E-editions on Tablets:
Examining the Influences of Print versus Digital Magazines, Involvement and Interactivity on Memory of Editorial and Advertising Content
University of Amsterdam Graduate School of Communication Research master in Communication Science
Thesis supervisor: Dr. Hilde Voorveld Student: Anneke Lisette Penders
Student number: 10439781 Date of completion: 26-06-2014
Acknowledgments
I would like to express my gratitude to Professor Dr. Peter Neijens and in particular my research supervisor Dr. Hilde Voorveld, for enabling me to conduct this study. Moreover, I would like to thank Hilde for her enthusiastic encouragement and valuable and constructive suggestions of this research work. I think I was very lucky to have you as my supervisor. Lastly, I would also like to thank Sanoma media for making the employed magazines available for this study.
Abstract
The aim of this study was to investigate the influence of medium type, involvement and interactivity on memory regarding editorial and advertising content. This was tested in two experiments: in the first experiment a less interactive magazine was employed and in the second experiment a highly interactive magazine was employed. Results showed that tablet was more effective than print regarding recognition of advertising content, while print was better than tablet regarding recall of editorial content in the first experiment. The second experiment showed a superiority of tablet regarding both types of content. In the first experiment, print was especially more effective than tablet for low involved individuals in memorizing advertising content, whereas in the second experiment tablet was especially more effective for high involved individuals in memorizing advertising content. Moreover, tablet readers did not recognize the interactive articles and advertisements more often than the non-interactive articles/advertisements. Hence, it can be concluded that although reading an e-edition on a tablet carries particular advantages and is especially more beneficial in case of the highly interactive magazine, current research shows that other differences between the
magazines and content should also be taken into account.
Introduction
In the last years there has been a sustained decline in magazine circulation, which resulted in a huge reorganization at the biggest magazine publisher of the Netherlands, Sanoma, which as a result was urged to abrogate a large number of well known magazine titles (Volkskrant, 29 Oktober, 2013). These developments are not only caused by the economic malaise, but also because of the digitalization the current market is facing (Volkskrant, 29 Oktober, 2013). This digitalization is caused by an evolution of new communication technologies and has dramatically changed the landscape of the media industry (Yoo, 2011). Therefore, combined with the year-on-year decline in circulation and advertising revenue, print media perceive the Internet as one of the key strategies to regain readership.
Moreover, currently half of the Dutch population owns a tablet and this has increased heavily over the last years (GfK Intomart, 2013). The total amount of tablet users is expected to increase even more next year, as more than a million people indicate that they are willing to purchase a tablet (GfK Intomart, 2013). In addition, more and more people also read
magazines and newspapers on their tablet. Reading a newspaper on a tablet for example is increased from 21% to 31% since 2011, whereas reading a magazine on a tablet is increased from 2% in 2011 to 11% in 2013 (GfK Intomart, 2013). Though magazine circulation on print is declining, the increasing media use on tablets may be seen as an opportunity for the
challenges the media branch is facing.
A result of the current developments is that print brands are increasingly digitally present (Intomart GfK, 2013 in: Neijens & Voorveld, 2013). This digital presence of print brands do not only come in forms of news sites or brand websites of a magazine title, newsletters and alerts distributed via email, but also in the form of a digital reproduction (Neijens &
Voorveld, 2013). These digital reproductions or ‘E-editions’, “are replicas of the printed versions as far as content and lay-out is concerned, with added navigation features for
selection of content and interactive features for additional content, communication, and transaction” (Chung, 2008, in: Neijens & Voorveld, 2013). Together with the increasing media use on tablets, magazine publishers perceive these e-editions as a solution to increase magazine circulation, reverse advertising revenue and to regain readership (Neijens & Voorveld, 2013).
It is however unknown whether such e-editions are as effective as print. Previous studies have examined the differences between print media and screen/online versions on several concepts such as attention, memory and learning and overall these studies conclude that print outperforms screen/online version (e.g. Sundar, Narayan, Obregon & Uppal, 1998;
Tewksbury & Althaus, 2000; Eveland & Dunwoody, 2001; Eveland & Dunwoody, 2002; Althaus & Tewksbury, 2002; Jones, Pentecost & Requena, 2005; Magee, 2012). Though, most studies focused on non-persuasive content (e.g. newspapers) and to date, no study has examined the differences between print magazines and their online edition on a tablet
regarding memory of editorial and advertising content. Therefore, the first aim of this study is to examine the influence of medium type regarding memory of editorial and advertising content.
A possible explanation for the differences between print and screen/online is interactivity. Interactivity is perceived as an unique characteristic of an e-edition on a tablet, as an e-edition on a tablet in itself is more interactive than a magazine on print, but also because particular interactive features are incorporated within the editorial and advertising content. However, previous research on the effectiveness of interactivity is contrasting. Whereas some argue that interactivity positively affects satisfaction, motivation, sense of fun, learning and persuasion (Chung, 2008; Liu & Shrum, 2009), others argue that interactivity can have detrimental effects (e.g. Bezjian-Avery, Calder & Iacobucci, 1998; Haseman, Nuipolatoglu &
what extent the influence of medium type on memory for editorial and advertising content is moderated by interactivity.
However, effects might not always be the same and may depend on involvement. Depending on whether an individual is low or high involved appears to result in differing effects (e.g. Petty, Cacioppo & Schumann, 1983; Liu & Shrum, 2009; Moorman, Neijens, & Smit, 2007). On the one hand, as an e-edition requires more cognitive resources due to interactivity, it is expected that reading an e-edition on a tablet is more effective for high involved individuals, because they are motivated and able to process information sufficiently (Liu & Shrum 2009). Print, on the other hand, may be more effective for low involved individuals, as this medium does not include features that may cause a cognitive burden (Magee, 2012). Therefore, the third aim is to examine to what extent the influence of medium type on memory for editorial and advertising content is moderated by involvement.
Theoretical framework
The influence of medium type on memory of editorial and advertising content
Most previous research on the differences between traditional and new media has focused on comparing media with different sensory modes (Jones et al., 2005). Overall, print media and the Internet are actually perceived as “largely similar considering the senses stimulated, the speed and control of information transfer, as consumers can process the information at their own pace and sequence” (Dijkstra, Buijtels & van Raaij, 2005, p. 378). However, in contrast to print media, online media can combine both written and multimedia content (Magee, 2012), which may result in using different senses when exposed to online media. Moreover, online media appear to differ from print considering the way humans respond to stimulation at three levels: physically, perceptually and cognitively (Dillon, 2004).
When reading, “documents must be handled and manipulated (physical level), before the reader can visually perceive the written information (perceptual level) and thus make sense of the material it contains (cognitive level)” (Dillon, 2004, p. 51). The first basic difference is that an e-edition on a tablet is physically different from a print magazine. Physical differences include stimulus characteristics such as orientation, aspect ratio, display size, and handling and manipulation (Dillon, 2004). Orientation means that paper has an advantage over screen, because it can easily be “picked up and oriented to suit the reader” (Dillon, 2004, p. 52). However, tablets are considered as very convenient as well. Aspect ratio refers to the
relationship of width to height, for example, “paper sizes are higher than they are wider, while the opposite is true for typical screen displays” (Dillon, 2004, p. 53). Though it can be argued that tablets are more similar to print regarding this aspect. Lastly, print and tablet differ regarding size and in contrast to print, content on tablets can be handled and manipulated (Dillon, 2004). These physical differences, which may for example cause a difference in the sensation of touch, can subsequently cause differences in memory (Magee, 2012).
Second, perceptual differences indicate how a (digital) magazine is visually perceived. These perceptual differences are aimed at image quality (Dillon, 2004). This includes
determinants such as fading and flickering characters, screen dynamics, visual angle in which the magazine is read, image polarity (i.e. contrast) and display characteristics (e.g. character size, line spacing and character spacing). For example, “the color saturation of images in a glossy magazine is greater than that of images viewed via a computer screen, which may result in greater attention in case of reading a print publication” (Magee, 2012, p. 87). Third, cognitive differences are about making sense of the material a (digital) magazine contains, and are closely related to the physical and perceptual differences (Dillon, 2004). For example, when the document becomes too large for the screen, other factors than image quality come into play: readers must manipulate the document and as a result they need to be
able to relate to previously displayed material that is not visible anymore (Dillon, 2004, p. 61). In such situations, memory for text and its location, schematic representation of documents and the ability to search for items gain importance (Dillon, 2004). First,
manipulating the document influences short term memory, as manipulating might place an extra burden on the limited capacity of working memory (Dillon, 2004). Second, readers have established a visual memory for the location of items within a printed text, based on a fixed relationship between an item and its position on a given page (Dillon, 2004). Memory for the location of items may however be weakened by a scrolling facility, causing visual memory for the location of items to be affected when reading a magazine on a tablet (Dillon, 2004). Third, schemas that are developed throughout the readers’ lives for an electronic information space, such as an e-edition on a tablet, are less clear-cut than for paper documents, which as a result may impede comprehension (Dillon, 2004). Lastly, electronic text enables word or term searches and other extra facility features, which particularly are an advantage for some readers, provided that they are familiar with the electronic medium (Dillon, 2004).
Besides that the physical, perceptual and cognitive differences between print and tablet may lead to differences in memory, memory differences might also depend on whether the content concerns editorial content or advertising content. More specifically, memory is expected to be greater for editorial content than for advertising content, because many individuals avoid being oversaturated by advertising messages (Speck & Elliot, 1997 in: Rumbo, 2002, 128). This is in accordance with the intentional exposure theory (Lord & Putrevu, 1993), suggesting that “readers primarily expose themselves to a magazine for editorial content rather than advertisements”, resulting in a higher memory for editorial content, as this intentional exposure is a more attentive state (Van Reijmersdal, Neijens & Smit, 2005, p. 42). In addition, whereas advertising is usually perceived as a persuasive attempt and solely in the benefit of the advertiser, editorial content is considered as objective
(Van Reijmersdal et al., 2005). This causes audiences to become more negative toward advertising than to editorial content and spend less effort processing the information in advertisements (Van Reijmersdal et al., 2005).
Overall, previous research yields that print is expected to outperform screen on memory (e.g. Sundar, Narayan, Obregon & Uppal, 1998; Tewksbury & Althaus, 2000; Althaus & Tewksbury, 2002; Jones, Pentecost & Requena, 2005; Magee, 2012). Though, one study showed that learning increases for screen through elaboration, whereas it decreases through selective scanning (Eveland & Dunwoody, 2001). Regarding type of content, earlier studies have shown that editorial content indeed receives more attention and is remembered better than advertising content (e.g. Cameron, 1994, Cameron & Curtin, 1995, Van Reijmersdal, et al., 2005; Jones et al., 2005). This suggests that “persuasive communications are processed differently than neutral consumer information, and therefore impacts attention and memory” (Jones et al., 2005, p. 629). Moreover, previous research regarding medium type and editorial and advertising content is scarce and has lead to conflicting results. One study showed that print readers remembered a higher amount of ad content in comparison to screen readers (Sundar et al.,1998), while others found no differences in recall and recognition for Web ads compared to print ads (Gallagher, Foster and Parsons, 2001a; 2001b). However, a more recent study found that although print is better for memory of both editorial and advertising content, editorial nor advertising content was better remembered in case of print or screen (Jones et al., 2005).
Hence, overall it is expected that print will outperform tablet regarding memory of editorial and advertising content, and that editorial content is better remembered than
advertising content in general. However, whether one type of content over the other is better memorized by print or tablet readers remains unclear. Therefore the following hypotheses and research question is posed:
H1. Reading a magazine in print will result in a higher memory of editorial and advertising content than reading an e-edition on a tablet.
H2. Memory is higher for editorial content than for advertising content.
RQ1. To what extent is there a difference between reading a (digital) magazine on print or on a tablet regarding memory of editorial and advertising content?
Involvement and interactivity
A very important explanation for the differences in print and e-editions on a tablet, is interactivity, as interactivity is an unique characteristic of an e-edition. Interactivity is defined as “the degree to which two or more communicating parties can act on each other, on the communication medium, and on the message and degree to which such influences are synchronized” (Liu & Shrum, 2002, p. 54). Interactivity contains three dimensions. The first dimension is active control and this dimension is “characterized by voluntary and
instrumental action that directly influences the controller’s experience” (Liu & Shrum, 2002, p. 54). This enables users to customize the information flow and voluntary acting according to their own goals and wills (Liu & Shrum, 2002). The second dimension is two-way
communication and refers to “the ability for reciprocal communication between companies and users, and users and users” (Liu & Shrum, 2002, p. 55). The third dimension is
synchronicity and this dimension refers to “the degree to which users’ input into a
communication and that the response they receive from the communication are simultaneous” (Liu & Shrum, 2002, p. 55).
Overall it is believed that interactivity is a good thing (Liu & Shrum, 2002). Though, research on the effects of interactivity shows inconsistent results (Liu & Shrum, 2009). On the one hand, research showed that interactivity has a beneficial influence on attention, memory and persuasion, due to a higher perceived efficiency, trustworthiness, quality, satisfaction, motivation and sense of fun (e.g. Liu & Shrum, 2002; Liu & Shrum, 2009; Pavlou & Steward,
2000; Rafaeli, 1988, in: Chung, 2008). On the other hand, interactivity has been shown to decrease time spent viewing ads and purchase intentions (Bezjian-Avery et al., 1998) or to have no effects on learning (Haseman et al., 2002) or attitudes (Coyle & Thorson, 2001). A possible explanation for these inconsistent results might be involvement.
Involvement is a situational factor that appears to influence the processing and effects of interactivity (Liu & Shrum, 2002; Liu & Shrum, 2009). Involvement can be defined as “the extent of cognitive elaboration that occurs in a communication process” (Liu & Shrum, 2002, p. 60). According to Elaboration Likelihood Model (Petty & Cacioppo, 1986), whether an individual is high or low involved influences whether an individual processes through the central or the peripheral route. Highly involved consumers process through the central route and for them motivation and ability to process information is sufficient. This subsequently stimulates the amount of message processing and thus causes a higher elaboration. Low involved consumers, on the other hand, process through the peripheral route and for them message-related thinking is limited. Whereas high involved individuals base their
consideration on message-relevant arguments, low involved individuals do this primarily through peripheral cues (Liu & Shrum 2009).
The level of involvement also determines how interactivity is processed (Liu & Shrum, 2002; Liu & Shrum, 2009). High involved individuals will attempt to fully utilize interactivity and are “motivated to allocate substantial cognitive resources to process a message and
attempt to evaluate relevant claims about the product advertised” (Liu & Shrum, 2009, p. 55). Moreover, as an e-edition on a tablet includes interactivity, also more cognitive involving experiences are created for this medium type (Liu & Shrum, 2002; Liu & Shrum, 2009). This can be explained through the active control and the two-way communication dimension of interactivity. First, to have active control, “one should be cognitively active and make choices, which requires closer attention and more cognitive processing” (Liu & Shrum, 2002, p. 60).
In addition, active control enables consumers to obtain information in a way that is most suitable for them and stimulates systematic elaboration on information that is relevant to their needs and processing style, which results in a higher memory (Liu & Shrum, 2002). Second, two-way communication requires more engagement than one-way communication, because the audience of one-way communication typically consist of passive message receivers, whereas two-way communication requires the audience to be active message creators (Liu & Shrum, 2002). Hence, as an e-edition on a tablet enables an interactive experience and requires a higher engagement, an e-edition on a tablet might be more suitable for high involved individuals, which eventually results in a higher memory.
On the other hand, as interactive features increase the demand on consumers’ resources (Liu & Shrum, 2009), it may also cause a cognitive burden and hinder the ease with which an individual finds information (Magee, 2012). This might be most likely for low involved individuals, as they possess a lower motivation level and ability to process the content than high involved individuals, which causes that interactivity requires too much cognitive
resources in contrast to their motivation level. This cognitive burden may result in a decrease in memory, because devoting the cognitive resources to process the interactivity leaves fewer resources available to encode the information (Magee, 2012). This also corresponds to the limited-capacity approach that sees “attention as a limited amount of mental capacity that can be allocated to only a few different tasks at a given time based on the category and priority of the tasks” (Hsieh & Chen, 2011, p. 936). Moreover, it might also be that the interactive content distracts the reader from paying attention to the core information, which is in accordance with the distraction hypothesis (Jones et al., 2005). Hence, based on these premises it can be expected that memory is negatively affected by interactivity for low involved individuals. As print does not include interactivity, it is expected that low involved individuals memorize more when reading a magazine on print.
There is also another reasoning regarding low involved individuals and the effects of interactivity. This reasoning is that low involved individuals have such a low motivation to engage in effortful processing, which causes that low involved individuals are unlikely to devote the cognitive resources necessary to engage in extensive interaction, even when interactivity is present (Liu & Shrum, 2009). Even more so, low involved consumers may not actually engage in extensive interaction, but see the mere presence of interactivity as a
positive peripheral cue, which subsequently results in more positive attitudes (Liu & Shrum, 2009). On the one hand, this might also be an explanation why print is better for low involved individuals, because low involved individuals do not devote the necessary cognitive resources that are needed for an e-edition on a tablet. However, on the other hand, it may also be argued that if low involved individuals in the first place are indeed unlikely to devote cognitive resources, then interactivity should not be a cognitive burden when reading an e-edition on a tablet, as they probably will not use and devote cognitive resources to the interactivity anyway. In that case, the benefit of one medium type over the other is eliminated for low involved individuals. Therefore, this study assumes that interactivity causes a cognitive burden for low involved individuals, causing a print magazine to lead to greater effects on memory for low involved individuals.
Hence, it is assumed that for high involved individuals reading an e-edition on a tablet increases memory for both type of contents, whereas for low involved individuals memory for both type of contents is expected to be higher when reading a magazine on print. This leads to the following hypothesis, tested for editorial and advertising content.
H3. High involved individuals show a higher memory for editorial and advertising content when reading a (digital) magazine on tablet than on print, whereas low involved individuals show a higher memory for editorial and advertising content when reading a (digital) magazine on print than on tablet.
Interactive articles and advertisements
Besides that an e-edition on a tablet is more interactive in itself, nowadays it is a common phenomenon to also add interactive features in articles and advertisements. Examples of interactive features embedded in articles and advertisements are hyperlinks and click-icons that forward readers to sites such as YouTube. As described earlier, even though it is intuitively believed that interactivity has positive effects on memory (Liu & Shrum, 2009), results of previous research are contrasting, suggesting beneficial effects of interactivity (e.g. Pavlou & Steward, 2000; Liu & Shrum, 2009), but also detrimental effects caused by
interactivity (Bezjian-Avery et al., 1998; Haseman et al., 2002; Coyle & Thorson, 2001). Hence, it remains unclear whether articles and advertisements that include interactive features are or are not better memorized by tablet readers, as compared to print readers who are only exposed to non-interactive versions of these articles and advertisements, and to what extent the previous expectations regarding high and low involvement can be applied.
Therefore, the following research question is posed:
RQ2. Do interactive features in editorial and advertising content lead to a higher memory of articles and advertisements for tablet readers and to what extent is this moderated by
involvement?
Method
As it was impossible to find a suitable magazine, that entailed a sufficient amount of interactivity and that addresses both males and females, two experiments were conducted. The procedure, design and measures were identical. However, in the second experiment, a more interactive magazine that was targeted at males was employed.
Experiment 1
Participants and Design
In total, 129 females (Mage = 21.48, SD = 3.12) participated in this study in exchange for
extra course credits or ten Euros. No participants were removed from the sample. The design was a 2 (Medium type: print versus tablet) x 2 (Involvement: high versus low) x 2
(Interactivity: absent versus present) experiment in which medium type and involvement were the manipulated factors and interactivity a within-subject factor.
Experimental stimuli
Medium type was manipulated by asking participants to read an existing magazine in print or an e-edition on tablet. The print version and the e-edition on a tablet were identical
regarding editorial content and advertising content. In contrast to the print version, the e-edition on a tablet included interactivity, existing of the interactive character of an e-e-edition in itself and interactive features, like click-icons, social media links and Internet and e-mail addresses that could be clicked on. See Appendix A for screenshots of the print magazine and the e-edition on a tablet.
Involvement was manipulated in line with the study of Liu and Shrum (2009) by telling the participants, who were assigned to the high involvement conditions, that the university was specifically chosen for the study and that their inputs were very important to the university and the publisher of the magazine (Sanoma). In addition, the publisher would use the results of the study on short notice. Participants in the low involvement condition were told that they were part of a large-scale study involving many participants, that their individual inputs would be averaged in the final analysis, and perhaps be used by the publisher (Sanoma) within a few years.
Procedure
Participants were recruited by spreading flyers, posters and online via an enrollment page for students. In these materials, the participants were told that the study was about reading magazines. The experiment itself was conducted in a library-like room in small group settings. On arriving at the location, participants were asked to sign the informed consent, by which the participants agreed to participate voluntarily in this study. Subsequently,
participants were randomly assigned to the print/low involvement, print/high involvement, tablet/low involvement or tablet/high involvement condition. Thereafter, participants were instructed to read a magazine like they would normally do for the next thirty minutes,
including the high or low involvement manipulation text. After reading a magazine/e-edition on print or tablet for thirty minutes, the participants received the questionnaire. Participants that were assigned to the print condition received a paper-pencil questionnaire, while for the participants that were assigned to the tablet condition, the questionnaire was opened digitally. The assignment of a paper-and-pencil questionnaire or a digital version of the questionnaire was to eliminate the possibility that the participant perceived a discrepancy between the medium type and the received questionnaire (Fransen, Fennis & Pruyn, 2010). It took about 20 minutes to complete the questionnaire.
Measures
Involvement. To check whether the manipulation of involvement was successful, involvement was measured by the revised Personal Involvement Inventory scale
(Zaichkowsky, 1994) and the Consumer Involvement Profile scale (Laurent & Kapferer, 1985). The ten items of the Personal Involvement Inventory scale were averaged to form an overall involvement score (α = .93, M = 4.36, SD = 0.97), as well as the fifteen items of the Consumer Involvement Profile scale (α = .81, M = 4.69, SD = 0.73).
Memory. As different measures of memory can be used to index different degrees of memory (Lang, 2000), three measures of memory were assessed: free recall, cued recall and recognition. Each measure was assessed for editorial and advertising content separately. Free recall was measured by asking the participants to write down all the articles and
advertisements they could remember seeing while reading the magazine from print or tablet. Cued recall was assessed by asking the participants whether they have seen following 15 articles (title-subtitle) or 15 advertisements (brand name), each consisting of 10 fillers and 5 true answers. Both free- and cued recall responses were coded as “1” for accurately
recalling/recognizing the article/advertisement, and “0” if the participant had not accurately recalled/recognized the article/advertisement. Both were transformed into a percentage of accurate free recall for editorial- (M = 98.65, SD = 5.83) and advertising content (M = 95.68, SD = 13.31) and cued recall for editorial- (M = 89.79, SD = 12.29) and advertising content (M = 56.95, SD = 34.59). Recognition for the editorial and advertising content was measured by asking the participants whether or not they have recognized a particular highly blurred article or advertisement as having appeared in the magazine (1 = no, 2 = yes, 3 = I am not sure). For both types of content the recognition measure was corrected, by recoding the ‘I am not sure’ responses into ‘no’ (Meditorial = 1.69, SD = 0.21; Madvertising = 1.35, SD = 0.25).
Other measures
Attitude toward the magazine in its entirety, the articles, advertisements and reading experience of the magazine was measured with a seven-point semantically differential scale composed of six bipolar adjectives (Voorveld, Neijens & Smit, 2011). The adjectives on which participants evaluated the advertisements were likable/not likable, interesting/not interesting, good/bad, appealing/not appealing, pleasant/unpleasant, positive/negative. To create a single attitude measurement for each type of attitude, the scale items were averaged
for each type of attitude measure and all measures resulted in an excellent reliability: attitude toward the magazine in its entirety (α = .95; M = 4.98, SD = 1.17), articles (α = .94; M = 5.03, SD = 1.08), advertisements (α = .91; M = 3.78, SD = 0.98), reading experience (α = .96; M = 5.06, SD = 1.08).
Use of interactivity was assessed with six items asking the participants in the tablet conditions, ‘While reading the magazine on tablet …’, for example, ‘I clicked on an advertisement’, ‘I clicked on the click symbol’ (1 = never, 2 = rarely, 3 = sometimes, 4 = often). A PCA, with rotation varimax, resulted in one unreliable component and another component that indicated an acceptable reliability (Component 1: EV = 2.46; R² = 0.41; α = .34; Component 2: EV = 1.17; R² = .19; α = .63; M = 1.12, SD = 0.45). Therefore, only the items of the second component were averaged. Participants were also asked whether they have used particular interactive features, such as the click-symbol, Internet and e-mail
addresses, social media links (1 = seen and used it, 2 = seen, but not used it, 3 = not seen it, 4 = I am not sure). As the scale indicated to be reliable, the items were averaged to a form a single scale of interactivity use (EV = 1.96; R² = .65; α = .71; M = 2.69, SD = 0.42). Whether the participants were aware of the function of these particular features (1 = no, 2 = yes) appeared to be unreliable (α = .46) and were therefore analyzed separately.
Familiarity/comfort with the Internet, tablet, paper and touch screens were measured with the statement ‘How experienced/comfortable are you with …’, resulting in an average score for familiarity/comfort with the internet (M = 6.08, SD = 1.0), tablets (M = 4.32, SD = 2.03), paper (M = 6.13, SD = 1.05) and touch screens (M = 5.95, SD = 1.25).
Moreover, participants were asked whether they have read the employed magazine before (1 = no, 2 = yes, 3 = I am not sure) (M = 4.22, SD = 0.41), how often they have read the employed magazine (1 = never, 7 = every edition) (M = 1.71, SD = 1.11), whether they have a subscription on the employed magazine (1 = no, 2 = yes) (M = 1.02, SD = 0.12), how much
interest they have in the topics of the employed magazine (1 = very little, 7 = very much) (M = 3.94, SD = 1.57), how important the topics are in the employed magazine (1 = unimportant, 7 = important) (M = 3.95, SD = 1.5), how much time they spend on the Internet (1 = less than 10 minutes, 7 = more than 4 hours) (M = 5.29, SD = 1.23) and on a tablet (1 = less than 10 minutes, 7 = more than 4 hours) (M = 1.57, SD = 1.15).
Experiment 2
In the second experiment, a total of 83 male participants participated (Mage = 23.93, SD =
5.85). No participants were removed from the sample. All measures were identical to the first experiment and unreliable scales were excluded from the analyses. See Table 1 in Tables for an overview of the Cronbach’s Alphas, means and standard deviations of the most important variables.
Results
Experiment 1
The participants that have read an e-edition on a tablet did not differ with respect to their attitudes, interest, familiarity/comfort with internet/tablet, medium preference, age and use of electronic devices from the participants that have read the magazine on print. However, frequency of reading the employed magazine (Flair) was significantly correlated with recall of advertising (r = -.287, p < 0.001) and differed significantly between the two groups(t(127) = 1.66, p < .01). Therefore, this variable was taken into account as a covariate in the analyses.
Manipulation check
To examine the effectiveness of the involvement manipulation, a multivariate analysis of variance was conducted with the Personal Involvement Inventory scale (Zaichkowsky, 1994) and the Consumer Involvement Profile scale (Laurent & Kapferer, 1985) as the dependent
variables and the involvement manipulation as the independent variable. Results revealed no significant main effects of the involvement manipulation on the Personal Involvement
Inventory scale, (F(1, 123) = 0.05, ns) and the Consumer Involvement Profile scale (F(1, 123) = 0.21, ns). Hence, it can be concluded that the manipulation did not succeed. Therefore, a median split was used to divide the Consumer Involvement Profile Scale variable into a low and high group of involvement (Mdn = 4.8).
The influence of medium type on memory of editorial and advertising content
A multivariate analysis of covariance was conducted to test the hypotheses, with medium type (print vs. tablet) and the involvement (high vs. low) as the independent variables and free recall of the editorial content, free recall of advertising content and the two recognition
measurements for editorial and advertising content as the dependent variables. Regarding editorial and advertising content, Hypothesis 1 described a main effect of medium type on memory, in which memory of editorial and advertising content is higher when reading a magazine on print than when reading an e-edition on a tablet. Hypothesis 2 expected a higher memory for editorial content than for advertising content and RQ1 questions whether one medium type over the other results in a higher memory regarding a particular content type. For the editorial content, the analysis yielded a significant main effect of medium type on free recall for the editorial content (F (1, 108) = 4.85, p < 0.05). Participants who have read an e-edition on a tablet recalled less editorial content (M = 97.9, SD = 0.65) in comparison to the participants that have read the magazine on print (M = 99.98, SD = 0.63). No main effects of medium type on cued recall (F (1, 108) = 0.419, ns) and recognition (F (1, 108) = 1.96, ns) were found. Regarding advertising content, the analysis yielded a non-significant main effect of medium type on free recall (F (1, 108) = 0.091, ns) and cued recall (F (1, 108) = 0.177, ns). However, regarding recognition a significant main effect of medium type was found,
indicating a benefit of tablet in comparison to print (F (1, 108) = 4.76, p < 0.05; Mtablet = 1.38,
SD = 0.03; Mprint = 1.28, SD = 0.03). Therefore, H1 is partially supported. Moreover, print
appears to be more suitable than tablet for readers regarding memorizing editorial content, whereas tablet appears to be more beneficial than print for readers when memorizing advertising content.
To examine H2, a GLM repeated measures analysis was conducted, with content type (editorial and advertising) as within-subjects variable, medium type as the independent variable and free recall, cued recall and recognition of editorial and advertising as the dependent variables. The analysis yielded a significant difference between editorial and advertising content (F(1.55, 176.37) = 955.42, p < 0.001). Moreover, contrasts show significant differences for free recall (F(1, 114) = 7.34, p < 0.01; Mfreerecall-ed = 99.9, SD =
0.50; Mfreerecall-ad = 95.65), cued recall (F(1, 114) = 118.9, p < 0.001; Mcuedrecall-ed = 89.46, SD =
1.15; Mcuedrecall-ad = 59.01, SD = 3.15) and recognition (F(1, 114) = 9108.7, p < 0.001;
Mrecognition-ed = 1.7, SD = 0.02; Mrecognition-ad = 1.34, SD = 0.02), indicating that editorial content
is better remembered than advertising content. Therefore, H2 is accepted.
Involvement and interactivity
Hypothesis 3 described a two-way interaction among medium type and involvement regarding editorial and advertising content. Under high involvement conditions, tablet readers were expected to have a higher memory, while under low involved conditions print readers were expected to have a higher memory. The hypothesis was tested for editorial and
advertising content. Regarding editorial content, the analysis yielded a non-significant main effect of involvement on free recall (F (1, 108) = 0.014, ns), cued recall (F (1, 108) = 0.203, ns) and recognition (F (1, 108) = 1.70, ns). Moreover, no interaction effects between medium type and involvement were found for free recall (F (1, 108) = 0.005, ns), cued recall (F (1,
108) = 0.311, ns) and recognition (F (1, 108) = 0.141, ns). Therefore, regarding editorial content, H3 is rejected.
Regarding advertising content, no significant main effect of involvement was found on free recall (F (1, 108) = 0.664, ns) and recognition (F (1, 108) = 1.29, ns). However, higher involvement did result in higher cued recall of advertising content than low involvement (F (1, 108) = 4.17, p < 0.05, Mlow = 53.00, SD = 4.40; Mhigh = 65.87, SD = 4.50). Moreover, no
significant interaction was present between medium type and involvement on free recall (F (1, 108) = 0.093, ns) and cued recall (F (1, 108) = 0.244, ns). Medium type and involvement did interact regarding recognition (F (1, 108) = 7.25, p < 0.01). High involved tablet readers recognized less advertising content (M = 1.29, SD = 0.05) than high involved print readers did (M = 1.32, SD = 0.05), while low involved print readers recognized much more advertising content (M = 1.47, SD = 0.04) than low involved tablet readers did (M = 1.25, SD = 0.05). Therefore, regarding advertising content H3 is partially accepted.
Interactive articles and advertisements
Before testing RQ2 regarding interactive articles and advertisements, analyses regarding the actual use and familiarity with the function of the interactive features showed that interactive features are not seen and used that often. Specifically, 69% of the tablet readers enlarged the texts, 44% reported to have read the text version of an article, 9% reported to scroll to the desired section, 8% clicked on an advertisement and 6% clicked on the click-symbol, whereas only 5% clicked on links. The participants were also asked whether they have seen and used particular interactive features and whether they knew the function of these features. Regarding the ‘click-symbol’, 6% reported to have seen and used it, while 32% only saw the feature. 34.4% was familiar with the function of this feature. No one saw and used the Internet and e-mail address feature and 27% reported to have only seen the feature. 56%
reported to be familiar with the function of the feature. Lastly, 1.6% had seen and used the social media links feature, while 23% only saw it. 69% of the participants reported to be familiar with the function of this feature.
To examine whether interactive features in editorial and advertising content lead to a higher memory for tablet readers or not and to what extent this was moderated by
involvement (RQ2), a GLM repeated measures analyses was conducted. Interactivity served as the within-subjects factor, representing the five articles, meaning that the within-subjects factor consisted of five levels, each level representing a different (non/less or interactive) article. Medium type and involvement served as the independent variables and recognition of the articles as the dependent variable(s). Regarding editorial content, the analysis yielded a significant main effect of the within-subjects factor on the recognition of the articles (F (3.6, 431) = 5.46, p < 0.001). Results showed that the most interactive article scored highest on recognition (M = 1.94, SD = 0.02) and differed significantly from the less interactive article (F(1, 119) = 27.5, p < 0.001; M = 1.6, SD = 0.04), the least interactive article (F(1, 119) = 10.85, p ≤ 0.001; M = 1.55, SD = 0.05) and the two non-interactive articles (F(1, 119) = 10.06, p < 0.01; Marticle4-noninteractive = 1.72, SD = 0.04; (F(1, 119) = 9.72, p < 0.01; M article5-noninteractive = 1.64, SD = 0.04).
No significant interaction effect between medium type and interactivity was found (F(3.6, 431.01) = 1.34, ns). The analysis did yield a significant interaction effect between
interactivity and involvement (F(3.6, 431) = 2.8, p < 0.05), with contrast tests revealing a significant difference between the most interactive article and the fourth non-interactive article (F(1, 119) = 4.04, p ≤ 0.05). A highly interactive article is better remembered by high involved individuals (Mhigh = 1.97, SD = 0.03 vs. Mlow = 1.91, SD = 0.03), whereas a
non-interactive article is better remembered by low involved individuals (Mhigh = 1.58, SD = 0.06;
within-subjects factor, involvement and medium type (F(3.6, 431) = 0.4, ns), indicating that tablet readers do not recognize interactive articles more often and that it does not depend on involvement.
Regarding advertising content, a similar repeated measures analysis was conducted. Only now the within-subjects factor represented five advertisements, instead of articles. The analysis indicated a significant main effect of the within-subjects factor on the recognition of advertisements (F (3.75, 453.6) = 7.2, p < 0.001). Contrasts tests showed that the interactive advertisement (M = 1.48, SD = 0.04) was significantly better recognized than the third non-interactive advertisement (F(1, 121) = 9.91, p < 0.01; Mnoninteractive-ad3 = 1.30, SD = 0.4), the
fourth non-interactive advertisement (F(1,121) = 15.64, p < 0.001; Mnoninteractive-ad4 = 1.25, SD
= 0.04), and the fifth non-interactive advertisement (F(1, 121) = 15.41, p < 0.001; M -noninteractive-ad5 = 1.26, SD = 0.04). No significant difference was found regarding the second
non-interactive advertisement (F(1, 121) = 0.77, ns). The analysis yielded non-significant interaction effects between medium type and interactivity (F(3.7, 453.6) = 1.94, ns),
interactivity and involvement (F(3.75, 453.6) = 1.23, ns) and interactivity, medium type and involvement (F(3.75, 453.6) = 0.15, ns), meaning that interactive advertisements are not more often recognized by tablet readers and that involvement does not influence this process.
Experiment 2
To examine whether results are stronger when employing a magazine that entails more interactivity, a second experiment was conducted and the same analyses were repeated. Again, tablet and print readers did not differ with respect to their attitudes, interest,
familiarity/comfort with internet/tablet, medium preference, age and use of electronic devices. However, frequency of reading the employed magazine (Autoweek) was significantly
-0.278, p < 0.05) and differed significantly between the two groups (t(81) = 1.11, p ≤ .05). Therefore, this variable was taken into account as a covariate. Similar to the first experiment, the involvement manipulation did not succeed, as the main effects of the involvement
manipulation on the Personal Involvement Inventory scale (F(1, 79) = 0.277, ns) and the Consumer Involvement Profile scale (F(1, 79) = 0.524, ns) were non-significant. Therefore, a median split was used to divide the Consumer Involvement Profile Scale variable into a low and high group of involvement (Mdn = 4.43).
The influence of medium type on memory of editorial and advertising content
To test H1-H3 and answer RQ1, a two-factor (medium type and involvement) MANOVA on free recall, cued recall and recognition as dependent variables was conducted. Regarding editorial content, the analysis yielded a non significant main effect of medium type on free recall (F (1, 60) = 0.148, ns) and recognition (F (1, 61 = 1.34, ns). Regarding cued recall of editorial content, a marginally significant main effect of medium type was found (F (1, 60) = 3.60, p = 0.06). Tablet readers recognized more editorial content (M = 72.31, SD = 6.01) than print readers (M = 56.96, SD = 5.33). Regarding advertising content, the analysis yielded a non significant main effect of medium type on free recall (F (1, 61) = 2.03, ns) and cued recall (F (1, 61) = 0.06, ns). However, a marginal significant main effect of medium type on
recognition of advertising content was found (F (1, 60) = 3.41, p = 0.07). Tablet readers recognized more advertising content (M = 1.33, SD = 0.03) than print readers (M = 1.24, SD = 0.30). Therefore, for the highly interactive magazine, H1 is rejected. Moreover, tablet is more suitable than print for memorizing both types of content.
Again, similarly to the first experiment, a GLM repeated measure analysis was conducted to test whether editorial content is better remembered than advertising content. A significant difference between editorial and advertising content was found (F(2.7, 173.1) = 99.57, p <
0.001). However, contrast tests yield no significant differences (p > 0.21), except regarding recognition (F(1, 63) = 14.63, p < 0.001), resulting in a higher recognition of editorial content than of advertising content (Mrecognition-ed = 1.55, SD = 0.03; Mrecognition-ad = 1.28, SD = 0.02).
Therefore, H2 is partially supported.
Involvement & interactivity
For editorial content, no main effect was found of involvement on free recall (F (1, 60) = 0.541, ns), cued recall (F (1, 60) = 0.805, ns), and recognition (F(1, 60) = 0.02, ns). Moreover, no significant interaction effect was present for medium type and involvement on free recall (F (1, 60) = 0.004, ns) and on recognition (F (1, 60) = 2.72, ns). However, regarding cued recall of editorial content, a significant interaction effect of medium type and involvement was found (F (1, 60) = 3.9, p ≤ .05; MTablet-high = 67.94 , SD = 6.34; MPrint-high = 68.54, SD =
7.24; MTablet-Low = 76.68, SD = 10.18; MPrint-low = 45.38, SD = 7.83). Surprisingly, high
involved individuals that read an e-edition on a tablet recognized somewhat less editorial content than high involved print readers do, whereas low involved tablet readers recognize much more editorial content than print readers do. Hence, as the results point in the opposite direction, H3 is rejected for editorial content.
Regarding advertising content, the analysis yielded a significant main effect of involvement on free recall (F (1, 60) = 9.59, p < 0.01, MLow = 69.83, SD = 6.12; MHigh =
93.51, SD = 4.55) and recognition (F (1, 60) = 5.92, p < 0.05; MLow = 1.34, SD = 0.04; MHigh =
1.23, SD = 0.03) and a marginally significant effect on cued recall (F(1, 60) = 3.46, p = 0.068, MLow = 61.03, SD = 6.42; MHigh = 68.24, SD = 4.79). The analysis yielded a non-significant
interaction effect of medium type and involvement on free recall (F(1, 60) = 2.72, ns) and cued recall (F(1, 60) = 0.015, ns). However, regarding recognition, a marginally significant interaction effect was found, showing that high involved tablet readers recognize much more
advertising content than high involved print readers, while the difference for low involved individuals is less pronounced (F (1, 60) = 2.83, p = .09, MTablet-high = 1.31, SD = 0.04; M Print-high = 1.15, SD = 0.04; MTablet-Low = 1.34, SD = 0.06; MPrint-low = 1.33, SD = 0.05). Therefore,
regarding advertising content, H3 is partially supported.
Interactive articles and advertisements
Similar to the first experiment, interactive features are not seen and used that often. However, the interactive icons in this experiment were seen and used more often.
Specifically, 42% of the tablet readers enlarged the texts, 37% have read a text version of an article, 4.7% used the navigation to scroll to the desired section, 4% clicked on an
advertisement and only 3% clicked on the click symbols, whereas 20% clicked on the links. Participants were also asked whether they have seen and used particular interactive features and whether they were familiar with the function of these features. There were two icons, one icon was a navigation icon (square with an open arrow in it), to directly go to the desired page, while the other icon was a typical ‘play’ icon, to forward the reader to videos or
additional information. Regarding the navigation icon, 22% reported to have seen and used it, whereas 71% reported to have seen it, but not used it. The majority reported to be familiar with the function of that particular feature (71%). Regarding the play icon, 34% reported to have seen and used it, while only 34% saw the icon. 49% was not familiar with the function of that particular feature. 10% of the participants saw and used the possibility to click on Internet and e-mail addresses, whereas 17% only saw the feature. 66% was familiar with the function of the feature. Lastly, only 10% saw the possibility to send a reaction. 78% reported to be not familiar with the function of that particular feature.
Similarly to the first experiment, RQ2 was assessed by conducting a GLM repeated measures analysis, with interactivity as the within-subjects factor, representing the five
articles, medium type and involvement as the independent variable and recognition of articles as the dependent variable(s). The analysis indicated that there was a significant main effect of interactivity on the recognition of articles (F(4, 300) = 3.80, p < 0.01). The most interactive article only differed significantly from the least interactive article (F(1, 75) = 4.36, p < 0.05; Mmost-interactive-5 = 1.60, SD = 0.06; Mleast-interactive-1= 1.49, SD = 0.06). Other contrasts were
non-significant (p > 0.12). Moreover, no non-significant interaction effects were present between medium type and interactivity (F(4, 300) = 1.53, ns), involvement and interactivity (F(4, 300) = 0.793, ns) and interactivity, medium type and involvement (F(4, 300) = 0.88, ns),
suggesting that regarding the highly interactive magazine, interactive articles are not more often recognized by tablet readers either, and that involvement does not influence this process. The same analysis was conducted regarding advertising content. A significant main effect of the within-subjects factor on recognition of advertisements was found (F(4, 304) = 2.97, p < 0.05). Contrasts reveal that the most interactive advertisement (M = 1.16, SD = 0.04) is significantly recognized less than the first non-interactive advertisement (F(1, 76) = 4.85, p < 0.05; Mnon-interactive-1 = 1.32, SD =0.05), the third least interactive advertisement (F(1, 76) =
8.88, p < 0.01; Mleast-interactive-3 = 1.38, SD = 0.06) and marginally significant difference with
the fourth non-interactive advertisement (F(1, 76) = 2.94, p = 0.09; Mnon-interactive-4 = 1.28, SD
= 0.05). No significant difference was present regarding the less interactive advertisement (F(1, 76) = 1.07, ns). Moreover, non significant interaction effects were found for medium type and interactivity (F(4, 304) = 1.54, ns), involvement and interactivity (F(4, 304) = 1.63, ns) and interactivity, medium type and involvement (F(4, 304) = 0.892, ns). Hence, similar to the previous results, interactive advertisements are not more often recognized by tablet readers and it does not depend on involvement.
Conclusion and Discussion
The aim of this study was to extend our knowledge regarding the effectiveness of print and digital magazines on a tablet, by investigating it’s influence on memory for editorial and advertising content. In addition, it studied to what extent involvement and interactivity influenced this process. The study shows that, overall, reading a digital magazine on a tablet leads to greater cognitive reactions than when reading a magazine on print. When looking at the differences between types of content, it can be concluded that print is more suitable for memorizing editorial content, while tablet is more beneficial regarding advertising content. Furthermore, it can also be concluded that individual’s involvement influences whether print or tablet is more effective in case of remembering advertising content, showing that print is particularly effective for low involved individuals, while tablet is especially effective for high involved individuals. Lastly, whereas reading an e-edition on a tablet, being more interactive in itself, leads to greater cognitive reactions, particular interactive features do not increase memory for particular articles or advertisements when reading an e-edition on a tablet. It therefore seems that, even though an e-edition on a tablet is especially more beneficial in case of a highly interactive magazine, other differences between (interactive-) articles,
advertisements and the magazine itself should also be taken into account.
Discussion of findings regarding the influence of medium type
Based on the two experiments, it can be concluded that tablet is superior in evoking cognitive responses. Though the first experiment showed that reading a magazine on print resulted in a higher recall of editorial content than when reading an e-edition on a tablet, both experiments showed that reading an e-edition on a tablet resulted in higher recognition of advertising content than when reading a magazine on print. In the second experiment, however, tablet was also more suitable when recognizing editorial content, suggesting that
tablet is especially better in case of a highly interactive magazine. Both studies confirmed that memory for editorial content is higher than that for advertising content.
The findings that a tablet is superior are in contrast with previous studies showing that print outperforms screen/online versions (Sundar et al., 1998; Tewksbury & Althaus, 2000; Althaus & Tewksbury, 2002; Jones et al., 2005; Magee, 2012). An explanation for these contrasting results between the literature and the findings of the current study, may be that learning differences are diminishing due to improved computer displays and because an interactive screen, like a tablet, offers a more natural comparison than a typical screen display (Jones et al., 2005). The diminishing advantage of print can also be due to a higher familiarity with tablets nowadays. Moreover, it can also be explained by structural isomorphism, which entails that navigation through Web sites, in this case an e-edition, ‘mimics the associative nature of human memory and information processing, suggesting Web superiority’ (Eveland & Dunwoody, 2001, p. 48).
Furthermore, reading an e-edition on a tablet may be perceived as more exiting and therefore stimulating participants to encode more information. Previous research has showed that consumers pay more attention to things they like and this higher appreciation
subsequently might lead to higher cognitive reactions (Van Reijmersdal et al., 2005). Hence, it can be perceived as a kind of “spill over” effect. An explanation for the superiority of print regarding editorial content and the superiority of tablet regarding advertising content, may be because articles encompass more text as compared to advertisements, and perhaps consumers rather read longer texts from print.
Moreover, in both studies, editorial content was better remembered than advertising content, which is in line with previous research, suggesting that editorial content indeed receives more attention than commercial messages and that it is therefore remembered better than advertisements (Jones et al., 2005; Cameron, 1994; Cameron & Curtin, 1995; Van
Reijmersdal et al., 2005). In accordance with the intentional exposure theory (Lord & Purevu, 1993), “readers pay more attention to formats that are perceived as more editorial, because readers intentionally expose themselves to editorial content of a magazine as opposed to advertising, which results in an attention profit for editorial content” (Van Reijmersdal et al., 2005, p. 50). In addition, a higher memory for editorial content may be because a magazine in general contains more editorial content than advertising content.
Furthermore, differences within the experiments regarding free recall, cued recall and recognition suggest that depending on medium type, editorial and advertising content are stored differently in memory. Recognition is a test of whether or not traces exist in memory, whereas recall may be affected by decay, information replacement, and cues present at
retrieval (Tulving & Pearlstone, 1966, in: Jones et al., 2005). Moreover, different measures of memory can be used to index different degrees of memory (Lang, 2000). While recognition is a measure of encoding of information, cued recall is a measure of thoroughness of storage, free recall, on the other hand, is a measure of retrieval processes (Lang, 2000). As most differences between the medium types were found for the recognition measures, it may be concluded that reading a magazine on print or an e-edition on a tablet mostly differ regarding the encoding of information and the thoroughness of storage. In addition, the lack of
difference regarding recall might be considered as a ceiling effect.
Discussion of findings involvement & interactivity
Based on the two experiments, it can be concluded that involvement influences whether print or tablet is more effective in case of memorizing advertising content. In case of the less interactive magazine, print is particularly more effective than tablet for low involved
individuals, whereas in case of the highly interactive magazine, tablet is particularly more effective than print for high involved individuals. On the other hand, regarding the highly
interactive magazine and editorial content, tablet appeared to be especially more effective than print for low involved individuals.
The findings regarding advertising content of these two experiments provide partial support for the expectations that high involved individuals perform better on a tablet and that low involved individuals perform better on print (e.g. Liu & Shrum, 2002; Ariely, 2000; Sicilia et al., 2005; Liu & Shrum, 2009). According to the literature, high involved individuals possess the ability and capacity to process interactivity, which may explain the effectiveness of a highly interactive magazine for high involved individuals, whereas interactivity can cause a cognitive burden for the low involved individuals, which may explain the effectiveness of print for low involved. However, as reading an e-edition on a tablet was especially more effective in the second experiment, it may also be related to the different types of magazines employed in each experiment. Therefore, more research into a possible explanation is required.
An explanation why reading an e-edition on a tablet appears to be more beneficial than print for low involved individuals regarding editorial content, may be that interactivity has increased elaboration, in such a way that higher interactivity levels may have shifted the focus of processing toward a more central route for some consumers (Sundar, Kalyanaraman & Brown, 2003, in: Liu & Shrum, 2009). Mean involvement levels show that participants in the low involvement conditions were still quite involved, causing a more mixed processing, instead of central or peripheral (Liu & Shrum, 2009). Moreover, reading an e-edition on a tablet could have served, besides a more exiting reading experience, as a peripheral cue, as “peripheral processing may on the one hand involve attention mainly to peripheral cues (and not content), or may, on the other hand, involve attention to content but simply less of it” (Liu & Shrum, 2009, p. 66). Aside from differences due to type of content, the different direction regarding editorial content as compared to advertising content in the second experiment,
might be also explained by the different degrees of memory on which the two recognition measures tap (Lang, 2000). As regarding editorial content, results were found for cued recall of editorial content, whereas the other results were found for recognition.
Both experiments showed that memory for the most interactive article was higher than for the less or non-interactive articles. Regarding advertising content, on the other hand, the two experiments point in different directions. In the first experiment, the interactive advertisement was recognized the most, while in the second experiment the most interactive advertisement was recognized the least. However, these differences between the interactive and less or non-interactive articles did not depend on whether the magazine was read from tablet or print and whether an individual was high or low involved, suggesting that other differences between the articles/advertisements may have came into play.
It might be that it was not interactivity that caused a difference in the recognition of editorial and advertising content. As the most interactive article in the first experiment was the first large article and the interactive advertisement was the first advertisement in the magazine, it might be that a primacy effect could have occurred, meaning that things or events that have been first in a series, make the most impression and are therefore better remembered (Li, 2010). Moreover, the most interactive article was also visually more outstanding from the previous smaller articles. In the second experiment, it might be that the most interactive article was more recognized due to other differences between articles. The most interactive advertisement in the second experiment, was not the first nor last advertisement. It was an outstanding advertisement with moving images and several clickable areas. Perhaps the moving images were perceived as annoying and therefore participants automatically avoid paying attention to the advertisement.
Future research and limitations
As the current study was the first to investigate differences between print and digital magazines on a tablet, this study has lead to several suggestions for future research. First, future research should focus on the specific reading experience of a consumer when reading an e-edition on a tablet, to get more insights into the reasons why and how consumers remember particular content. For example, conducting an eye-tracking study would give a more precise insight into attention, and subsequently the memory differences between print and tablet.
Second, it remains unclear whether interactivity has shifted the focus of processing to a more central route for some consumers. The same applies for whether peripheral processing has involved attention to peripheral cues only or to the content, though less than when
centrally processing (Liu & Shrum, 2009). Therefore, future research is needed to reveal these underlying processes. Third, as the current study employed real life magazines, content, including particular interactive features, could not be manipulated. Therefore, future research could manipulate these interactive features and empirically test its effects in the context of reading a print magazine and an e-edition on a tablet.
Fourth, as it remains unclear to what extent the interactive features have influenced the results, more knowledge is needed regarding the function of interactivity in e-editions on tablets and the possible (dis)advantages this might have. Future research should therefore focus on why particular (interactive) articles and advertisements are or are not better
remembered for print or tablet and whether consumers are actually aware of the presence of interactivity in the articles or advertisements.
Lastly, a main limitation of the current study is that the manipulation of involvement did not succeed, which may be because the manipulation was not strong enough. Although this
manipulation was validated in the study of Liu and Shrum, (2009), future research may pre-test this manipulation in the context of their own studies.
Practical implications
Even though this study was not in line with all expectations, this study points out that publishing an e-edition on a tablet is a possible solution for the challenges the media branch is currently facing. It shows that the longstanding superiority of print is not so clear cut
anymore. Though print remains to have its benefits, reading an e-edition on a tablet not necessarily means that consumers have less attention regarding editorial and advertising content. Moreover, when advertisers want the readers to memorize the advertising content in a more interactive magazine, they should take into account the target’s group level of
involvement to achieve attention. In addition, while interactive content was not better memorized by tablet readers, that does not necessarily mean that magazines should not incorporate interactivity, as an e-edition on a tablet being more interactive in itself has lead to positive results. Besides the differences in interactivity, differences between specific
magazines should also be taken into account, as differences may not exclusively depend on whether it is an e-edition, but on the magazine itself. However, more studies are needed to give answers and until then e-editions seem a good attention trigger.
References
Althaus, S.L., & Tewksbury, D. (2002). Agenda setting and the ‘new’ news: patterns of issue importance among readers of the paper and online versions of the New York Times. Communication Research, 29, 180-207.
Bezjian-Avery, A., Calder, B., & Iacobucci, D. (1998). New media interactive advertising vs. traditional advertising. Journal of Advertising Research, 38, 23–32.
Cameron, G.T. (1994). Does publicity outperform advertising? An experimental test of the third-party endorsement. Journal of Public Relations Research, 6(3), 185-207.
Cameron, G.T., & Curtin, P.A. (1995). Tracing sources of information pollution: A survey and experimental test of print media’s labeling policy of feature advertising. Journalism and Mass Communication Quarterly, 72, 178-189.
Chung, D. (2008). Interactive features of online newspapers: Identifying patterns and predicting use of engaged readers. Journal of Computer-Mediated Communication, 13, 658-679.
Coyle, J.R., Thorson, E. (2001). The effects of progressive levels of interactivity and vividness in Web marketing sites. Journal of Advertising, 30, 65–77.
Dijkstra, M., Buijtels, H. E., & van Raaij, W. F. (2005). Separate and joint effects of medium type on consumer responses: a comparison of television, print, and the Internet. Journal of
Business Research, 58(3), 377-386.
Dillon, A. (2004). Designing Usable Electronic Text (2nd ed.). London: CRC Press. Eveland, W., & Dunwoody, S. (2000). User control and structural isomorphism or
disorientation and cognitive load? Learning from the web versus print. Communication Research, 28(1), 48-78.
Eveland, W., & Dunwoody, S. (2002). An investigation of elaboration and selective scanning as mediators of learning from web versus print. Journal of Broadcasting & Electronic
Media, 46(1), 34-53.
Fransen, M.L., Fennis, B., & Pruyn, A.T.H. (2010). Matching communication modalities: The effects of modality congruence and processing style on brand evaluation and brand choice. Communication Research, 37(4), 576-598.
Gallagher, K., Foster, K. D., & Parsons, J. (2001a). The medium is not the message: Advertising effectiveness and content evaluation in print and on the web. Journal of Advertising Research, 41(4), 57-70.
Gallagher, K., Parsons, J., & Foster, K.D. (2001b). A tale of two studies: Replicating advertising effectiveness and content evaluation on the Web. Journal of Advertising Research, 41, 71–84.
Haseman, W.D., Nuipolatoglu, V., & Ramamurthy, K. (2002). An empirical investigation of the influences of the degree of interactivity on user-outcomes in a multimedia
environment. Information Resources Management Journal, 15 (2), 31–48.
Hsieh, Y. C., & Chen, K. H. (2011). How different information types affect viewer’s attention on internet advertising. Computers in Human Behavior, 27(2), 935-945.
Intomart GfK (2013). Trends in Digitale Media. Hilversum.
Jones, M. Y., Pentecost, R., & Requena, G. (2005). Memory for advertising and information content: comparing the printed page to the computer screen. Psychology and Marketing, 22, 623–648.
Lang, A. (2000). The limited capacity model of mediated message processing. Journal of Communication, 50(1), 46-70.
Laurent, G., & Kapferer, J. (1985). Measuring Consumer Involvement Profiles. Journal of Advertising Research, 43, 207-216.
Li, C. (2010). Primacy effect or recency effect? A long-term memory test of Super Bowl commercials. Journal of Consumer Behavior, 9, 32-44.
