100) that they would buy the product. Some of these products were products from their shopping list (planned purchases) and some products were not on their initially stated shopping list (unplanned purchases). The different treatments are implemented by different intermezzo’s that the subjects will experience during the shopping trip. The different treatments and how they are simulated during the shopping trip are listed below.
3.2 Independent variables
The independent variables are the different types of shopping-unrelated mobile phone usage that are being tested. These are assigned as different treatments over the total number of respondents. In the table below (Table 1) each variable with the corresponding treatment that was used is explained. All different treatments put the subject in a decision-making mode, which ensures that they are actively involved in the activity.
Variable Treatment
Playing a game The subject will be interrupted twice to play a Wordfeud game.
Social Media The subject will be interrupted twice by a social media post on Instagram and given the choice to like the post.
E-mailing The subject will be interrupted twice by an e-mail and asked to write a reply. The first email will be from the RUG and the second e-mail will be for a business meeting.
Engaging in private conversation
The subject will be interrupted by a recording of mother asking them what they have been up to this weekend. Afterwards, the subject will be asked to type how they would answer. Next, the subject will be interrupted by a Whatsapp conversation and asked to type a reply to their friend.
Listening to music The subject will be interrupted twice to decide which of a list of songs they want to listen to.
Control condition The subject will not be interrupted by any mobile phone usage.
Table 1: Variables with corresponding treatment
3.2 Dependent variable: probability to make unplanned purchases
In the survey all subjects are confronted with these two products during their shopping trip and asked to answer on a scale of 0 - 100 what the probability is that they would put this product in their shopping cart. The control group sees a video of the aisle with the products on the shelf, with the marketing stimuli clearly shown as well. The other treatment groups are in front of the shelf, but are in the meantime interrupted by their phone. They see the products and the marketing stimuli as well, but in the background behind their phone.
3.3 Mediators
To find out whether the effect of mobile phone usage on unplanned purchases can be explained via the cognitive load theory as Scandria et al. (2019) stated or through the prevention of noticing marketing stimuli as Bellini & Aiolfi (2017) argued, both variables are included as possible mediators in this research.
3.3.1 Cognitive load
For measuring the cognitive load of the subjects, earlier literature and experiments were reviewed. Sweller (1988), who is one of the founding fathers of cognitive load theory, evaluated different scales and concluded that a simple subjective rating scale has surprisingly shown to be the most accurate measure available to differentiate the cognitive load imposed by different instructional procedures. This single item category rating scale is also adopted by Paas (1992) who also contributed greatly to the cognitive load theoretical framework with many studies (Paas & Meriërenboer 1994; Paas et al., 2003; Paas et al., 2004). Therefore, in this research the single item category rating scale is used as well. Respondents are presented with one item on which they have to translate their perceived amount of mental effort they experience during the shopping trip into a numerical value. Mental effort is defined as “the aspect of cognitive load that refers to the cognitive capacity that is actually allocated to accommodate the demands imposed by the task; thus, it can be considered to reflect the actual cognitive load” (Paas, Tuovinen, et al., 2003, p. 64). At the end of the survey the subjects are asked to rate the amount of mental effort on a 7-point Like scale.
3.3.2 Noticing marketing stimuli
whether they saw whether the Ben & Jerry’s and the beer was on discount. They can indicate whether they saw both marketing stimuli, one of them or none. So, values here range from 0 - 2.
3.4 Control Variables
At the end of the survey also some additional questions will be asked concerning the control variables stated earlier. In the table below (Table 2) all variables are shown with the corresponding questions that are asked in the survey.
Variable Questioned in survey
Age Age in numbers
Gender Male / female
Mood Indication of mood on a scale of 1 -7 Likert scale (with visual input)
Disposable income Monthly disposable income (rounded on hundreds in euros) Impulsiveness
(personality trait)
Impulsiveness was measured using a six-item 7-point Likert scale adapted from Puri (1996) and used in earlier research from Scandria et al., (2019). Respondents were provided with six adjectives and asked to indicate their level of agreement with how well each attribute described them: impulsive, easily tempted, enjoy spending, a planner, self-controlled, and restrained, on a scale where 1 = strongly disagree and 5 = strongly agree. The last three items were reversed coded for analysis.
Table 2: Control variables
3.5 Analysis
The results of this experiment will be analyzed in IBM SPSS Statistics 25. This is a program produced by IBM that is designed to execute a wide range of statistical procedures (Cronk, 2019). Different types of analysis will be used for the different types of data. Beforehand a factor analysis will be executed for the multi-item scale of measures impulsivity as a personality trait. Factor analysis is a technique that is used to reduce a large number of variables into fewer numbers of factors (Young & Pierce, 2013). The items will then be merged into one or more factors, which can be used for further analysis.
Analysis Of Covariance (ANCOVA) tests will be performed. An ANOVA test is used to compare means of dependent variables between groups. An ANCOVA is used for the same purpose, but also takes into account control variables (Cronk, 2019). Control variables are factors that can strongly influence the results of an experiment. With ANCOVA the variance explained by the control variables is neutralized, so that the relationship between the dependent and independent variables that we do want to examine is more accurate (Birks & Malhotra, 2006).
4. RESULTS
In this section the data analysis and the results from this study will be presented. Here only the plain results are stated and in the next chapter these findings will be further interpreted and discussed.
4.1 Data cleaning
First all data from respondents who did not complete the survey until the end were deleted. This left us with 679 respondents. Next, the data was inspected for missing values. There were two respondents who did not fill in their age. These two respondents were deleted from the survey, which left us with 677 respondents. Lastly, it occurred that people interpret a question in a lot of different ways, which made that the answers on this question were not consistent and reliable. This question was about the disposable income per month of the subject. Some respondents filled in their disposable income per month, some respondents filled in how much they spent on groceries per month and some respondents answered that they did not feel comfortable with sharing this information. Unfortunately, this control variable is therefore deleted from the data. However, this is not a big issue for the analysis of the results, since in the beginning of the survey it is mentioned twice that people can assume that they have more than enough money to pay for all the groceries. So, they were told to not let money restraints play a role in their purchase decisions. This insight made the control variable ‘income’ unnecessary anyway. So, the removal of this variable from the data is not an issue for this research.
4.2 Descriptive statistics
4.3 Factor Analysis
Before the data is ready to use for testing our hypotheses, a factor analysis is performed to merge the different items into one or more factors, for the items that measure impulsivity as a personality trait among the respondents. This is measured by 6 items, of which 3 are reverse scored variables. So, people who are very impulsive score high on the normal scored three items and low on the reverse scored variables. This helps to make sure that the respondents are paying attention when filling in the scales in the survey. However, for our analysis it is necessary to recode them into new variables that all have the same way of scoring. This was done in SPSS. After that a Factor Analysis was performed with the 6 items. By examining the eigenvalues and total percentage of variance explained two factors were initially formed by SPSS.
Figure 2: Factor analysis - correlation matrix
All items significantly correlate with each other, except for the item about how restrained and how easily tempted the respondent is (Figure 2). When looking at the Rotated Component Matrix (Figure 3), it is clear that this negative correlation also prevents the ability to load all the items on one component. The item about how easily tempted someone is has the highest possible loading, so that is the reason why the item about how restrained someone is, was removed from the data.
Figure 4: Factor Analysis - new Correlation Matrix
Now all items correlate with each other significantly (Figure 4). To test whether the factor analysis makes sense and is reliable the KMO & Bartlett’s Test and Cronbach’s Alpha are checked (Appendix 8.4). The KMO should be higher than 0.5 (Malhotra, 2009) and is 0.712 which is great and the Bartlett’s Test is significant (p = 0.00). The Cronbach’s alpha should be higher than 0.6 (Malhotra, 2009) and is 0.652 which is good.
All items can be loaded on one component (Figure 5). And they all have a loading of > 0.5. Therefore, a new variable was computed CV2_impulsive personality. This variable will be taken into account as a control variable for the rest of the analysis.
4.3 Testing the hypotheses.
In this section the main hypotheses will be tested through ANCOVA and ANOVA tests in SPSS. This section is divided into two parts. Because one part of our research is testing the effect of mobile phone usage in general on the probability to make unplanned purchases, the cognitive load and the ability to notice marketing stimuli. While another part is testing whether different kinds of mobile phone activities have different effects on the probability to make unplanned purchases, the cognitive load and the ability to notice marketing stimuli.
4.3.1 Analysis of shopping-unrelated mobile phone usage effect in general
A one-way analysis of covariance (ANCOVA) was performed to examine the effect of shopping-unrelated mobile phone usage on the amount of unplanned purchases made. For the dependent variable was a new variable computed (DV_Combined_Average). This is the average of the two probabilities of buying an unplanned purchase that the respondent indicated. For the independent variable a dummy variable was created (0 = control group, 1 = mobile treatment). And for the control variables gender, age, mood and the new variable from the factor analysis for impulsivity as a personality trait were used.
When looking at the estimated marginal means (adjusted for covariation) in the ANCOVA (Figure 6) it shows that the control group has a mean probability to make unplanned purchases of 27.803 and the mobile treatment group has a mean probability of 46.237. The Test of Between- Subjects effects shows that this difference between groups is significant (P = 0.000).
Figure 6: ANCOVA - effect of shopping unrelated mobile phone usage on the probability to make unplanned purchases
To learn more about how the mediators, cognitive load and noticing marketing stimuli, play a role in this effect, the means of these variables are compared between groups (Figure 7). The mean of the cognitive load (on a scale of 0 - 7) is 2.1066 for the control group and 3.4595 in the mobile treatment group. The mean cognitive load is thus higher in the mobile treatment group. The ANOVA table on the next page (Figure 8) shows that this difference between groups is significant (P = 0.00).
Figure 7: Compare means - mediators
Figure 8: ANOVA - mediators
4.3.2 Analysis of different shopping-unrelated mobile phone usage effects (all treatments)
that indicates the treatment that the subject was assigned to (0 = control, 1 = game treatment, 2 = social media treatment, 3 = listening music treatment, 4 = conversation treatment, 5 = e-mail treatment). This variable named ‘Treatment’ is used as the independent variable in this ANCOVA.
When looking at the covariate-adjusted means in this ANCOVA it stands out that the control group has again by far the lowest probability to make unplanned purchases (= 27.849). The e-mail and conversation treatment group have the highest probabilities to make unplanned purchases with values of 51.588 and 53.167.
Figure 9: ANCOVA - effects of different treatments on probability to make unplanned purchases
Figure 10: Compare means - mediators
Figure 11: ANOVA - mediators
4.3.3 Regression analysis of mediator effects
model. So, both the control variables and the mediators are entered here. The regression model has an adjusted R-Squared value of .113. This is not very high, but in this research it is not a problem since the model is not going to be used to make predictions. We can still draw important conclusions about how changes in the predictor values (cognitive load and noticing marketing stimuli) are associated with changes in the dependent variable (probability to make unplanned purchases). When looking at the ANOVA table (Appendix 8.5) it shows that p = 0.000, which means that the independent variables reliably predict the dependent variable. When looking at the standardized coefficients in the Coefficients table (Figure 12) it is visible that the cognitive load has a strong positive effect (0.180) on the probability to make unplanned purchases and is significant (p = 0.00). Noticing marketing stimuli has a small negative coefficient (-0.039) and is not significant (p = 0.292).
Figure 12: Linear regression analysis - coefficients
Below the results are summarized by indicating which hypotheses are supported and which hypotheses are rejected (Table 3).
Hypothesis Hypothesis description Result
H1 Shopping-unrelated mobile phone usage has an effect on the probability to make unplanned purchases
Supported
H2a Shopping-unrelated mobile phone usage has a negative effect on the amount of marketing stimuli noticed.
H2b The amount of marketing stimuli noticed has a positive effect on the probability to make unplanned purchases.
Rejected
H3a Shopping-unrelated mobile phone usage has a positive effect on cognitive load
Supported
H3b Cognitive load has a positive effect on the probability to make unplanned purchases
Supported
H4 Different shopping unrelated mobile phone tasks have different effects on the probability to make unplanned purchases
Supported
H5 Respondents that engage in activities that have a higher impact on cognitive load and amount of marketing stimuli noticed, as stated in the order mentioned above, will have a higher probability to make unplanned purchases.
Rejected
5. DISCUSSION
5.1 Findings
The main goal of this research was to find out what the effect of shopping-unrelated phone usage is on the probability to make unplanned purchases. Hypothesis 1 was supported, which means that there is definitely an effect of shopping-unrelated mobile phone usage on the probability to make unplanned purchases. People who did not use their phone during the shopping trip had a significantly lower probability to make unplanned purchases than people who did use their mobile phone for unrelated activities. This seems to support the earlier findings made by Scandria et al. (2019) and the point of view of Bellini & Aiolfi (2013) seems to invalid.
We also tested the mediation effects that both researchers proposed. Scandria et al. (2019) reasoned that people who use their phone make more unplanned purchases, because of the cognitive load theory. She argues that shopping-unrelated mobile phone usage would increase the cognitive load of the shopper and that this would cause the increase in unplanned purchases made. Our results supported this reasoning. Our research confirmed that people who were interrupted by shopping-unrelated mobile phone tasks experienced a higher cognitive load. And this cognitive load had a significantly positive effect on the probability to make unplanned purchases.
The mediator variable ‘noticing marketing stimuli’ is higher for people in the control Group. So, people who used a mobile phone during the shopping trip noticed significantly less marketing stimuli. This is in line with the argumentation of Bellini & Aiolfi (2013). However, the amount of marketing stimuli noticed does not have a significant effect on the probability to make unplanned purchases. So, people who use their phone do see less marketing stimuli, but seeing marketing stimuli is not a significant trigger for making more unplanned purchases.
comes to noticing marketing stimuli. An explanation for this could be that the interruption where the respondents gets a call is not visually demanding. There is nothing to see on the mobile phone, only the sound is important here. This may cause the respondent to pay more attention to the environment. The game treatment is quite distracting in both categories as expected. The social media treatment has less of an effect on cognitive load than expected beforehand. Even though the proven distracting nature of social media (Brooks, 2015) it did not have a big effect on cognitive load, but it did prevent people from noticing marketing stimuli. So the task was more visually distractive, but did not cost a high mental effort. Lastly, the music treatment was as expected, not that distracting.
Distraction level Expected
Distraction level (in terms of cognitive load)
Distraction level
(in terms of missing marketing stimuli)
E-mail treatment Conversation treatment Game treatment
Game treatment E-mail treatment Social media treatment Social media treatment Game treatment E-mail treatment Conversation treatment Social media treatment Music treatment
Music treatment Music treatment Conversation treatment Table 4: Distraction level of shopping-unrelated mobile phone activities
5.2 Limitations
A limitation of this research is that the experiment was not conducted in a neutral environment for everyone, since there was no lab available for this project. Respondents filled in the survey in their own time and environment, which may have caused differences in the results. For instance, respondents who were in a noisy environment or only had a limited time to fill in the survey could have experienced a different cognitive load than people who were in a calmer environment. Looking back, this could have been included in the survey and used as a control variable.
respondent. This deviates from how the real-life situation would have been and causes a bit of a lower external validity on this aspect.
5.3 Future research
For future research I would recommend to run this experiment again in a lab to control for the effect that a noisy vs. a calm environment may have on the cognitive load of the respondents. Or when there are enough resources, the shopping trip could even be better simulated in a recreated supermarket in a lab. This would also neutralize the factors of the environment, since all respondents shop in that same recreated supermarket and it would make the shopping simulation more realistic for the respondents.
6. CONCLUSION
6.1 Summary of findings
The results of this research show that shopping-unrelated mobile phone usage increases the probability to make unplanned purchases among consumers. When looking at the possible variables that mediate this effect, it appears that shopping-unrelated mobile phone usage causes an increase in cognitive load of the consumers. The limited cognitive resources of the shoppers causes them to be less efficient in sticking to the shopping plan and make more unplanned purchases. In addition to this, this study indicated that different shopping-unrelated mobile phone activities have different effects on the cognitive load and amount of unplanned purchases made. An activity that involves an interaction with someone seems to cause a higher cognitive load than other activities. Lastly, the mediation effect of noticing marketing stimuli was also examined. From the results can be concluded that shopping-unrelated mobile phone usage does visually hinder people to notice marketing stimuli. However, noticing marketing stimuli does not have a significant effect on the probability to make unplanned purchases. So, this mediation effect is not relevant.
6.2 Academic and practical implications
7. REFERENCES
Banerjee, S., & Longstreet, P. (2016). Mind in eBay, body in Macy’s. Journal of Research in Interactive Marketing.
Babin, B. J., Darden, W. R., & Griffin, M. (1994). Work and/or fun: measuring hedonic and utilitarian shopping value. Journal of consumer research, 20(4), 644-656.
Ball-Rokeach, S. J., & DeFleur, M. L. (1976). A dependency model of mass-media effects. Communication research, 3(1), 3-21.
Bellini, S., & Aiolfi, S. (2017). The impact of mobile device use on shopper behaviour in store: An empirical research on grocery retailing. International Business Research, 10(4), 58-68.
Barratt, E. S. (1993). Impulsivity: Integrating cognitive, behavioral, biological, and environmental data.
Bashar, A., Ahmad, I., & Wasiq, M. (2013). A study of influence of demographic factors on consumer impulse buying behavior. Journal of Management Research, 13(3), 145-154.
Birks, D. F., & Malhotra, N. K. (2006). Marketing Research: an applied approach. Pearson Education UK.
Brooks, S. (2015). Does personal social media usage affect efficiency and well-being?. Computers
in Human Behavior, 46, 26-37.
Caird, J. K., Willness, C. R., Steel, P., & Scialfa, C. (2008). A meta-analysis of the effects of cell phones on driver performance. Accident Analysis & Prevention, 40(4), 1282-1293.
Caramia, C., Bernabucci, I., D'Anna, C., De Marchis, C., & Schmid, M. (2017). Gait parameters are differently affected by concurrent smartphone-based activities with scaled levels of cognitive effort. PloS one, 12(10).
Dhar, R., & Wertenbroch, K. (2000). Consumer choice between hedonic and utilitarian goods. Journal of marketing research, 37(1), 60-71.
Deloitte (2017). Global Mobile Consumer Survey: US Edition. Retrieved from
https://www2.deloitte.com/content/dam/Deloitte/us/
Documents/technology-media-telecommunications/us-tmt-2017-global-mobile-consumer-survey-executive-summary.pdf Deloitte (2019), Deloitte Global Mobile Consumer Survey 2018 - 2019, NL edition. Retrieved from
https://view.deloitte.nl/rs/502-WIB-308/images/deloitte-nl-tmt-gmcs2019-report.pdf
Dosher, B. (2003). Working memory. In Encyclopedia of cognitive science (Vol. 4, pp. 569–577). New York: Wiley
Hausdorff, J. M., Yogev, G., Springer, S., Simon, E. S., & Giladi, N. (2005). Walking is more like catching than tapping: gait in the elderly as a complex cognitive task. Experimental brain research,
164(4), 541-548.
Hildebrand, C., & Schlager, T. (2019). Focusing on others before you shop: exposure to Facebook promotes conventional product configurations. Journal of the Academy of Marketing Science, 47(2), 291-307.
Inman, J. J., Winer, R. S., & Ferraro, R. (2009). The interplay among category characteristics, customer characteristics, and customer activities on in-store decision making. Journal of
Marketing, 73(5), 19-29.
Jalees, T. (2009). An empirical analysis of impulsive buying behavior in Pakistan. Market Forces,
5(3).
Garlin, F. V., & Owen, K. (2006). Setting the tone with the tune: A meta-analytic review of the effects of background music in retail settings. Journal of business research, 59(6), 755-764.
Grimes, D. A., & Schulz, K. F. (2002). Bias and causal associations in observational research. The
lancet, 359(9302), 248-252.
Kim, H. M. (2013). Mobile media technology and popular mobile games in contemporary society.
International Journal of Mobile Marketing, 8(2).
Kollat, D. T., & Willett, R. P. (1967). Customer impulse purchasing behavior. Journal of marketing research, 4(1), 21-31.
Kollat, D. T., & Willett, R. P. (1969). Is impulse purchasing really a useful concept for marketing decisions?. Journal of Marketing, 33(1), 79-83.
Hildebrand, C., & Schlager, T. (2019). Focusing on others before you shop: exposure to Facebook promotes conventional product configurations. Journal of the Academy of Marketing Science,
47(2), 291-307.
Iyer, E. S. (1989). Unplanned Purchasing: Knowledge of shopping environment and. Journal of
retailing, 65(1), 40.
Leppink, J., Paas, F., Van Gog, T., van Der Vleuten, C. P., & Van Merrienboer, J. J. (2014). Effects of pairs of problems and examples on task performance and different types of cognitive load.
Learning and Instruction, 30, 32-42.
Malhotra, N. K., 2009, Marketing Research, an Applied Orientation Global Edition, 2009, 6th edition, Prentice Hall
Martin, C. (2016). In-Store Phone Usage Tied Directly to Shopping Activity. Retrieved from https://www.mediapost.com/publications/article/288150/in-store-phone-usage-tieddirectly-to-shopping-act.html
McKnight, A. J., & McKnight, A. S. (1993). The effect of cellular phone use upon driver attention.
Accident Analysis & Prevention, 25(3), 259-265.
Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological review, 63(2), 81.
Paas, F., Tuovinen, J. E., Tabbers, H., & Van Gerven, P. W. (2003). Cognitive load measurement as a means to advance cognitive load theory. Educational psychologist, 38(1), 63-71.
Rook, D. W. (1987). The buying impulse. Journal of consumer research, 14(2), 189-199.
Rippé, C. B., Weisfeld‐Spolter, S., Yurova, Y., Dubinsky, A. J., & Hale, D. (2017). Under the sway of a mobile device during an in‐store shopping experience. Psychology & Marketing, 34(7), 733-752.
Schmeck, A., Opfermann, M., Van Gog, T., Paas, F., & Leutner, D. (2015). Measuring cognitive load with subjective rating scales during problem solving: differences between immediate and delayed ratings. Instructional Science, 43(1), 93-114.
Sciandra, M. R., Inman, J. J., & Stephen, A. T. (2019). Smart phones, bad calls? The influence of consumer mobile phone use, distraction, and phone dependence on adherence to shopping plans. Journal of the Academy of Marketing Science, 47(4), 574-594
Soman, D. (2003). The effect of payment transparency on consumption: Quasi-experiments from the field. Marketing Letters, 14(3), 173-183.
Strayer, D. L., & Johnston, W. A. (2001). Driven to distraction: Dual-task studies of simulated driving and conversing on a cellular telephone. Psychological science, 12(6), 462-466.
Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive science,
12(2), 257-285.
Stilley, K. M., Inman, J. J., & Wakefield, K. L. (2010). Planning to make unplanned purchases? The role of in-store slack in budget deviation. Journal of consumer research, 37(2), 264-278.
Tariq, W., Mehboob, M., Khan, M. A., & Ullah, F. (2012). The impact of social media and social networks on education and students of Pakistan. International Journal of Computer Science Issues
(IJCSI), 9(4), 407.
Thomas, A., & Garland, R. (1993). Supermarket shopping lists: their effect on consumer expenditure. International Journal of Retail & Distribution Management.
Vale, R. C., & Duarte, J. (2013). Classification of fmcg product macro-categories on the utilitarian vs. hedonic dimensions. Laboratório de Psicologia, 11(1), 29-36.
Vohs, K. D., & Faber, R. J. (2007). Spent resources: Self-regulatory resource availability affects impulse buying. Journal of consumer research, 33(4), 537-547.
Ward, A. F., Duke, K., Gneezy, A., & Bos, M. W. (2017). Brain drain: The mere presence of one’s own smartphone reduces available cognitive capacity. Journal of the Association for Consumer Research, 2(2), 140-154.
Wood, M. (1998). Socio-economic status, delay of gratification, and impulse buying. Journal of economic psychology, 19(3), 295-320.
8. APPENDIX
8.1 Survey
8.2 Descriptive statistics between control group / mobile treatment Report
dummy_mobileused CV4_Age CV3_Sex CV1_Mood
control group Mean 32.80 1.85 3.9180
N 122 122 122
Std. Deviation 13.903 .356 .75613
Minimum 17 01.00 2.00
Maximum 72 02.00 5.00
Median 26.00 2.00 4.0000
mobile treatment Mean 34.30 1.84 3.9477
N 555 555 555 Std. Deviation 14.308 .367 .77586 Minimum 14 01.00 1.00 Maximum 83 02.00 5.00 Median 29.00 2.00 4.0000 Total Mean 34.03 1.84 3.9424 N 677 677 677 Std. Deviation 14.238 .365 .77187 Minimum 14 01.00 1.00 Maximum 83 02.00 5.00 Median 28.00 2.00 4.0000
8.3 Descriptive statistics between all groups Report
Treatment CV4_Age CV3_Sex CV1_Mood
