Abstract
The e-book has gained a large number of readers in many countries. However, the growth rate, as well as the market penetration of e-books, varies considerably between countries and language areas. In the United States and other English-speaking countries, the use of e-books is more widespread than in many European countries, such as Sweden. In both the United States and Swedish markets, the growth rate has slowed down during the latest 2 years, and the majority of readers still seem to prefer printed books. This article aims to reveal the relationship between attitudes and use of e-books and to contextualize this in relation to socio-demographic background, library use and reading habits. The results, based on national, representative surveys in Sweden during 2013 and 2014, show that people’s attitudes are a significant and important factor for the choice of e-books instead of print, but this is probably a two-way relationship. It is also evident that e-books serve a complementary rather than displacement function in relation to print.
Keywords
To understand the development of the e-book market, one needs to study the actors in the field, how they consider the e-book and how these actors relate to each other, as well a number of aspects related to supply, pricing, technology access and reading preferences. Both publishers and authors point to the audience as a key factor for previous and further development (Nilsson et al., 2015). The Swedish audience, however, has thus far been somewhat reluctant to adopt e-books (Bergström et al., 2017; Wischenbart, 2014).
The uptake of new technology for media use, such as reading e-books, relies on access to both e-book titles and reading devices on which to access them (Jung et al., 2012; Westin, 2013). E-book access also is related to accountable content, attractive e-books and pricing for content and reading devices (Jung et al., 2012). Access is not, however, the only parameter in the process of diffusion of new media technology and its applications; how it is perceived among potential users also matters. If the strengths exceed the weaknesses, people might consider using the new distribution platforms and the media get domesticated (Haddon, 2006). Actions and attitudes are strongly intertwined in diffusion processes such as the transformation from printed books to e-books (Ajzen, 2005; Solomon et al., 2010).
This study is part of a large research programme in which these factors and actors are captured in relation to e-book production, distribution and reading. The study aims to reveal the relationship between readers’ attitudes and use of e-books and to contextualize this with regard to socioeconomic background and experience of book reading and library use.
The e-book reading context
In many countries, e-books already make up a significant part of the total book market. The use of e-books is widespread in English-speaking countries, such as the United States and the United Kingdom, but in Europe and in small-language countries like Sweden, e-books have reached only a small part of the potential market (Steiner, 2015; Wischenbart, 2014). By the time the surveys for this study were conducted, e-books in Sweden represented only about 25% of the total number of titles published, most of them fiction. In terms of sales, income from e-books was less than 2% of the total revenue (The Swedish Publishers Association, 2017).
Actors like Amazon do not operate directly in the Swedish market (although Amazon plans to do so from 2018 to 2019 (Kozlowski, 2018)); instead, other online booksellers have emerged, with the aggregator Elib being the largest. The pricing of e-books generally is the same as for printed books and with high pricing comes lower sales (Wikberg, 2015). There are a few streaming services on the market, but they have been only moderately successful (Bergström et al., 2017; Steiner, 2015). There seems to be an imbalance between production and demand (Borg, 2014).
One plausible explanation for the Swedish development is the fact that e-books in Sweden reach their readers through public libraries rather than through sales; in 2011, public libraries accounted for as much as 90% of the Swedish e-book market (Steiner, 2015). Public libraries in Sweden have been victims of a monopoly-like situation in the provision of e-books. Having been referred to Elib and to a financial model charging per loan, e-books have been a rather costly business for libraries, thus constraining the supply of titles to library users.
For the individual, access to reading devices also affects use, as do personal reading habits. In Sweden, most readers access e-books via tablet or laptop computers (Findahl, 2014), while, for instance, in the United States, specialized tablets like the Kindle are most common (Zickuhr and Rainie, 2014). Access to reading devices is fairly good in Sweden. In 2014, 50% of households had access to tablets, 90% to a computer and 72% to a smartphone (Bergström et al., 2015).
On a general level, the prerequisites for e-book reading are in place. There are, however, well-known differences when comparing different groups of the population. Access to reading devices is higher among people aged 16–29, while, among those aged 65–85, fewer have access to a smartphone or a tablet (smartphone 95% and 47%, respectively; tablet 58% and 34%, respectively, in 2014). Access also is higher among those with higher formal education and family income above the mean (Bergström and Höglund, 2014). In other words, the differences in access seem to correspond with what is known about the diffusion of new media innovations (Rogers, 2003).
Theory
E-book reading and attitudes towards e-books can be understood from several perspectives. One, already mentioned, is the perspective of technology diffusion and acceptance. Another, which will shed light on the adoption of e-books, is the substitution or displacement theory. Psychological research on attitudes and behaviour also can contribute to the understanding of e-book development among readers.
Diffusion and technology acceptance
Innovations of most kinds develop and diffuse in complex social contexts affected by social factors or supervening social necessities (Winston, 1998). In his theory, Winston highlights a concentration of social forces working directly on the processes of innovation. There is no limitation of the forces that can act as supervening social necessities. They can range from the objective requirements of changed social circumstances to the subjective whims of perceived needs. The action of supervening necessity does not account for the entire development and reception of a technology. It is reasonable to assume that there is a variety of supervening necessities, ranging from economic to cultural, among potential readers of e-books.
However, just as one can reveal supervening social necessities, there also are, according to Winston (1998), social constraints limiting the potential of the innovation. The diffusion of the invention depends upon the law of the suppression of radical potential, the source of which may vary according to the nature of the invention, but the result of which is a failure to fully develop the possibilities of the invention. For example, restrictions on access to various book titles and publishers’ reluctance to invest in a new format can work as restraining factors.
On an individual level, there are several theories for understanding why people adopt or reject an innovation such as the e-book. Two of the most common are the diffusion of innovations theory (Rogers, 2003) and the technology acceptance model (TAM; Davis, 1989). The TAM emphasizes an item’s perceived usefulness and ease of use, institutional norms and organizational characteristics as powerful determinants of the use of information systems at work. Perceived usefulness and perceived ease of use are closely related to Rogers’ (2003) concepts of relative advantage and complexity. The unified theory of acceptance and use of technology is a comprehensive synthesis of prior technology acceptance research (Venkatesh et al., 2012).
Some groups in a society have proved to be more eager than others to adopt and learn new techniques. In Rogers’ terms, these are called innovators and are characterized as trendsetters who actively seek the new. They are mostly younger than the average age. The innovators are, according to Rogers, followed by the early adopters, who tend to be well-educated and well-off as a consequence of new technologies and their applications, often being relatively expensive and demanding a certain amount of knowledge. Social outcomes, such as higher status in the community or being unique in the group, may be important additional drivers of use. Further, patterns of diffusion are well-known in the context of the development of digital media (Venkatesh et al., 2012). There are several reasons to expect the groups just mentioned to be more eager than others to use e-books.
Substitution and displacement
The issue of how upcoming media interact with existing ones has been studied for a long time. Scholars have discussed the substitution or displacement of old media since at least the mid-20th century (Lasswell, 1948). Basically, there are two approaches: replacement and complementary. The former takes a point of departure in different media having different functions and properties, which, in turn, affect how they might compete for people’s time and resources. One can expect people to decrease their use of, or even stop using, an old medium, when a new one is made available (Dimmick, 2003). The replacement theory suggests that existing habits are transferred to new platforms (Chan and Leung, 2005).
The second approach emphasizes that people incorporate new media in a complementary way. The needs or motives for adopting new media are there, but the forms for satisfying them might differ (Nguyen and Western, 2006). To satisfy your need for reading, for instance, you might use several platforms to be able to read in different situations. One might complement printed books with e-books to get more access to content of interest. The complementary theory argues that use of new media is added to use of the old (Chan and Leung, 2005).
The displacement or complementary relations are not mutually exclusive but strongly connected to functions and needs (Dutta-Bergman, 2004; Flavián and Gurrea, 2009). When a new medium is regarded as more functionally desirable than an old medium, the audience may abandon the old and replace it with the new (Carey and Elton, 2010; De Waal and Schoenbach, 2010; Lin, 2001). The competition is about finding a niche in a certain context, meaning that the new medium might be suitable in one situation and the old in another (Dimmick, 2003). It has been found, for instance, that printed newspapers are read at home in the morning, online news from the desktop computer at work and mobile news when on the go (Dimmick et al., 2011).
To study the uptake of e-books from a substitution or displacement perspective, one needs to know something about the habits of using the old medium – in this case, the printed book. Sex is an important variable in predicting both the amount and type of reading material an individual pursues. Socialization, reading proficiency and work status account for some of the variation in fiction reading between men and women (Tepper, 2000). Other factors that explain book reading behaviour include level of education, library use, household income and age (Clark, 2013; Höglund, 2012; Verboord and van Rees, 2003).
Attitudes and behaviour
A deeper understanding of the development of e-book reading can be ascertained by studying attitudes concerning the digital book format. Attitudes and behaviour are related at a general level (Solomon et al., 2010) and in more specific situations (Ajzen, 2005; Miesen, 2003). An attitude is, according to Ajzen (2005: 3), ‘a disposition to respond favourably or unfavourably to an object, person, institution or event’. Attitudes are about differing preferences (Johnson and Boynton, 2010). When talking about how attitudes are related to behaviour, the research field is somewhat inconsistent. The interrelation between the two sometimes is found to be in one direction – attitudes are assumed to underlie behaviour but sometimes behaviour affects attitudes (Ajzen and Fishbein, 2005).
The degree of attitude–behaviour consistency is affected by the situation in which the behaviour is performed, as well as who is performing it (Ajzen and Fishbein, 2005), and by whether attitudes are easily activated from memory and easily reconstructed (Glasman and Albarracín, 2006). The context of the object influences the attitude (Wigfield and Eccles, 2000). The more multidimensional the object, the stronger the correlation between attitude and behaviour. Attitudes that are stable over time also are more likely to influence behaviour (Regan and Fazio, 1977). When the discrepancy between attitude and behaviour is large, people tend to adjust their attitude to be more consistent with behaviour.
Studies of children’s motives for reading show that attitudes are important factors when intentions to read are used to predict reading. A positive attitude towards reading in general, together with previous reading habits, is one of the best predictors for future reading (Miesen, 2003).
There are two dimensions of attitudes which are interesting in the context of this study. The cognitive dimension focuses on the degree of usefulness, and the affective dimension focuses on whether the object is considered more or less pleasant (Ajzen, 2005). The attitude questions in the conducted survey were constructed according to this distinction.
Research on attitudes towards e-books and how these relate to e-book reading is still in its infancy. An American study, focusing on format preferences in different reading settings, concluded that the printed book is preferred when reading to children and when people want books which can easily be shared with others. E-books, on the other hand, are preferred when travelling or commuting and when people want to be able to get a book quickly. When asked about reading in bed and when a wide range of titles is wanted, people perceived printed books and e-books as equally functional (Rainie et al., 2012).
Although we can find some general principles in the adoption of innovations, the use of new media and the relationships between attitudes and behaviour, knowledge about how this works in the case of e-books is very limited. The extent to which the specific conditions in a small-language area, such as Sweden, will result in different behaviours when compared with other countries, especially English-speaking ones, remains unclear. Using representative surveys, this study investigates e-book-related behaviour and how this can be related to certain aspects of attitudes in Sweden. The research questions we address are as follows: How is e-book reading diffusing in the Swedish context? To what extent can background factors contribute to the understanding of e-book reading? Do people have positive or negative affective and cognitive attitudes to e-books? Is it possible to illuminate some of the sometimes complex relationships between attitudes and behaviour regarding e-books?
Method and data
We captured e-book reading in the Swedish population using quantitative mail surveys with representative samples of the Swedish population aged 16–85. The study was based on data collected in the Swedish National Society, Opinion, Media (SOM) surveys conducted in the autumn and winter of 2013 and 2014. Both surveys were made up of four slightly different questionnaires. The relevant part of the 2013 survey was administered to 3400 persons, with a net response rate of 54% (1706 respondents) (Vernersdotter, 2014). The relevant part of the 2014 survey was administered to 3400 persons, with a net response rate of 54% (1742 respondents) (Vernersdotter, 2015). For the first research question, we also have access to data for 2015–2017 from the annual SOM surveys.
The average questionnaire consists of approximately 20 pages and 80–90 questions, most of them with fixed response options. The question about e-book reading was built into a set of questions in the questionnaire: How often during the last 12 months have you read the following types of books? The answering options were divided into fiction and non-fiction with three distribution forms: as a printed book, as an audiobook and as an e-book. A seven-point scale was used: never, once during the last 12 months, once during the last 6 months, sometime in the last quarter, monthly, weekly and several times a week.
The attitude questions were constructed as two different batteries. In the 2013 survey, the questions were formulated in line with the affective dimension of attitudes and tried to capture attitudes towards the object on a more general level (Ajzen and Fishbein, 2005). The question was posed as follows: Which is your assessment of the following statements about e-books? An 11-point scale between 0 = Completely wrong and 10 = Completely right was used. In the 2014 survey, we designed a question similar to one posed in the Pew Internet & American Life Project (Rainie et al., 2012), which captures the cognitive dimension of attitudes (Ajzen and Fishbein, 2005). The question was Which is better for these purposes, a printed book or an e-book? Three answering options were used: printed books, e-books and no opinion. The statements and purposes are found in tables in the Findings section.
The independent variables in the study derive from theories of diffusion, supplement and displacement, and theories explaining the relation between reading behaviour and attitude formation as outlined in the theoretical section. These explanatory factors are traditional demographics, such as age, sex, level of education and household income, which in different ways have proven to affect both book reading and the uptake of digital devices. Sex and age (year of birth) were integrated with the data set from public registers. Education was measured in eight categories. Household income was measured in 12 categories and then divided into 3 equal groups based on national statistics on income in Swedish households.
To study the relationship among e-book reading, experiences of printed books (outlined in the theoretical section) and library use (most Swedish e-books are accessed via libraries), variables showing reading of printed books and library visits were chosen. Book reading habits were captured in the same set of questions as e-book reading. Library habits were put in a battery of lifestyle questions, measured with the same scale as e-book reading habits. Household access to reading devices also was included in the analysis and captured in single questions as either yes or no.
Together, the respondents in the survey provide an accurate reflection of the original sample including statistically significant data and possibilities of recognizing patterns in a large population, as well as in smaller groups. As with other statistical surveys, this one gives a general picture of e-book reading. The methodological choices and the limited number of questions, however, do not give more in-depth information about issues such as how much time is spent on reading, what e-books are read or how readers interpret reading on a digital device compared to in print.
Findings
The first and second research questions deal with the development of e-book reading and who reads e-books. E-book reading is not very common in the Swedish population, but it doubled between 2012 and 2014 (Table 1). In 2012, about 9% of participants had read an e-book within the last 12 months, compared to 18% in 2014. The largest increase was found between 2013 and 2014. The differences between reading fiction and non-fiction are small and not statistically significant. The frequency for e-book reading is diverse, and the growth is seen among both more and less frequent readers. After the analysis of e-book reading and attitudes to e-books, the reading habits were updated in 2015 and 2016 surveys. A stagnation then is noticed since 2014, and e-book reading was still found on the same level, 18%, in 2016 (Bergström et al., 2017). In 2017, however, the number of users again seemed to increase as did the sales of digital books, e-books and audiobooks. This is mainly attributed to new actors in the market and new business models with subscription offers for digital books (Wikberg, 2018). Thus, business models allowing for lower prices per unit and easy access to many titles seem fundamental for further diffusion of digital books. Compared to the situation in the United States, the proportion of the population reading e-books at least yearly is still lower in Sweden, but the use of audiobooks is higher in Sweden than in the United States (Perrin, 2018; Höglund and Wahlström, 2018).
E-book reading in total, fiction and non-fiction, 2012–2014 (per cent).
E-book reading is not evenly spread when comparing groups in society. As outlined previously, one can identify several factors contributing to an overall understanding of e-book reading (Table 2). For comparison, we included a regression analysis for reading of printed books (Table 3). Firstly, there are patterns resembling those of reading printed books; secondly, there are patterns of innovation diffusion. The most important factors explaining e-book reading overall were found to be tablet access, book reading habits and online library visits (Table 2). The more frequently participants read and use public libraries online, the more likely they were to have read an e-book. Age also plays a significant role in the take-up of e-books. E-book reading, especially reading of non-fiction, is more prevalent among young people than among older people.
E-book reading, 2014 (Ordinary Least Square [OLS], standardized beta).
*p < 0.05; **p < 0.01; ***p < 0.001.
Reading of printed books, 2014 (OLS, standardized beta).
*p < 0.05; **p < 0.01; ***p < 0.001.
It turns out that socio-demographic factors, such as sex, education and household income, do matter, but the impact of these factors is weak. Tablet access also contributes to the overall understanding of e-book reading but only to a limited extent and mainly for reading of fiction e-books, illustrating that access is a necessary but insufficient condition for use of e-books. Habits related to library visits do not have any significant impact on e-book reading, but it is the strongest predictor for reading of print books.
When studying fiction and non-fiction separately, some differences in the explanatory model occur. Although age and education contribute to the overall understanding of digital non-fiction reading, there are no significant effects of these factors for reading of fiction. Fictional e-book reading, on the other hand, is significantly related to both tablet access and reading of printed books. These factors also are significant for non-fiction reading, but the correlations are much weaker.
Table 3 includes a similar regression analysis for reading of printed books, which shows several differences in comparison with e-books. As a result of women’s higher preference for reading in print, sex is significant but only when we consider fiction. E-book reading also is related to reading of printed books, but to a larger extent, this is the case for non-fiction. Library visits is the most important factor explaining reading of printed books both for fiction and non-fiction literature. The total impact of the model in Table 3 is much higher than in Table 2, suggesting that e-book reading is more complex to predict from this kind of variable. This contributed to our interest in including attitudes in the analysis.
E-book diffusion, to a large extent, follows well-known patterns of technology acceptance found in previous research (Rogers, 2003; Venkatesh et al., 2012). It is evident that book reading experience and factors associated with it (Clark, 2013; Höglund, 2012; Verboord and van Rees, 2003) are driving forces for e-book reading.
Even though correlations are weak, it is interesting to note that e-book reading is slightly more common among men than among women. This is opposite to what is known about reading of printed books (Tepper, 2000; Höglund, 2012). One possible explanation is men’s higher access to media technology. In the process of diffusion of digital media applications, men have adopted several facilities earlier (Rogers, 2003).
The third research question deals with the affective and cognitive dimensions of attitudes (Ajzen, 2005) towards e-books and how they relate to use and to background factors. Affective dimensions are captured in a 2013 survey distributed to a representative sample of the Swedish population. The first question in a battery of four is a more general assessment of the likelihood of the e-book hypothetically replacing the printed book. A mean of 7.24 of 10 clearly indicates a strong position for the printed book (Table 4).
Attitudes to e-books and e-book reading, 2013 (mean, standard deviation and per cent).
Note: Percentage shares of respondents answering 8–10 on an 11-grade scale from 0 to 10 are presented.
The respondents also provided information on three more e-book-specific issues. The majority do not consider technical issues or the handling of technical devices to be major problems. E-books are only perceived to be too complicated to a relatively small extent (M = 4.96). The format and length of books also do not seem to be factors that prevent people from reading longer e-book texts (M = 5.34).
Another dimension of attitudes measured is how e-books affect the purchase or loaning of printed books. The possibility of reading e-books does not lead to people buying or borrowing fewer printed books (M = 3.89). One conclusion of this is that e-books complement rather than replace printed books.
However, even though people prefer printed books rather than e-books, there also is evidence for e-book support. Therefore, attitudes to e-books might differ regarding socio-demographic factors, reading habits, library visits and technology uptake (Table 5).
E-book attitudes in different groups (OLS, standardized beta).
Note: Respondents who have taken a position on each item are included in the analysis.
*p < 0.05; **p < 0.01; ***p < 0.001.
The most important factor related to attitudes towards e-books is book reading habits – especially the experience of digital book reading. E-book reading is the strongest predictor for the assessment of format and access, whereas reading of printed books significantly contributes to the understanding of the opinion that printed books never can be replaced by e-books. E-book readers express a stronger support for e-books (or a weaker support for the printed book). The overall experience of reading books in different formats clearly indicates that reading of printed books is strongly related to support for them. One should be somewhat cautious about fully comparing habits of reading e-books and printed books as the former must be considered immature in relation to the latter among all respondents in our survey. Currently, no one has grown up as an e-book native but all have had their first reading experience from printed books.
A few significant, but small, differences were found between sexes – for the replacement item and for the format and length items – and women were slightly less positive to e-books. Tablet access and attitudes co-vary, but the correlations are very weak. It is also evident that frequent library visitors find that the e-book format is best suited for shorter books and that they access fewer printed books in favour of e-books.
Thus, we find that the support for the e-book is stronger in groups with experiences related to digital reading and with access to, or current use of, digital tools, such as tablets and smartphones. It also is stronger among those with experience of library use through the Internet and with experience reading literature or non-fiction in digital form. It is evident that the relation between attitudes and behaviour found in previous research (Ajzen, 2005; Ajzen and Fishbein, 2005; Miesen, 2003; Solomon et al., 2010) also is valid for the e-book situation. Considering the high level of access to tablets, the factors holding back the diffusion of e-books seem to be related more to structural constraints, habits and attitudes than to access to technology (Bergström et al., 2017).
Cognitive dimensions of attitudes to e-books were captured in a 2014 survey that replicated an American study from 2011 (Rainie et al., 2012). In this context, the overall impression again is one of strong supports for the printed book when functions are studied (Table 6). This is especially evident regarding reading with a child. Printed books also are undoubtedly perceived to be better when sharing books with others and when reading books in bed. When reading books while travelling or commuting, printed books still are favoured, but the differences are smaller. Moreover, regarding having a wide selection of books from which to choose, differences are small, but e-books are perceived to be better when people want to get a book quickly.
Format preferences, 2014 (per cent).
Note: Respondents who have taken a position on each item are included in the analysis.
It is striking that many respondents do not take a standpoint on the issues. For four of the items, 30–40% said No opinion, while 15–20% answered No opinion for two of them. It seems that some situations are harder to evaluate than others. The shares of respondents who did not take a position on the matter vary when comparing different groups and are larger among older people than among younger people (58% and 23%, respectively). The percentages also decrease with higher level of education, and people with access to tablets have a lower share of No opinion answers than do not. Book readers – both of print and digital – and people who visit public libraries and online libraries more frequently also have significantly lower percentages of No opinion answers than people with less contact with books and libraries. It is evident that it seems harder to express preferences for format when you have less experience with books in either form. In this sense, attitudes, on a general level, are related to use.
With respect to this issue, a multivariate analysis revealed factors that might have an impact on attitudes. It is evident that reading printed books is the single most significant factor when trying to explain format preferences in the book reading sector (Table 7). The more frequently people read a printed book, the less support they have for e-books, no matter the context or function. For three of the situations, e-book reading habits have a significant but weak correlation with format preferences. E-book readers are more positive towards reading e-books in bed but more negative to e-books when wanting a wide selection of books and when sharing with other people. Habits and experience affect assessments, but the habit of reading e-books is not yet as settled as the reading of printed books is. The printed book seems to be the benchmark for readers’ assessment.
E-book attitudes in different groups (OLS, standardized beta).
Note: Respondents who have taken a position on each item are included in the analysis.
*p < 0.05; **p < 0.01; ***p < 0.001.
As for the rest of the analysed factors, few and very weak significant co-variations were found for format preferences. It is clearly a matter of book reading habits. Cognitive attitudes are only more strongly related to reading habits, whereas affective attitudes had a somewhat larger range of explanatory factors. In both cases, it is evident that use correlates with attitudes on a more general level, just as suggested in the literature (Ajzen, 2005; Ajzen and Fishbein, 2005; Miesen, 2003; Solomon et al., 2010). This study cannot, however, reveal the direction of the correlations.
As outlined in the theoretical section, transformation from one medium to another depends on functions and needs (Dutta-Bergman, 2004; Flavián and Gurrea, 2009). Our findings reveal that the need for e-books is limited and the digital format should be seen as complementary to printed books both when analysing reading and attitudes.
The relation between reading and attitudes
Although regression coefficients may seem low, there are rather strong correlations between different sets of attitudes and book reading, especially if we look at respondents with the most positive attitudes to e-books. There we reach a proportion near the actual proportion of readers, which for the whole population studied was 11% in 2013 and 18% in 2014. Even in the groups most favourable towards e-books, the proportion of reading printed books is still much higher than for e-books but diminishes as the attitude towards e-books increases. Whether the attitudes precede behaviour or come as a result of trying e-books may vary, but it seems to go both ways, and we can assume that attitudes become more favourable with more frequent use of e-books.
The two sets of attitudes from the 2013 and 2014 surveys were summarized in indexes. The four items of affective attitudes were added into low, middle and high positive attitude. The cognitive attitudes were added into three groups where the preference for e-books varied between 0 and 6 (Table 8). The indexes then were related to reading of literature in print and as e-books.
Proportion reading literature in print and e-book after attitudes towards e-books and format preferences (per cent and gamma).
The e-book attitude indexes are both clearly related to the reading of literature in print (negative correlation), as well as to the reading of literature in the form of e-books (positive correlation). The correlation between attitudes and reading is stronger for e-book reading than for reading printed books. This suggests that attitudes may be more favourable after trying e-book reading, but also that reading in print will remain the dominant form for quite some time.
As a summary of the empirical part of the article, we repeated the regression analysis on e-book reading, now including the cognitive attitudes to e-books. The attitude index on functions was used in the model. It is evident that attitudes to print and digital formats in different settings can contribute greatly to the understanding of e-book reading (Table 9). The explanatory value of the model increased to 0.18 (compared to 0.11, Table 2, total e-book reading).
E-book reading, 2014 (OLS, standardized beta).
*p < 0.05; **p < 0.01; ***p < 0.001.
Attitude towards print and digital books is the single most important factor in the model when considering all the other factors. The more positive the attitude, the more use and vice versa. No matter the direction of the correlation, the interrelation is strong, just as in many other situations (Ajzen, 2005; Miesen, 2003; Solomon et al., 2010). The correlation between attitudes and behaviour in the e-book context is stronger for affective issues than for the cognitive items (see Ajzen, 2005). Habits of book reading and library visits also contribute to the overall understanding of e-book reading but to a lesser extent. The effects of background factors, income and tablet access are negligible when analysed together with other factors (Table 9).
Conclusions
This study used data from representative mail surveys of the Swedish population aged 16–85. The first research question investigated who the readers are and to what extent they seem to fit the general pattern found in theories on diffusion of innovation and new technology. Readers of e-books remain a small but growing minority. They tend, more often, to be young or in early middle age, to have a higher formal education and to have access to reading devices. The most important factors explaining e-book reading, in addition to attitudes towards e-books, are, however, library visits and the habit of reading printed books. E-book use is more even between men and women than was expected given what is known about book reading in general. Men tend, more often, to be early adopters of the necessary technology, but women read literature and visit libraries more often. It can be assumed that these two tendencies are in balance and that the proportions of male and female e-book users may change in later stages of the adoption process, especially when reading of fiction literature is considered.
The users, so far, seem to correspond with many of the characteristics of early adopters (Rogers, 2003), and theories of innovation diffusion, therefore, will be relevant for future analysis. Turning to theories on substitution and displacement, it is obvious that e-books fill a complementary role, as even frequent users of e-books also read print frequently, and a majority still prefer the print format. It is evident that e-books have found a niche in certain contexts and in situations where fast access or mobility is important and that old and new formats are suitable in different settings (Dimmick, 2003).
Turning to functions of e-books, perceived usefulness and perceived ease of use (Davis, 1989) – or to speak of Rogers’ (2003) relative advantage and complexity – the findings imply limitations on these issues on the part of the readers. The relative advantage is not perceived to be particularly high, and with little or no incentives in terms of for instance lower prices, the arguments for choosing e-books are few.
The findings also point to the perceived usefulness of the e-book when travelling or commuting (Davis, 1989). Taken together, there are functions of the e-book that attract users (cf. Carey and Elton, 2010; De Waal and Schoenbach, 2010) but not to an extent of making them displace the printed book. Up to now, the e-book format has, to a large extent, resembled that of the printed book. Future e-books might well take more advantage of interactive features, and with the development of e-books into multimedia products, attitudes and future patterns of use might change.
The set of affective attitudes towards e-books show a general correlation with e-book reading in all four cases. Two of the questions also were related to reading printed books. The explanatory power of age, sex, level of education and library visits, together with book reading, is still limited, partly as a result of the small proportion of respondents who have read an e-book. It is reasonable to assume that attitudes may change with increased use of e-books. The reading of printed books is, however, an important factor indicating a need or an interest in reading. It is important to emphasize that the respondents in the conducted surveys all belong to a generation of print book readers. The printed book is their benchmark when evaluating e-books. Future research will have the opportunity to study differences between e-book immigrants and e-book natives to reveal the impact of the situation in the formative years on consumption and attitudes.
When comparing functions of print and e-books in different reading situations, the impact of print experience also is evident, as is experience of e-books, level of education and online library visits. When adding attitudes to the explanatory model, significant interrelation between attitudes and behaviour occurs, just as suggested in previous research (Ajzen, 2005; Ajzen and Fishbein, 2005; Miesen, 2003; Solomon et al., 2010).
Attitudes can stem from different sources. Habits and earlier experience are important aspects which also seem to be relevant here. Attitudes, however, also can be related to taste, social status and the social circles to which people belong, none of which were assessed in this study. Moreover, other social aspects, such as conspicuous consumption (very limited for e-books), can work as constraints suppressing the radical potential of the e-book (Winston, 1998). Future research should study different aspects of attitudes and their relation to the reading of e-books in more detail.
This study was conducted in a small-language context where the native e-book market is relatively small compared to English-speaking countries. This is reflected in the findings not least with regard to the overall uptake of e-books being lower than in countries such as the United States. Assumedly, a small market as Sweden will be more easily affected by tiny incremental factors (e.g. change in pricing, successful distribution channel, competition of foreign language books) than larger markets. Continuous studies of similar contexts are necessary to reveal whether the supervening social necessities will out-rule the law of suppression of radical potential.