Books in the Time of Screens: The Reading Habits of Slovenian Students

This research paper brings up the question of recreational reading habits among the generation of university students that were born after 1990 and grew up in a digital environment. The paper focuses on their self-reported reading habits. The students belong to two disciplinary domains (humanities and social sciences) at the University of Ljubljana, Slovenia. A significant number of them were pre-service teachers. The research was conducted on a non-random sample of 429 students. Data were gathered using a printed questionnaire that included assessment scales of students’ reading habits. These data were compared with data gathered in the online reading habits survey of the general Slovenian population that took place in autumn 2019 on a sample of 1000 participants, demographically representing Slovene general population. The key finding was that university students read more than members of the general population do, but still less than would be expected of future educators who will introduce book reading to pupils as the number of non-readers in student population was higher than expected. In addition, the data revealed that practically all students were able to read in at least one foreign language (predominantly in English) what is significantly higher than in general population. The paper indicates a few reasons for such outcomes and addresses the question of why recreational book reading matters in educational settings.


Introduction
The authors examine a range of data taken from a study on the reading and studying habits of students of the Faculties of Arts and Education at the University of Ljubljana (Slovenia) and compare them with the findings from a Slovene reading habits survey (The Book and Readers VI, 2019). The main research question this paper deals with is whether the decrease in recreational book reading among the general population walks hand in hand with the reading habits of university students, who represent the recruitment base for pre-service teachers and promotors of reading in primary and secondary education. The study also aims to address the impact of family reading on reading habits.
The starting point of this research was an observation that deep and long-form reading is losing ground in the digital landscape (see Baron, 2015;Kovač & van der Weel, 2018;Wolf, 2018). If this is, indeed, the case, such developments might call into question readers' abilities for abstract thinking, which is one of the known benefits of deep reading. In other words, one of the central cultural and pedagogic dilemmas of our time is whether the deep reading of long-form texts still matters as one of the main components of training for abstract thinking, or have we developed new, more suitable training processes for abstract thinking by using digital learning tools? Is so, has the deep reading of long-form texts become a cultural residuum with little or no contemporary practical value? We have no clear-cut answer to this question, but answering it is not one of the aims of this paper. Our goal is more modest: Research studies such as ours are trying to determine whether deep reading is, indeed, becoming an endangered cultural practice or not. In other words, if book reading habits are not going away, then training for thinking persists in a business-as-usual manner; but if reading habits do go away, or if long-form, deep reading is even not practiced by pre-service teachers, then the question outlined above turns from an academic research question into a pressing political issue, especially if reading habits are being lost among pre-service teachers. To answer this question, what needs to be checked is whether the debate on diminishing book reading practices can be backed with data and, then, what trends do these data reveal?

Indicators and Measurements of Reading Habits
Together with other data, publishing statistics represent useful indicators of reading habits. For the last 50 years, for example, book production (measured in terms of the number of titles published annually) has been on the increase globally, but book sales and library book borrowing have not been keeping pace. Thus, in developed countries, the number of newly published titles increased by more than 1000% between 1950 and 2008, but the copies of books sold only increased cumulatively by about one third, which is approximately the same figure as the increase in the population (Kovač & Wischenbart, 2019). Since 2008, such trends have very likely been even more intensive, but due to the rise of ebooks,which are outside regular book statistics, we have no reliable figures.
Comparisons among countries with different reading cultures shed additional light on reading processes. In Norway, a country with highly developed book reading habits (Leserundersøkelsen …, 2018), there were, on average, about 10,000 new titles published yearly (i.e., 2000 per million inhabitants) in the first decade of the new millennium. This is similar to the figures in Slovenia (2000-2500 titles per million inhabitants) (Kovač, 2019). In 2014, 14 books (three in English) per inhabitant were sold in Norway, while fewer than three were sold in Slovenia, which is the same as the number of books in English sold per head in Norway. In the same period, four books per adult were borrowed from Norwegian public libraries, while Slovene public libraries loaned seven per capita (loan period extensions excluded) (Kovač, 2019, p. 73;Leserundersøkelsen …, 2018).
In sum, an average reader in Norway consumes twice as many books each year as his or her counterpart in Slovenia. By contrast, fewer than 1000 books per million inhabitants are published in Romania, one book per inhabitant is sold, and library book borrowing is practically unknown (Kovač, 2017).
Book reading habits can also be analysed using the findings of Southerton et al. (2012), who looked at recreational activities in different countries. Such studies, based on time diaries, were conducted in the Netherlands, Norway, Germany, France and the UK for several decades. A couple of hundred selected people, ones demographically representative of the entire population, kept diaries about their recreational activities, and book reading was always one of the activities included. The findings of these studies show that the time devoted to reading books has decreased in all of the countries, except Norway. This is what confirms Norway's status as one of the countries with the most developed reading habits.
There are no such studies of recreational activities in Slovenia; therefore, we have no comparable data to say how much the amount of time spent on book reading has changed, on average, over time.
Despite the above, Slovenia has a long tradition of reading habits surveys. The first study, called Knjiga in bralci (The Book and Readers), was published in 1974 (Kocjan, 1974), and then the study was revisited in regular 5-or 10-year intervals. The Slovenian reading surveys use similar questionnaires as those in other countries, with many questions recurring in the same form throughout the decades; for instance, 'How many books have you read in the last 12 months', 'How many have you bought and/or borrowed from a library?', 'What literary genres do you read' and so on. Thanks to such recurring questions, changes in reading habits are revealed.
It is also important to note that the methodology of reading surveys has significantly changed in the last 40 years; until the late 1980s, reading research was usually done as fieldwork -interviewers called on a pre-determined sample of interviewees at their homes and filled in pre-prepared questionnaires with their answers (Rupar, 2019, p. 11). After 1990, the studies began to be conducted over the phone, and then online. Today, they are generally done online with part of them -often for older interviewees -done via telephone. The samples of participants reflect the educational, regional, gender and age structures of the population where the study takes place (Rupar, 2019). 1 1 In Slovenia, online surveys started in 2014, whereas until 1999 they were done as fieldwork. The studies done in 2014 and 2019 were financed by the Slovenian Book Agency, which is organisationally managed by the UMco publishing company, and the surveys were carried out by the Valicon agency as a combination of online and phone interviews. Both the surveys included a little over 1000 respondents. We draw attention to the change in research methodology mainly because it has some impact on the respondents' replies: According to Holbrook et al. (2003) and Fricker et al. (2005), respondents tend to give fewer socially desirable responses online than to in-person interviewers, so online answers are more honest and accurate. This means that the findings on reading habits from 1979 are almost certainly a little too optimistic, and they show a better reading image than was actually the case. Despite the probable optimism of the 1979 findings, the picture of today's reading habits remains poor.
Research findings indicate that the share of readers in society (who are broadly defined as those people who have read at least one book in the last 12 months) was practically the same in 2019, in 2014 and in 1979. The findings also show that, for the past 40 years, half of Slovenia's population had not read a single book during the year in which the survey took place. The number of self-defined 'non-readers' in Slovenia is smaller (37%) than the number of those who say they have not read any book in the last 12 months, which suggests that reading books is still a socially professed, desirable value (for more on that, see Rupar, 2019, pp. 13-14). However, similar research conducted in Norway in the period indicates a 20% growth in the number of book readers ((Leserundersøkelsen …, 2018) With the above in mind, at least Norwegian data do not confirm the thesis that, in the pre-digital past, the number of book readers was higher than it is now and that people today read fewer books than they once did. On the contrary, in Slovenia, the data are telling, in terms of the failure of reading culture in the last four decades.
In such a context, the reading habits of the Slovene tertiary-educated population are especially interesting. In 2019, only 80% of people with a tertiary degree were readers. In addition, between 2014 and 2019, the share of intensive readers decreased: For those who read more than 20 books per year, the number fell from 9% to 6%, and for those who read more than 10 books per year, it declined from 18% to 14%. By contrast, in Norway, the share of intensive readers has remained virtually the same: The number of those who read more than 20 books per year has only decreased from 18% to 17%, and the share of those who read more than 10 books per year has remained unchanged (37%) (Leserundersøkelsen …, 2018; Kovač, 2019).
Further, in Slovenia, the number of families who report reading often to children also fell by 3% (Kovač, 2019, pp. 71-72), whilst in Norway, this percentage went up by 3% (Leserundersøkelsen …, 2018). Notably, the number of families who often read to children is about three times higher in Norway than in Slovenia. Such data on reading to children are intertwined with the data on the number of books that people have at home, that is, home libraries. The average home library in Slovenia has 80 books, while the averages for Norway, the Czech Republic and Sweden are about 200 books (Sikora et al., 2019).
Since the findings of the first Trends in International Mathematics and Science Study (TIMSS) were made known (1995; see also 1999, 2003), it has generally been accepted that there is a positive correlation between the number of books at home and children's performance in TIMSS assessments (Vehovar et al., 2009). International studies have also demonstrated that the children of parents who have more books at home achieve better learning outcomes than those who have fewer books at home (see Following these data, Slovenia is at the bottom of OECD member states regarding reading literacy, far behind Sweden, the Czech Republic, Norway and Estonia, and accompanying countries such as Italy, Spain and Greece (Sikora et al., 2019). In such a context, the reading habits of students and pre-service teachers could be seen as an indicator of the past, present and future of reading literacy in a given country.

Research Problem
This paper focuses on three aspects of student reading habits: how students spend their leisure time, what their reading habits are and what the correlation is between reading education in the immediate family and later reading habits. The research investigated four variables, namely: 1) Activities to which students dedicated the majority of their spare time, and the placement of reading among them. 2) Students' reading habits, regarding the number of books they read, the foreign languages in which they read, library visiting and the types of books they read.
3) The influence of the mother's education level on the frequency of reading in the immediate family. 4) The impact of family reading on reading habits.
On this basis, two main hypotheses were tested: H1: The mother's education is an important factor in influencing the frequency of reading in the immediate family.
H2: Through frequent reading, parents have the most important influence on students' reading habits.

Research Methodology
The study was conducted in the academic years 2017/18 and 2018/19 and included students from two faculties that educate pre-service teachers. Students completed a questionnaire about their leisure activities and reading and study habits.
The non-random sample consisted of 429 students of the Faculties of Arts and Education at the University of Ljubljana, 305 (71.1%) of whom were studying at the bachelor level, and 124 (28.9%) of whom were master's students. Taking into account the number of students enrolled in the two faculties (6689 students in the year 2018/19 (Univerza v številkah, 2020)), our sample represented 6.4% of the overall population, which indicates an adequate level of representation.
A majority of students included in our sample were female (86.9%, male 13.1%); their average age was 22.5 years.
The data was collected using a paper-and-pencil questionnaire consisting of 34 closed-ended questions and an assessment scale. The questionnaire was structured in three thematic blocks: leisure activities, reading habits and study habits.
The reliability of the assessment scale was tested with Cronbach's alpha (α = 0.71) and validated using factor analysis (the first factor explained more than 20% of total variances). The survey was carried out during lectures and seminars with university teachers present.
The data were processed at the levels of descriptive and inferential statistics. Structure tables were used (f, f%) to show the data, and the hypotheses were tested using the chi-square test. In cases when the conditions for this test were not met (when more than 20% of the expected count were under 5), the Kullback test was used instead.
We were first interested in what our responding students did in their free time. They were asked to choose their five most-frequent activities, which we then ranked (see Table  1). They were also asked to provide us with the average number of hours per week that they spent engaged in these leisure activities. As the frequencies of the students' answers shown in Table 1 demonstrate, the responding students spent their leisure time most frequently working and looking for entertainment on a computer and/or smartphone (89.7% out of 429), followed by socialising with friends (83.4%), listening to music (72.2%), playing a sport (56.6%) and watching films (44.5%). The least frequent activities were visiting shopping centres (9.1%), reading daily newspapers (7.9%) and reading magazines (4.2%). Reading books ranked in the middle; it was one of the leisure-time activities chosen by 44.3% of the students.
Looking at the average number of hours that the respondents spent on each activity, work and entertainment on a computer and/or smartphone came first (17.4 hours), followed by listening to music (11.8 hours) and socialising with friends (10.9 hours). They spent the least time reading magazines (5.1 hours) and daily newspapers (3.8 hours) and going to shopping centres (4.0 hours). According to the respondents, they read more (8.1 hours) than they played sports (6.6 hours), but the average number of hours spent on reading was significantly lower than the number of hours spent on computers and/or smartphones.
Just under 90% of the respondents said they spent their free time on a computer and/or smartphone, and 44% of the respondents chose reading as a leisure activity. For humanities and social science faculties, and especially for teacher-training programmes, this is not a very optimistic figure. The share of students who did not consider themselves to be readers (16%) was lower than the general population's mean (49%) by two thirds, but it was still high, given that these were the students of humanities and social science programmes who might end up working in educational institutions. Among those who defined themselves as readers, the share of the students who read every day was smaller (6%) than the same share in the general population (9%). The share of students who read 20 or more books per year was comparable with the general population (10%, 11%), but it was smaller than the share of intensive readers in Norway (17%) (Leserundersøkelsen …, 2018).
The figures in the Slovene reading survey (Knjiga in bralci VI 2019) regarding people with higher education show that 20% among them are non-readers, while our study of students shows that 16% of them self-defined as non-readers. The highest share of non-readers was among the students of educational sciences in the Faculty of Arts (20%) and the students of the Faculty of Education (17.5%).
The reasons for non-reading cited by those who had not read a single book in the previous year (16% out of 429) were also interesting. Among the 58 respondents who replied to this question, 43.1% said they preferred doing something else, just over one third (34.5%) said they did not have time for reading and 15.5% felt no need to read. Of the 83.9% of the students who said they were readers, 15% did not remember even one book they had read in the previous 12 months (in the entire sample, the share of those who self-identified as non-readers and those who selfidentified as readers but could not remember a single title was 28.5%). Just under 10% listed the titles of the books they had read as compliance reading (in secondary or even primary school, which casts further doubt on their responses regarding the number of books they had read in the previous year). If the students who did not remember any titles or who listed as the last books they had read compliance reading in primary-/secondary-school were categorised as non-readers, the share would equal the share of non-readers in the general population.
In addition, as seen in Table 3, students do not read demanding books in their free time. When categorising the books, we formed four categories. A special category includes the books that we know were part of school syllabuses in primary and secondary schools, which today's students had to read as part of their homework. We also included the share of students who could not remember any title. The distribution of the books in tables like this is necessarily subjective, but we believe the table reflects the general trends in the levels of difficulty.
One-third of the students (33.3%) listed the titles of less demanding contemporary fiction, popular fiction and somewhat more demanding books for personal growth. More demanding popular literature was cited by just over 35% of the students, and prominent literature and essays requiring intellectual effort were listed by only 7% of the responding students. When listing the books they had read, 26% of the respondents (also) cited works in English; other foreign languages were negligible (German 1.1%, Italian 0.8%), but if we look at the students' responses to the question about what languages they read in, the percentages are significantly higher ( Table 2).
The figures in Table 4 show that the students whose mothers had higher (tertiary) education levels had been read to statistically significantly more frequently in the family (just under 53%), compared to those students whose mothers had lower education levels (secondary, 39.6%, and primary school or lower, 20%). Although the share of students whose mothers had lower education levels (primary school or lower) was small, our findings show that more than half of them (53.3%) had been only 'occasionally' or 'never' read to in the family, while the share among the students with better-educated mothers was significantly smaller (25% and 14.1%, respectively).
Similarly, high deviations were found when we compared data in the student survey with data from the general reading survey (Table 4). Those who had often been read to in childhood accounted for just under 80% of the students, which is almost three times the share of the general population (29%) and practically the same as in Norway (Kovač, 2019, p. 71). This Norwegian figure, together with the small share of non-readers, confirms the thesis on the importance of family reading (Leserundersøkelsen …, 2018).  The findings presented in Table 5 suggest that the frequency of family reading is a key factor influencing future reading habits.
The figures in Table 5 indicate that the students whose parents had read to them every day statistically significantly more frequently (just under 66%) stated that it was their parents who had had the biggest influence on their reading habits. By contrast, only a quarter of the students whose parents had occasionally read to them or had not read to them at all quoted their parents as an influence. The impact of school as a factor in reading habits was greater for those students whose parents had occasionally read to them. Just above one fifth and about one quarter (21.4% and 25%, respectively) of the students who stated that their parents had occasionally read to them or had not read to them reported no influence on their starting to read.

Discussion
Various research findings (DeBaryshe, 1995; Weigel et al., 2006) have indicated that the mother's education level, income and reading habits have a positive link on the frequency and quality of reading to a child (Marjanovič Umek, 2011,p. 19). The best readers have educated parents (Duursma, 2014;Lyytinen et al., 1998;West et al., 2002), and it is important that adolescents start to engage with books at home early on, because it 'not only effects on literacy, but also adult education, occupation, and reading at home', and it creates 'a lifelong propensity to routinely include books into one's cultural and material environment' The findings in our survey are in accordance with findings in the above-mentioned papers. In short, our research confirmed our main hypothesis. Also, in Slovenia, the mother's education is an important factor for influencing the frequency of reading in the immediate family, and through frequent reading, parents had the most important influence on students' reading habits.
However, if we return to Slovene reading habits in general and compare them to the reading habits of students in university teacher-training programmes, we see that not all indicators show better results for the latter than for the Slovenian average; even more so, some indicators reveal the students' reading habits to be even worse than country's average (Table 2). Consequently, the reading habits of Slovene students are worse than the reading habits of the average population in Norway.
The findings regarding study habits were not really encouraging, either. More than three-quarters of the respondents (76.4%) studied only from their notes for more than half of their courses, and significantly fewer consulted original texts, that is, books and articles. Based on smaller-scale studies in the USA, Hoeft (2012), too, observed that students do not read original texts, and those who do have problems understanding them. Thus, students are only asked to read extracts and notes (see also Cirino et al., 2013;Datta & Macdonalds-Ross, 2002;Huang et al., 2014).
As these students are the future 'guardians' of reading habits in the educational system, all of this confirms our second hypothesis, that Slovene students read less than would be expected of pre-service teachers.

Conclusions
As stated at the beginning, the aim of this paper was to check whether the debate on diminishing book reading habits can be backed with data and, if so, to identify what trends these data reveal. We believe that, at least in Slovenia, our findings confirmed the existence of such trends. This legitimises the question of whether or not the deep reading of long-form texts still matters as one of the main components of training for abstract thinking, or if can we substitute it with new, more suitable digital learning tools. In other words, if book reading matters less, what is the quality of its substitute(s) that introduces students to abstract thinking? On the contrary, if book reading still matters as training for thinking, who will educate future readers if some of the educators are losing ground with their own reading habits?
As the answers to such questions are closely connected to the quality of primary and secondary education, we do not consider this to be only an interesting academic research issue, but rather it is also a pressing political question that, as such, needs to be addressed not only by teachers and the producers of learning tools, but also by policy makers and parents. The authors of this paper do not pretend to have an answer to the problem; additionally, we do not purport that our findings have an international value per se, as they are based on research that took place in only one country. Nevertheless, on basis of our findings, we are convinced that it makes sense to address such issues in an international environment and call for a wider debate on reading trends and the consequences of these trends for the quality of education.