Media Multitasking and Persuasion:
The Moderating Role of Emotions and Screen DifferencesName: Malou Bromberg Student number: 11851805
Master’s Thesis Graduate School of Communication Master’s programme Communication Science
Supervisor: Hilde Voorveld Date: 1-02-2019 Word count: 10.804
Abstract
The aim of this study is to examine whether media multitasking (engaging in- or quickly switching between two or more media activities at the same time) has an effect on cognitive (recall and recognition) and affective (text likeability and advertisement likeability) reactions towards advertisements and text. Moreover, this study aims to determine the influences of two moderating variables on this relationship; emotion (happy versus sad) and screen differences (computer versus smartphone). Based on the Limited Capacity Model, Elaboration Likelihood Model, Counterarguing Inhibition Hypothesis and Heuristic Systematic Model it was expected that media multitasking would lead to more negative cognitive outcomes, and more positive affective outcomes, in comparison to single tasking. Further, interaction effects were expected between media multitasking and emotion, as well as media multitasking and screen differences, on cognitive and affective reactions. By means of an online experiment (N = 95), people were asked to read an article from www.nu.nl about Nespresso, and watch an advertisement about Chevrolet. Half of the participants received a sad advertisement, whereas the other half of the participants received a happy advertisement. Through convenience sampling, 45 participants completed the questionnaire on their smartphone and 50 participants on their computer. The study did not show any main effects. However, two interaction effects were found. When participants were media multitasking on a computer screen they were less likely to recognize a message, but they had a more positive attitude towards the text in comparison with media multitasking on a smartphone. Conversely, when participants were single-tasking on a computer screen, they were more likely to recognize a message, but they had a less positive attitude towards the text in comparison with single-tasking on a smartphone. Thus, an interaction effect between media multitasking and screen differences with regard to recognition and text likeability was found.
Introduction
We live in a world in which the digital media environment is a dominant driver for consumer behaviour where the rate of change is moving at a fast pace. In this technology-dependent era, people are regularly engaging in two or more media activities at the same time. Whether at home, at work, at school or somewhere in between, the tendency to multitask with media is evident. When one takes a moment of their day to look up from their screen, they will notice people constantly switching tabs on their smartphones or computer (Jeong & Hwang, 2016), listening to music while reading, playing a game while studying, quickly switching between reading an online article and shopping online, and so on. The concurrent consumption of- or rapid switching between two or more forms of media is referred to as media multitasking (Brasel & Gips, 2017; Jeong & Hwang, 2016).
Media multitasking has become an integral part of the digital world of today (Deloitte Development, 2018; Veer, Boekee & Hoekstra, 2018). Even though media multitasking is often portrayed as something negative, it plays an unavoidable role in our society (Lui & Wong, 2012; Veer et al., 2018). Several studies have shown that media multitasking can even help with integration of information (Lui & Wong, 2012) and result in a more positive brand attitude (Segijn, Voorveld & Smit, 2017). This research aims to build on literature highlighting the positive aspects and ways in which media multitasking can be used to one’s advantage (Lui & Wong, 2012; Segijn et al., 2017).
Research has shown that media multitasking is especially prevalent among younger adults, in comparison to older adults, and they are more likely to adopt media multitasking habits than older generations (Carrier et al., 2009; Voorveld & van der Goot, 2013). For that reason, this research focuses on young adults, between the ages of 18 and 26.
Media multitasking affects the way we see, process and use information that we are exposed to. One of the principal indicators for measuring the persuasive effectiveness of
messages and advertisements is the attitude towards the advertisement. Research has shown that while media multitasking, comprehension and counterarguing are reduced, resulting in less resistance towards advertising messages and a more positive brand attitude (Jeong & Hwang, 2012; Segijn et al., 2016). However, other research has shown that worsened recall, due to media multitasking has led to a more negative brand attitude (Segijn, Voorveld, & Smit, 2016). Researchers have also brought up several concerns when it comes to cognitive reactions to media multitasking. Research has shown that media multitasking can lead to worsened recognition and factual recall, due to higher cognitive load (Angell et al., 2016; Jeong & Hwang, 2016; Segijn et al., 2016; Van Cauwenberge, Schaap, & Van Roy, 2014; Voorveld, 2011). Due to the previously found evidence with regard to cognitive and affective reactions to media multitasking, the first aim of this research is to examine the influence media multitasking has on cognitive and affective reactions. Cognitive reactions will be measured by looking at recognition and recall of the text. Affective reactions will be measured by measuring text likeability and advertisement likeability.
While there are many factors that may influence the relationship between media multitasking and cognitive, as well as affective reactions, one possible explanation is the moderating effect of emotions, which has not been widely covered in scientific research. The importance of emotions with regard to solely information processing has been widely studied (MacInnis & Jaworski, 1989; Shiv & Fedorikhan, 1999; Szczygiel, Buczny, & Bazinska, 2012; Weinberger, Spotts, Campbell & Parsons, 1995), and the power of emotion in advertising is evident (Osaka, Yaoi, Minamoto & Osaka, 2013). Research supports the notion that a higher level of emotional experience (positive or negative) facilitates information processing (MacInnis & Jaworski, 1989; Shiv & Fedorikhan, 1999). Using positive emotions in advertisements has shown to reduce negative cognitions through distraction, while it enhances attitudes, positive affect, attention, and recognition (Bradley, Greenwald, Petry &
Lang, 1992; Lewinski, Fransen & Tan, 2016). It has been suggested by research that negative emotions negatively affect working memory, impairing recall (Kominos, 2017; Osaka, Yaoi, Minamoto & Osaka, 2013; Tyng, Amin, Saad & Malik, 2017). In general, negative stimuli reduce information processing (Peterson & Kerin, 1979). Further, it has also been studied that positive and negative emotions affect working memory, cognitive and affective task performance. (Osaka et al., 2013).
The current research seeks to explain how emotions may affect the way that information is processed cognitively and affectively, with regard to media multitasking. As positive emotions have a different effect on information processing than negative emotions do, two sorts of emotions will be compared. This will be done so using advertisements that are either happy or sad. Therefore, the second aim of this research is to find out whether positive (happy) or negative (sad) advertisements result in different affective and cognitive reactions while media multitasking versus single-tasking.
Another factor to be of possible influence in the relationship between media multitasking and cognitive or affective reactions is the role that differences in screens (computer versus smartphone) play. Technological innovation is not only transforming the amount of media that is used, but also the way it is being used. Clickable screens, zoom options, different sizes of screens. These aspects of screens influence consumers’ information processing (Hou et al., 2012). Touch ability- the ability to click or zoom into your screen with your finger, has shown to be of influence on affective outcomes (Wei Shi & Kalyanam, 2018). It has shown to result in closer attention being paid to items at hand, as well as stronger purchase intent (Wei Shi & Kalyanam, 2018). Large screens, in comparison to smaller screens, are rated higher in quality and result in more positive affective outcomes, while resulting in more negative cognitive outcomes (Hou et al., 2012; Kim, 2017; Kim & Sundar, 2014; Lombard, 1995).
Differences in screens; size, touch- and click ability seem to be evident. The third aim of this research is therefore to take the differences between screens into account as playing a moderating role in the relationship between media multitasking and cognitive, as well as affective reactions. Where computers have a larger screen size, smartphones have more touch- and click ability. These two specific types of technology have not yet been compared in previous research.
Taken together, this research aims to determine the influence of emotions conveyed in an advertisement, and screen differences, on the relationship between media multitasking and recall, recognition and text- or advertisement likeability. With that, this research intends to inform readers into possible ways for- and factors that may allow companies, marketeers and researchers to use media multitasking to its advantage.
Theoretical Background Media Multitasking Effects on Cognitive Responses
Limited Capacity Model of Motivated Mediated Message Processing. To examine how media multitasking influences the consumer advertising processing, a starting point is to look at the way consumers process messages. People are assumed to have a limited capacity in the information they can process (Lang, 2000). Lang (2000) developed the limited capacity model of motivated mediated message processing (often abbreviated to LCM) to explain this. In this model, Lang explains how processing messages involve three sub-processes: encoding, storage, and retrieval (Lang, 2000; Segijn et al., 2016).
Lang (2000) defines the encoding subprocess as the selection of stimuli that are later stored as mental representations. The most important pieces of information are chosen to be stored. Storage involves linking of the encoded information to previously stored information, and retrieval is the activation, or bringing back of previously stored information (Lang,
2000). People generally use these cognitive capacities for information processing, but when the requirements from the message demand more resources than available, it results in cognitive overload. Information processing becomes even more difficult when there are two tasks happening simultaneously, as is the case with media multitasking. This requires more cognitive resources than performing a single task (Srivastava, 2013), and performance, in terms of recall and recognition have been shown to suffer as a result (Angell et al., 2016; Segijn et al., 2016; Srivastava, 2013). The LCM has been frequently studied. Studies include those by Angell et al., (2016) and Segijn et al., (2016), investigating recall and recognition of advertisements while participants were media multitasking. Participants who were media multitasking had a more difficult time recalling and recognizing the advertisement than participants who were not media multitasking (Angell et al., 2016). Segijn et al., (2016) argued that encoding and storage are affected during media multitasking, making recognition of the brand more difficult. This study indeed found that people who were media multitasking had a more difficult time recognizing the brand (Segijn et al., 2016).
Capacity Interference Theory and Limited Capacity Model of Information Processing. Two other theories in line with- and support of the LCM are the capacity interference theory (Armstrong & Chung, 2000) and the limited capacity model of information processing (Kahneman, 1973). These theories suggest that people have a limited cognitive capacity for attention and information processing that they can give to different cognitive tasks. When the cognitive load exceeds the attentional and processing capacity, the performance degrades (Jeong & Hwang, 2015). Studies in line with these two theories have shown that media multitasking had a detrimental effect on recall and recognition (Armstrong & Chung, 2000; Jeong & Hwang, 2012).
The detrimental effects that media multitasking can have on cognitive responses can be further explained by structural interference. While capacity interference refers to two or
more tasks competing for cognitive resources, structural interference is about two or more tasks occupying the same sensory channel (Jeong & Hwang, 2015). For example, when you are reading a text while watching television, you are utilizing the same sensory channel, namely the visual channel (Jeong & Hwang, 2015; Kahneman, 1973). Studies showed that comprehension and recognition were reduced during media multitasking due to structural interference (Jeong & Hwang, 2015; Pool, van der Voort, Beentjes & Koolstra, 2000). In the current experiment, the same channel is occupied in the media multitasking condition because participants read an article while simultaneously watching an advertisement.
Elaboration Likelihood Model. Next to the set of theories explained above, another theory to explain cognitive responses to media multitasking is the elaboration likelihood model (ELM). The ELM is best known and often used to explain affective responses, however research has shown that it is also a noteworthy model to explain cognitive responses to media multitasking. (Barden & petty, 2008; Jeong & Hwang, 2012; Zhang et al., 2010). The ELM is about the amount of effort someone uses in order to process and evaluate a message. It poses that when faced with a message, people process according to two possible routes – the central or peripheral route (Petty & Cacioppo, 1986). The central route involves a high level of elaboration, attention and conscious thinking. The peripheral route involves a low level of elaboration and conscious thinking. While peripherally processing, we pay less attention to the message and are more easily swayed by surface characteristics (Petty & Cacioppo, 1986).
Processing happens through the central route when people have the cognitive resources, motivation and opportunity to fully process a message. If not all of these factors are available or accessible, the peripheral route is likely to be taken. When people are media multitasking, cognitive overload causes central processing to be reduced while peripheral processing is induced (Jeong & Hwang, 2012). Previous research on media multitasking has
demonstrated that this peripheral processing leads to reduced comprehension of information (Jeong & Hwang, 2012), as well as attention (Barden & Petty, 2008), recall and recognition (Zhang et al., ).
Based on the limited capacity model of motivated mediated message processing, the capacity interference theory, the limited capacity model of information processing and the elaboration likelihood model, as well as previous empirical and scientific research in accordance with these theories, it is expected that:
Hypothesis 1: People who are media multitasking have more difficulty recalling and recognizing information than people who are not media multitasking (single-tasking).
Media Multitasking Effects on Affective Responses
Elaboration Likelihood Model. While media multitasking has shown and is expected to have negative effects on cognitive reactions, the affective reactions towards text and advertisements while media multitasking may differ. According to the elaboration likelihood model (ELM), while media multitasking, one is more likely to process information through the peripheral route (Jeong & Hwang, 2012). Peripheral processing reduces attention and comprehension, thereby reducing counterarguing of the message. Due to this, people are more likely to change their attitude and accept a message or perceive it as being real (Jeong & Hwang, 2012). A meta-analysis based on 212 studies by Jeong & Hwang (2016) supports this claim. The affective outcomes that they measured were: the agreement to messages, reduced counterarguing and attitude change. They found that media multitasking had an overall positive effect on these affective outcomes (2016). Several other studies investigated this relationship as well, supporting the notion that media multitasking distracts people from centrally processing a message, leading to peripheral processing. This led to a reduced
amount of attention being paid to a message (Barden & Petty, 2008) and higher advertisement evaluations (Chinchanackochai, Duff, & Sar, 2015).
Counterarguing Inhibition Hypothesis. A second theory to support how media multitasking influences affective reactions is the counterarguing inhibition hypothesis, developed by Keating & Brock (1974). Counterarguing refers to the resistance to a persuasive message (Segijn et al., 2016). Counterarguing is more likely to take place when a message is elaborately thought about. Elaboration of a message is more difficult while media multitasking because attention is divided over the different tasks. Counterarguing will therefore be reduced when media multitasking. Reduced counterarguing leads to an increased message acceptance (Moyer-Guse & Nabi, 2010; Segijn et al., 2016). Empirical support showed that media multitasking decreased counterarguing and resulted in a more positive brand attitude (Segijn et al., 2016). The meta-analysis conducted by Jeong & Hwang (2012) also supported the hypothesis, again showing that media multitasking reduced counterarguing.
Media multitasking may increase the text and advertisement likeability (attitude) due to the suppression of counterarguing and inducement of peripheral processing. The ELM and counterarguing inhibition hypothesis as well as previous research in accordance to these two theories have led to the formulation of the following hypothesis:
Hypothesis 2: People who are media multitasking are more likely to have a positive brand attitude (text likeability and advertisement likeability), than people who are not media multitasking (single-tasking).
The effects of media multitasking can be stronger or weaker in specific circumstances. One important distinction is the emotions used in the type of media that is shown. The role of emotions in media multitasking will be considered and further explored in the current research. Emotions actively facilitate information processing (Brocas & Carillo, 2007). Studies have shown that people rely heavily on emotions when making decisions with regard to advertisements (Bagozzi, Gopinath & Nyer, 1999; Barbosa, 2017). As media multitasking has become an integral part of our advertisement society, it is important to explore the moderating role that emotions could play on the relationship between media multitasking and cognitive as well as affective reactions.
Positive Emotions. Amusing, or positive advertisements are often defined as ‘smile-inducing stimuli’, intended to create attitude change (Duncan, 1979). Happiness is frequently used in advertisements and has shown to be an effective emotion for attitude change (Anastasiei & Chiosa, 2014; Bagozzi et al., 1999). The emotion happiness makes one feel good, has the possibility to make one smile or laugh (Duncan, 1979). There are several studies that have shown the effect of emotions on cognitive and affective reactions.
The effect of emotions on cognitive processes has been extensively studied and it has been found that the positive emotional state of someone influences information processing. Positive emotions influence retrieval and encoding such that positive stimuli cause better recall of information than negative stimuli (Isen, Shalker, Clark & Karp, 1978; Nasby & Yando, 1982 as cited by Bagozzi et al., 1999). Research has shown that positive words stimulate memory whilst negative words do not (Isen et al., 1978; Nasby & Yando, 1982). Encoding and retrieval clearly play an important role in media-multitasking. As explained earlier with the LCM, participants have a more difficult time recalling and remembering information while media multitasking, compared with when they are not media multitasking. As research has linked stronger cognitive responses to positive emotions, it will be expected
that positive advertisements, in comparison with negative advertisements, will result in stronger cognitive responses.
Another possible cognitive, as well as affective explanation with regard to positive emotions and information processing is given by the information processing approach (McGuire, 1979 as cited in Eisend, 2011). This approach says that when you are exposed to an advertisement, attention comes into play. When attention is grabbed, it allows you to elaborate on the information. Happiness has a high attention-attraction ability. It has been shown that happiness, in the form of humour, leads to more positive cognitions (Eisend, 2011). Besides attention, counterarguing plays a role as well in the information processing approach. Happiness has the ability to distract consumers from making counterarguments, making them more likely to accept a message (Eisend, 2011). The counterarguing inhibition hypothesis says that reduced counterarguing while media multitasking leads to increased message acceptance (Moyer-Guse & Nabi, 2010; Segijn et al., 2016). This suggests that when media multitasking is present, and happy advertisements are shown, consumers will rate these advertisements more positively, compared with when unhappy advertisements are shown. To further support the effect positive emotions have on affective responses, the simple evaluative condition is examined. Happiness creates an ‘affect transfer’. This means that happiness elicits affect, which in turn, is carried over to the general positive affective response to the advertisement (Eisend, 2011). Happiness suppresses negative affect due to distraction as well (Eisend, 2011).
Existing research has supported the notion of happy emotions having a positive influence on cognitive as well as affective responses to advertisements (Beard, 2005; Cline & Kellaris, 2007; Weinberger et al., 1995). It causes distraction, reduces counterarguing, enhances attitudes, positive affect, attention, recall and recognition (Bears, 2005; Eisend, 2011; Lewinski, Fransen & Tan, 2016). These factors play an important role when it comes
to media multitasking. In sum, the LCM poses that we have a limited capacity to process information (Lang, 2000). The ELM (Petty & Cacioppo, 1986) and the Counterarguing Inhibition Hypothesis (Keating & Brock, 1974) pose that media multitasking reduces counterarguing. Happiness is expected to facilitate this and result in stronger cognitive- and more positive affective responses, in comparison with media multitasking with negative advertisements, which are sad advertisements in this experiment.
Negative Emotions. In contrast to positive emotions, negative emotions refer to emotions which bring about some sort of unpleasant feeling like sadness or anger (Osaka et al., 2013). Negative affect has the potential to make tasks more difficult (Lewinski et al., 2016). Osaka et al., (2013) researched the influence of positive, negative and neutral texts with regard to their influence on working memory and cognitive performance. In this research, brain activity was measured during the encoding and retrieval phases of reading tests. The findings show activation in different brain areas related to working memory. Working memory holds information that is later used for processing and has a limited capacity (Osaka et al., 2013). It was found that activation in working memory was related to cognitive performance, and therefore suggests that emotions have an influence on working memory and capacity. When comparing negative to positive and neutral stimuli, it was found that the recognition accuracy; the extent to which the words remembered matched the stimuli; was significantly lower in the negative condition than in the positive and neutral condition. It was also found that the response time; the time it took to recall words; was significantly lower in the negative condition than in the positive and neutral condition. This indicates that negative emotions have a stronger negative influence on cognitive performance than positive emotions (Osaka et al., 2013). Further support is reached by the finding that negative stimuli have a more negative influence on information processing than positive stimuli. Negative stimuli negatively impact working memory, impairing recall and recognition (Tyng, Amin,
Saad & Malik 2017). The LCM poses that recall and recognition are diminished during media multitasking (Lang, 2000). This suggests that when people are media-multitasking and are doing so with negative stimuli (like sad advertisements), recall and recognition will be even more negatively affected.
Sadness is a negative emotion that is often used in advertisements (Oetting, 2018). Sadness, will therefore be used in the current experiment as the negative emotion- and counterpart of happiness. Support shows that sadness slows the cognitive system (Izard, 1993). This, with regard to information processing, leads to more central and elaborate processing, assumingly because one has to focus more on remembering information (Izard, 1993; Nabi, 1999). As there are practically no studies about the specific effects of sadness on information processing, there has not been an unambiguous explanation found for this. However, with regard to media-multitasking, a link can likely be made with the ELM. The ELM poses that central processing is reduced- and people are more likely to peripherally process information while media-multitasking. As Izard (1993) shows that sad information leads to central- as well as slowed processing, it will be expected that sad advertisements lead to worsened cognitive responses while media multitasking. The effect of sadness in advertisements on affective reactions towards the advertisements in programs was examined (Lacoste-Badle, Malek & Droulers, 2013). The study found that sadness had a negative effect on attitudes towards the brand (Lacoste-Badle et al., 2013). Further research on the effects of sadness on brand attitude is limited, thus current research hereby aims to touch upon new scientific grounds.
According to the existing literature, positive emotions generally enhance information processing and with that cognitive and affective responses, where negative emotions diminish information processing. With regard to media-multitasking and the theories explained above, the following hypotheses were formulated:
Hypothesis 3: People who are media multitasking have more difficulty recalling and recognizing information than people who are not media multitasking. This effect is stronger when a sad advertisement is shown compared with when a happy advertisement is shown.
Hypothesis 4: People who are media multitasking are more likely to have a positive brand attitude (text likeability and advertisement likeability), than people who are not media multitasking. This effect is stronger when a happy advertisement is shown compared with when a sad advertisement is shown.
The Role of Screen Differences
This research will also investigate the effect that smartphones and computers have on the relationship between media multitasking and cognitive, as well as affective reactions. There are two poignant differences between smartphones and computers, namely the screen size and the touch- and click ability.
Screen Size. The first difference between smartphones and computers considers screen size. The screen size of computers is clearly larger than that of smartphones. Previous research has consistently shown that a larger screen size has generally led to more positive effects with regard to affective responses (Hou et al., 2012; Kim, 2017; Kim & Sundar, 2014; Lombard, 1995). Larger screens led to more favourable impressions and attitudes of tasks completed (Hou et al., 2012), more positive ratings with regard to subjective experience and level of satisfaction (Kim, 2017), positive attitude (Kim & Sundar, 2014), and more positive emotional responses to what was shown on the screen (Lombard, 1995).
One explanation of these results are the hedonic and pragmatic qualities of different technology items. Hedonic quality refers to the positive subjective experience of a product,
like a smartphone being pretty, original or appealing (Kim, 2017). Pragmatic quality refers to the level of satisfaction, something being practical or useful (Kim, 2017). Hedonic and pragmatic qualities have shown to be promoted more by large screens in comparison with smaller screens. On one hand, large screens satisfy hedonic needs, such that the screen itself- or items on large screens, are generally found to be more attractive (Kim, 2017; Kim & Sundar, 2014). Large screens trigger what is said to be ‘the bigger the better heuristic’ (Hirschman & Holbrook 1982). On the other hand, it has been found that when objects are rated as more attractive, the overall usefulness and positive evaluations are strengthened as well, satisfying pragmatic qualities (Hou et al., 2012; Kim, 2017).
Research asking participants to watch television on different sized screens found that participants’ impressions of- and emotional response to the television shows were significantly more positive on larger screens (Lombard, 1995). Similar results were found in further research. Usage of larger screens led to a more positive media experience overall (Hou et al., 2012), more positive attitudes ( Kim & Sundar, 2014), and greater perceived attractiveness (Kim, 2017). With regard to the current experiment, the hedonic and pragmatic qualities of screens are promoted more by larger screens, which have been shown to result in more positive affective responses. It is therefore expected that participants who are media multitasking will evaluate the advertisement and text more positively when they see it on a computer screen compared with when they see it on a smartphone.
To reach further support with regard to affective and cognitive responses, the literature investigated the role of information processing in relation to differences between screens. Earlier, the ELM was explained. In line with the ELM, another theory with regard to dual processing has specifically reached past support concerning screen differences, namely the Heuristic Systematic Model (HSM) (Hou et al., 2012; Kim & Sundar, 2016). Argumentation regarding the HSM is in line with argumentation concerning the ELM, such
that the two modes of information processing, systematic (central) and heuristic (peripheral) processing, function accordingly (Chaiken, 1980; Chaiken, Liberman, & Eagly, 1989). These theories pose that people will automatically heuristically process information if obvious heuristics are available. Kim & Sundar (2008; 2016) tested this dual processing theory with regard to screen size. They found that information via small screens was processed systematically, whereas information via larger screens was processed heuristically. This, again, triggered ‘the bigger the better heuristic’ and led to more positive evaluations, and acceptance of the message (Kim & Sundar, 2016).
Past research has shown that information is processed heuristically, or peripherally, when one is media multitasking. This leads to more positive attitudes. Thus, when one is media multitasking on a larger screen, information will be processed heuristically and this will result in more positive attitudes. However, the ELM and HSM also pose that peripheral processing leads to a reduced comprehension of information, attention, and recall (Barden & Petty, 2008; Jeong & Hwang, 2012; Zhang et al., 2010). This suggests that cognitive responses will diminish on larger screens in comparison with smaller screens while media multitasking. This is due to the fact that closer attention must be paid when completing tasks on smaller screens, thus processing information centrally, or systematically.
Touch- and Click Ability. The second difference between smartphones and computers regards the touch- or click ability. Where most computers do not have the option to physically touch the screen or zoom in with your fingers, smartphones do have that option. Wei Shi & Kalyanam (2018) studied the usage of information touch features on apps and their influence on engagement. Informational touch features are features like being able to zoom in with your finger or clicking on links or photos to change the view. Research has shown that information touch ability of apps improves stay-likelihood, and implies stronger purchase intent, as well as closer attention being paid to the item being zoomed-in- or clicked
on, in comparison with navigational touch features. Navigational touch means simply swiping and scrolling (Wei Shi & Kalyanam, 2018).
There is little research on the specific aspect of informational touch and the influence of this on cognitive responses. However, due to the closer attention that is paid when you are able to click on- or zoom into something on a page, it is expected that cognitive responses to media-multitasking will be more positive when this is the case. Such is the case when it comes to smartphones, as you are able to zoom in, and scroll down the page with your fingers. The existing literature concerning screen differences, as well as supportive research studies led to the formulation of the following hypotheses:
Hypothesis 5: People who are media multitasking have more difficulty recalling and recognizing information than people who are not media multitasking. This effect is stronger on larger screens (computer) in comparison with smaller screens (smartphones)
Hypothesis 6: People who are media multitasking are more likely to have a positive brand attitude (text likeability and advertisement likeability), than people who are not media multitasking. This effect is stronger on larger screens (computer) in comparison with smaller screens (smartphones).
Figure 1
Media multitasking effects on cognitive and affective responses. The moderating role of emotions and screen differences.
Method Sample
A total of 95 young adults between the ages of 18 and 26 (M = 22.63, SD = 1.95) completed the experiment. Initially, 200 respondents filled in the questionnaire. When checking for missing values, however, many respondents were eliminated. 89 respondents either missed essential questions or did not finish the questionnaire. Further, 2 respondents did not fall in the age category that was pre-determined, thus were eliminated as well. The third- and last check, was the time spent on the stimulus material. It had been pre-determined that participants should have spent at least 60 seconds on the material, to allow for an accurate exposure to the manipulation material. This was due to the fact that you had to read a text and watch an advertisement. 14 respondents spent less than 60 seconds on the stimulus page, with a mean time of 23.69 (SD = 19.82) seconds on the page. These respondents were therefore eliminated as well.
The participants were recruited through convenience sampling, distributing questionnaires on social media channels like Facebook, WhatsApp, and LinkedIn, as well as recruitment via e-mail. 67 (70.5%) participants were females, and 28 (29.5%) participants were males. The nationality of the participants was homogenous, the study was conducted in the Netherlands and all participants were Dutch. Thus, part of the stimulus material consisted of Dutch information. This allowed avoidance of disassociations with brands and amount of knowledge of the brands used in the experiment.
Design
The study consisted of an online experiment. Although the aim was not to test a three-way interaction, two different moderators were tested in the experiment. Therefore, formally a 2 (media-multitasking vs. single-tasking) x 2 (positive emotion vs. negative emotion) x 2 (smartphone vs. computer) between-subjects design was conducted. In essence, two separate 2 (media multitasking vs. single-tasking) x 2 (happy vs. sad) (1) and 2 (media multitasking vs. single-tasking) x 2 (smartphone vs. computer) (2) designs were conducted and analysed. Participants were proportionally divided among the stimulus conditions, condition (1): Chi2(3) = 1.884, p = .597 and condition (2):Chi2(3) = 2.698, p = .441.
In the media multitasking condition, participants were prompted to switch quickly between two tasks, which is in line with the definition of media multitasking (Jeong & Hwang, 2016). In the single-tasking condition, participants completed one task after another. In both the media multitasking- and single-tasking conditions, participants were either shown a happy advertisement (positive emotion) or a sad advertisement (negative emotion). Participants were randomly assigned such that some participants were asked to perform the experiment on their smartphone, while others were asked to perform the experiment on their
computer. 45 (47.4%) participants completed the experiment on their smartphone, while 50 (52.6%) participants completed it on their computer.
Stimulus
Independent Variable. The independent variable in this experiment was media-multitasking (or single-tasking). This variable was manipulated such that some participants received a media-multitask manipulated text, whereas other participants received a single-tasking manipulated text. Media-multisingle-tasking, in this research, referred to the extent to which people could read and remember a text, while quickly switching to an advertisement (Brasel & Gips, 2017; Jeong & Hwang, 2016). The advertisements were placed such that participants would read half of the article, and then be interrupted by the advertisement. After the advertisement, participants would finish reading the article. This interruption which caused quick switching between two tasks is called in-stream advertising (Kastidou & Cohen, 2006; Thawani, Gopalan & Sridhar, 2004). In-stream advertising is similar to commercial breaks, it interrupts a television program for example. This type of media-multitasking can also be called sequential multitasking (Salvucci & Taatgen, 2008). It entails spending time on one task and then switching to another task (Salvucci & Taatgen, 2008). Single-tasking, in this research, is referred to as participants reading the text completely, and receiving the advertisement upon completion of the text. There was no time limit to completing the experiment. However, a hidden timer was installed to measure how long people spent on the stimulus page on average. This was done so it could be controlled for in the analyses. The average time that people spent on the stimulus page was 159.37 seconds (around two and a half minutes).
The content of the text consisted of information taken from the website www.nu.nl. The text was about the coffee manufacturer Nespresso wanting to make more sustainable
coffee capsules (nu.nl, 2018). Since 2000, worldwide turnover of Nespresso has been growing with 30% (Nespresso.com). When actor George Clooney became spokesperson for Nespresso, turnover even doubled (Nespresso.com). Along with that, the Netherlands is known to be a coffee-drinking country. Research has shown that Dutch people drink 2,4 cups of coffee a day, putting them in the top 3 most coffee consuming countries in Europe (Geraedts, 2014). For these reasons, it was decided that a text about Nespresso would be a good choice. While the brand Nespresso was expected to be relatively known, such that the text was interesting to read, the information in the text was expected to be relatively unknown to the majority of the participants. This would prevent bias and pre-existing attitudes or knowledge. The text took around one minute to read.
Moderators. The first moderator in this experiment was emotions. The emotions chosen for this study were happiness and sadness. They were manipulated such that some participants received a humorous advertisement and other participants received a sad advertisement. The advertisements had to meet the following conditions. They had to be advertisements from the same brand, the same length, and portraying the two chosen emotions. Two advertisements were found for the car brand Chevrolet. The sad advertisement was about the love of a dog and its owner, the journey they take together in life and how that journey comes to an end (YouTube, 2014). The happy advertisement was about a recently graduated boy receiving a present from his parents for his graduation. He thinks he is getting a car when he is really getting a fridge (YouTube, 2011). The advertisements were in in English, so it was decided consciously that the questionnaire would be in English as well. The article was, however in Dutch. The information in the article was complicated and this ensured that everyone understood the information, as well as knowing the brand (Nespresso). The recruitment text for the participants asked you to participate in the study if
you were confident in both English and Dutch, so this would pose no problem with regard to the results.
The second moderator in this experiment was screen differences. The two screens chosen for this experiment were smartphone and computer. At the start of the experiment, participants were asked to either complete the experiment on their smartphone or on their computer. At the end of the experiment, the question was asked: “I completed this experiment on my…” with two options: (a) smartphone or (b) computer. All participants remained on the same device they started the experiment with.
Procedure
The experiment was designed such that participants were asked to take part in their natural environment, to approximate a natural multitasking environment. Before commencing with the experiment, participants read and signed an informed consent form. The participants were given a cover story about taking part in an experiment measuring media use in daily life. At the beginning of the study, participants were either asked to perform the experiment on their smartphone or on their computer. At the end of the questionnaire, participants were asked once again if they completed the study on their smartphone or on their computer.
Upon starting the experiment, participants were asked to read a text closely and watch an advertisement. The text, as explained before, was taken from the Dutch website www.nu.nl. The topic of the text being Nespresso. The text and advertisement were shown as explained before. Some participants received a text where they had to multitask (being interrupted by an advertisement while reading a text), while other participants were able to finish reading the text before seeing the advertisement.
After the stimulus material, participants were to fill out a questionnaire. It was explained that this questionnaire would ask a couple of questions with regard to what they
just read and saw. The questionnaire asked questions regarding emotions, recall, recognition, text likeability, advertisement likeability and background variables (i.e., age and gender).
Measures
Dependent Variables. The dependent variables consisted of cognitive and affective responses. Cognitive responses were measured through recall and recognition. Affective responses were measured through text likeability and advertisement likeability.
Cognitive Responses. The first dependent variable was cognitive reactions. This
variable was measured in the form of recall and recognition; being able to remember something that you read or saw and being able to recognize it. Recall is a proven indicator of memory retrieval and later stages of information processing (Kazakova et al., 2016; Lang, 2000). Recognition taps more into the earlier phase of information processing, memory encoding (Kazakova et al., 2016; Lang, 2000).
Recall was measured by asking participants which brands they saw. It was an open ended question. Participants either answered correctly if they recalled at least one of the two brands (Nespresso or Chevrolet), or answered wrong if they could not remember both brands. This method was based on research measuring recall in the past (Voorveld, 2011). Recognition was measured based on methods used in earlier experiments measuring cognitive responses (Segijn et al., 2016; Srivastava, 2013; Voorveld, 2011). Participants received a multiple choice question, a list of three options of items that came up in the text. They were then asked to choose the item they could remember. Participants were assigned a score of 1 when they remembered the correct information, and a score of 0 when they remembered the wrong information. There were three multiple choice questions, thus the responses ranged from: 0, 1, 2 to 3. 54.5% of the participants recalled at least one of the
three multiple choice questions correctly. Thus, 45.5% of the participants recalled none of the questions correctly.
Affective Responses. The second dependent variable that was measured in this
experiment was affective responses, the attitude participants have towards a brand. This was measured in the form of text likeability and advertisement likeability. Brand attitude regards the standpoint that people have with relation to a certain brand or product (Segijn et al., 2016). To measure the attitude participants had towards the text (Nespresso) and the advertisement (Chevrolet), a five-point semantic differential scale was used. This scale was modified from the scale created by Holbrook & Rajeev (1987), and research by Chang & Thornson (2004), Segijn et al., (2016) and Voorveld et al., (2011). The answer responses were: very likeable/not very likeable, interesting/not interesting, good/bad, appealing/not appealing, and positive/negative. The scale was proven reliable for measuring the brand attitude towards the text (M = 3.22, SD = .72, = .85) and the brand attitude towards the advertisement (M = 3.33, 1.05, = .94).
Moderator. To measure emotions, whether participants perceived the advertisement as happy or sad, the Discrete Emotions Questionnaire (DEQ) was used (Harmon-Jones, Bastion, & Harmon-Jones, 2016). Past research has proven its validity with regard to positive and negative emotions. This scale was modified such that it fit the current research. The scale asked participants the following question: While watching the advertisement, to what extent did you experience these emotions in the advertisement? Participants were asked to respond on a scale from 1 (not at all) to 7 (an extreme amount). The words to describe happiness were: happy, satisfaction, enjoyment and liking. The words to describe sadness were: sad, grief, lonely and empty. The scale proved to be a reliable measure of emotion, Cronbach’s alpha for the DEQ was = .85 (M = 3.49, SD = 1.26).
Manipulation check
Several manipulation checks were conducted to determine the internal validity of the experiment. The manipulation checks were conducted to see whether the manipulation was successful, and participants were able to identify to their right condition. On the basis of these checks, we were able to determine whether the study measured what it was supposed to measure.
Independent Variable. For the independent variable of media multitasking versus single-tasking, two questions were asked. Namely, during the experiment, I felt like I had to switch between tasks and during the experiment, I felt that I needed to divide my attention. These two questions were measured based on five answer possibilities, ranging from 1 (completely disagree) to 5 (completely agree). The scale was proven reliable, thus measuring what it was supposed to measure (M = 3.05, SD = 1.22, = .89).
Moderator Emotion. For the moderator emotion (happy versus sad), two manipulation questions were asked. The questions were: the advertisement I just saw was geared towards happiness and the advertisement I just saw was geared towards sadness. These questions were measured using a Likert scale, again ranging from 1 (completely disagree) to 5 (completely agree). For this manipulation check, there were no existing scales that were applicable to this research. For that reason, existing Likert scales were used and adjusted such that they were applicable for the manipulation checks in this research. The scale to measure emotion was also shown to be a reliable measure (M = 3.07, 1.26, = .88).
Moderator Screen Differences. At the start of the experiment, participants were asked to complete the experiment either on their smartphone or computer. Participants had to select either option. At the end of the questionnaire, a manipulation check was done regarding the different screens. The following question was asked: “I have completed this
study on my (a) smartphone, (b) computer. Analysis showed that all participants used the same device throughout the entire experiment.
Control Variables
In the current research, the variables of age, gender and time that was spent on the stimulus material were kept constant. These three variables acted as control variables and were measured in order to make sure that they did not have a confounding effect on the relationships measured.
Results Control Variables
To check whether it was necessary to control for certain variable items, several analyses were conducted. A chi-test was conducted for gender (Chi2(3) = 2.215, p = .529), showing that gender was proportionally divided over the conditions. T-tests were conducted for age (t = 1.643, p = .408) and for time spent on the stimulus page (t = 1.398, p = .608). The analyses showed that they were both not significantly different between the two types of mediums that were used. Thus, if there is a difference between the conditions, this will be due to the conditions (manipulation), not due to the control variables.
Cognitive Responses: The Influence of Multitasking, Emotions and Screen Size on Recall
The first hypothesis stated that people who are media multitasking, have a more difficult time recalling information than people who are not media multitasking (single-tasking). Recall was measured on a dichotomous scale, getting either nothing correct (0) or one of the two brands correct (1). Due to the fact that there were 90 participants who recalled
at least one brand, only 5 participants were left who were not able to recall any brand. Therefore, a logistic regression was not possible. The hypotheses with regard to recall could not be tested.
Cognitive Responses: The Influence of Multitasking, Emotions and Screen Size on Recognition
To test whether participants have a more difficult time recognizing information when they are media multitasking compared to when they are single-tasking (H1), whether this difference is smaller when a happy advertisement is shown in comparison with when a sad advertisement is shown (H3 and H4), and whether this difference is larger when media multitasking (or single-tasking) on a computer in comparison to a smartphone (H5) an ANOVA was carried out.
An ANOVA with media multitasking, emotions and screen size as factors, and recognition as the dependent variable showed that there was no main effect of multitasking on recognition F (1, .686) = 1.895, p = .467. A main effect of emotion on recognition was also not found F (1, 1) = .113, p = .794, just as there was no main effect concerning screen differences on recognition F (1, 1) = .527, p = .600. An interaction effect between multitasking and emotion on recognition was not found either F (1, 89) = .056, p = .814. There was, however, an interaction effect found with regard to screen differences F (1, 89) = 5.627, p = .019.
This interaction effect was explored further by looking at the means of the different groups. The analysis looked at the differences between the conditions, to see whether the difference between media multitasking versus single-tasking was larger on a computer or on a smartphone. The combination multitasking x computer led to a mean of M = .546 (SD = .350) and the combination single-tasking x computer led to a mean of M = .845 (SD = .248).
The difference here is .299. The combination multitasking x smartphone led to a mean of M = .772 (SD = .224), and the combination single-tasking x smartphone to a mean of M = .808 (SD = .214). The difference between these conditions is .036. Hereby, it can be concluded that the difference in the computer conditions is noticeably larger. Thus, the effect of media multitasking versus single-tasking on a computer screen is stronger with regard to recognition than media multitasking versus single-tasking on a smartphone. When participants were media multitasking on a computer screen, they were less likely to recognize a message. Conversely, when participants were single-tasking on a computer screen, they were more likely to recognize a message. In sum, the hypothesis stating that people who are media multitasking are less likely to recognize a message, and this effect is stronger on larger screens (computer) in comparison to smaller screens (smartphone) is hereby accepted.
Affective Responses: The Influence of Multitasking, Emotions and Screen Size on Text Likeability
To test whether media multitasking would lead to a more positive affective response with regard to the text, than single-tasking, and to test whether there was an interaction effect with emotion and screen size, an ANOVA was conducted. The ANOVA with multitasking,
0,4 0,5 0,6 0,7 0,8 0,9 Single-tasking Multitasking M ea n
Figure 1: Interaction effect media multitasking and screen differences on recognition
emotions and screen size as factors and text likeability as the dependent variable showed that there was no main effect of multitasking on text likeability F (1, .764) = .164, p = .770. A main effect of emotion on text likeability was not found either F (1, 1) = 4.514, p = .280. This also counts for the main effect of screens on text likeability F (1, 1) = .001, p = .985. An interaction effect between multitasking and emotion on text likeability was not found either F (1, 89) = .567, p = .454. An interaction effect between multitasking and screen differences on text likeability was, however, found F (1, 89) = 3.977, p = .049. To explore this interaction effect further, a closer look was taken at the differences between the conditions.
Firstly, the difference between the combination multitasking x computer (M = 3.47, SD = .64) and single-tasking x computer (M = 3.01, SD = .66) was 0.47. The difference between the combination multitasking x smartphone (M = 3.08, SD = .89) and the combination single-tasking x smartphone (M = 3.32, SD = .711) lay at 0.24. Thus, the difference in the computer condition is larger than in the smartphone condition. When participants were media multitasking on a computer screen, they had a more positive attitude towards the text compared with media multitasking on a smartphone. When participants were single-tasking on a computer screen, they had a less positive attitude towards the text in comparison with single-tasking on a smartphone. With that, the hypothesis stating that people who are media multitasking have a more positive brand attitude than people who are not media multitasking, and this effect is stronger on larger screens than on smaller screens, is partially accepted.
Affective Responses: The Influence of Multitasking, Emotions and Screen Size on Advertisement Likeability
To test whether media multitasking would lead to a more positive affective response with regard to the advertisement, than single-tasking, and to test whether there was an interaction effect with emotion and screen size, an ANOVA was conducted. The ANOVA, with multitasking, emotions and screen size as factors and advertisement likeability as the dependent variable showed that there was no main effect of multitasking on text likeability F (1, .419) = 1.594, p = .581. A main effect of emotion on advertisement likeability was not found either F (1, 1) = .159, p = .759. This also counts for the main effect of screens on advertisement likeability F (1, 1) = .004, p = .959. An interaction effect between multitasking and emotion on advertisement likeability proved to be insignificant F (1, 89) = 2.338, p = .126. Lastly, an interaction effect between multitasking and screen differences on advertisement likeability was not found F (1,89) = .196, p = .659. Thus, all hypotheses with regard to advertisement likeability were rejected.
2,7 2,8 2,9 3 3,1 3,2 3,3 3,4 3,5 3,6 Single-tasking Multitasking M ea n
Figure 2: Interaction effect between media multitasking and screen differences on text likeability
Smartphone Computer
Discussion
The general aim of this research was to determine the influence of emotions conveyed in an advertisement, and screen differences, on the relationship between media multitasking and recall, recognition, text- and advertisement likeability towards the text and the advertisement. This study showed that multitasking had no direct effect on recognition, text likeability or advertisement likeability (brand attitude). This study also showed that media multitasking in combination with happy or sad emotions had no effect on recognition, text likeability or advertisement likeability. However, it was indeed found that media multitasking on a computer in comparison with a smartphone had an influence with regard to recognition as well as text likeability. Such that, when participants were media multitasking on a computer screen, they were less likely to recognize a message, but they had a more positive attitude towards the text in comparison with media multitasking on a smartphone. Conversely, when participants were single-tasking on a computer screen, they were more likely to recognize a message, but they had a less positive attitude towards the text in comparison with single-tasking on a smartphone.
The first aim of the study was to show that media multitasking had a direct influence on recall, recognition, text- and advertisement likeability. The limited capacity model and elaboration likelihood model theorized that recall and recognition would diminish while media multitasking (Lang, 2000; Petty & Cacioppo, 1986; Segijn et al., 2016). On the other hand, the ELM and counterarguing inhibition hypothesis thought text- and advertisement likeability to increase due to media multitasking (Jeong & Hwang, 2012; Keating & Brock, 1974; Petty & Cacioppo, 1986). Thus, contrary to these theories and empirical results (Angell et al., 2016; Barden & Petty, 2008; Jeong & Hwang, 2011; 2015; 2016; Moyer-Guse & Nabi, 2010; Pool et al., 2000; Segijn et al., 2016; Voorveld, 2011; Zhang et al., 2010), there were no effects found of multitasking on recall, recognition or text- and advertisement likeability.
A plausible explanation as to why these effects were not found, is the fact that the experiment was not executed in a controlled setting. Therefore, we were not able to explicitly control for tasking and media multitasking. It is therefore possible that participants in the single-tasking condition were still multisingle-tasking, even if it was not on the device they were performing the experiment on. For example, listening to music or even making a sandwich. The participants in the multitasking condition could have just as likely been performing other tasks, besides the task they were asked to do for the experiment. This could have influenced the results. Future research should conduct an experiment in a controlled environment, such that the participants who are single-tasking are truly only doing one task. Conversely, participants who are media-multitasking are undeniably only media-multitasking.
A second aim was to determine whether happy or sad advertisements had a different effect on the role between media multitasking on recall, recognition, text- and advertisement likeability. No differences were found. Although research has shown emotions to play an important role in advertisements (Bagozzi et al., 1999; Barbosa, 2017), the current research does not support this claim. Expectations that positive emotions would increase recall, recognition (Isen et al., 1978; Nasby et al., 1982) and text- and advertisement likeability (Bears, 2005; Eisend, 2011) while media multitasking, were not met. There are several possible reasons as to why these effects were not found. A first explanation builds on the information processing approach (McGuire, 1979 as cited by Eisend, 2011). It talks about the idea that attention must be drawn to be able to get the desired reaction (Eisend, 2011; Pieters & Wedel, 2004). When attention is grabbed, advertisements have a larger influence, they are more easily remembered and liked (Pieters & Wedel, 2004, Solomon, 2013). Pieters & Wedel (2004) studied the importance of attention in advertising through an eye-tracking experiment. On the basis of their study, they discussed important factors and ways to increase the attention-grabbing possibility in advertisements. They suggest that the size of the brand name
and text are important, larger is better. Along with that, product familiarity is a crucial factor as well, when consumers know a brand, they are more likely to pay attention to it (Pieter & Wedel, 2004). Taking this into account with regard to the current study, the brand name of the advertisement (Chevrolet) only came into view at the very end in small letters. Future studies should do a pre-test to know whether there is familiarity with regard to the brand, and the brand name should be made clear- and large enough such that it grabs attention. Spillover effects are more likely to occur in this case (Ahluwalia, Unnava & Burnkrant, 2001). Spillover effects are secondary effects that follow a primary effect (2001). Research has shown that attention and familiarity with a brand allows for positive spillover effects (2001) and result in more positive brand attitudes (Ross & Hajjat, 2016). A second explanation for the insignificant effects is the possibility that the advertisements did not cause strong emotional reactions that were expected beforehand. The average response on the emotion scale lay at ‘somewhat’, ranging from ‘slightly’ to ‘moderately’. There were barely any responses that showed an extreme positive or negative emotional reaction. To prevent this in future studies, a pre-test should be administered with several different brands. This way, one can determine which brands elicit the greatest emotion with participants and should then be used in further study.
A third aim was to measure whether completion of the experiment on a smartphone or on a computer would result in differences between the effect of media multitasking on recall, recognition, text- and advertisement likeability. Indeed, differences were found. There were as many as two effects with regard to screen differences. Analysis showed that when participants were media multitasking on a computer screen, they were less likely to recognize a message in comparison to when participants were media multitasking on a smartphone. Conversely, when participants were single-tasking on a computer screen, they were more likely to recognize a message in comparison to when they were single-tasking on a
smartphone. This specific interaction effect has not yet been found in previous research, but supports the heuristic systematic model (Hou et al., 2012; Kim & Sundar, 2016) as well as the elaboration likelihood model (Barden & Petty, 2008; Jeong & Hwang, 2012; Zhang et al., 2010). The HSM and ELM pose that heuristic/peripheral processing leads to worsened recognition. Additional research showed that information is likely to be processed heuristically on larger screens (Kim & Sundar, 2014), and systematically on smaller screens (Kim & Sundar, 2014) – further strengthening the results that were found. The current research also showed that when participants were media multitasking on a computer screen, they had a more positive attitude towards the text in comparison with when they were media multitasking on a smartphone. Conversely, when participants were single-tasking on a computer screen, they had a less positive attitude towards the text in comparison with when they were single-tasking on a smartphone. Thus, research has clearly shown that screen size matters, not only for cognitive, but for affective responses as well. The current research supports existing literature. Literature has posed that a larger screen size has led to positive effects with regard to affective responses (Hou et al., 2012; Kim, 2017; Kim & Sundar, 2014; Lombard, 1995). This is due to hedonic and pragmatic qualities and ‘the bigger the better heuristic’ that are promoted by larger screens (Hirschman & Holbrook, 1982; Kim, 2017; Kim & Sundar, 2014). The HSM further supports this, by saying that larger screens cause heuristic processing, due to cues that are more easily available – which then result in positive affective responses (Hou et al., 2012; Kim & Sundar, 2016). The current research supports ongoing literature with regard to screen differences, therefore contributing to scientific literature.
There are several limitations in this study that give opportunity to future research. Besides the obvious limitations of a small sample size and the use of a convenience sample, there were two important limitations with regard to the scales that were used that will be discussed below.
The first limitation lies in the scale that was used for measuring recall. Recall was measured on the basis of one question only and immediately after the stimulus material was presented. The results showed that the question was too easy, as nearly all participants answered the question correctly- remembering at least one brand. Goldstein (2011) investigated the relationship between memory and recall, and showed that recall is better tested after some delay (Goldstein, 2011). Research found that memory decreased after some time had passed and several tasks were completed. This is called proactive interference (PI), when previously learned information interferes with new information (Goldstein, 2011). Recall is often measured in laboratory settings (Goldstein, 2011; Kazakova et al., 2016). This allows not only for a larger number of items to be shown, it also allows the researcher to create some delay in answering time, as this will be something you can control for in a laboratory setting. It has been shown as well that information is encoded along with its context, this is called encoding specificity (Goldstein, 2011). When controlling the environment, the context in which the information was first seen and will later be recalled can remain the same. Due to time- and money constraints, it was not possible to measure recall after some delay, as well as measuring it on the basis of a larger number of items. Future research should therefore make use of a controlled environment. It would also be useful to test recall immediately after exposure to stimulus material and with some delay. This would allow for a useful comparison and probably lead to more conclusive results with regard to recall. An additional point to think about for future research, that was not considered in the current experiment, is intelligence. Intelligence was not measured in the
