ABSTRACT
Several researchers have applied Bourdieu’s concept of habitus to studies regarding the experience of space and architecture, concluding that respondents often experience surroundings very differently, depending on their social class. Therefore, it can become problematic when public institutions opt for an architectural style that could discourage members of lower/working classes to visit. We believe this to be the case for public libraries, which are housed increasingly frequent in very abstract, iconic buildings. In this article, we present different analyses into this topic performed on data that was gathered in a large scale user survey in public libraries in Flanders and a literature review that explores both the subject of iconic architecture and the relationship between class, architecture and aesthetics.
Introduction
The ‘social-pedagogic motive’ (Huysmans, 2006), the idea that every individual must be offered the chance to develop themselves through culture so that they might increase their opportunities for societal advancement, has been a driving force behind public library policy since the end of the Nineteenth century. The notion of ‘societal advancement’ of course suggests that public libraries are predominantly aimed at the more vulnerable members of society: members of the working class or those with lower incomes and lower levels of educational attainment. While public libraries have always been tasked by policymakers to reach especially this segment of society (De Pauw, 2005; Huysmans, 2006), it could be argued that they are not meeting this objective: the majority of library visitors are (and always have been) middle class, both in Flanders (Glorieux, Kuppens, & Vandebroeck, 2007) and in most other Western countries (Huysmans & Hillebrink, 2008; Manzuch & Maceviciute, 2014; Sei-Ching & Kyung-Sun, 2008). Results from the Flemish Participation Survey (2014), a large scale research in which a standardised questionnaire on recreational time use is presented to a representative sample of the Flemish population, suggests a worrying trend concerning library attendance with regard to lower educated visitors especially. Library attendance has dropped with 6.6 per cent on average between 2004 and 2014, but disproportionately so among visitors with lower levels of education (Lievens, 2015). We would therefore argue that a suboptimal situation (a historic underrepresentation of visitors with lower levels of education) has gotten worse in recent years.
Research aimed at understanding why lower educated visitors are underrepresented in public libraries tends to focus on factors of a very practical and economical nature (such as membership fees or opening hours). These proved very useful when looking at, for example, the underrepresentation of men in Flemish public libraries (Glorieux et al., 2007), but seem to fall short of providing any type of meaningful explanation for the underrepresentation of lower educated users. Therefore, in this contribution, we will approach the topic from a more ‘class-cultural’ perspective in which lower educated visitors are conceptualised as a (more vulnerable) social group that experiences (a form of) symbolic violence when trying to participate to a cultural institution, in this case public libraries, that is dominated by the middle class. Of course, we are not the first to point out that certain public institutions that should be ‘open’ to all, appear to (unintentionally) exclude more vulnerable groups in society: Dawson (2014), for instance, found that science museums tend to discourage the attendance of low-income and minority ethnic groups by maintaining ‘problematic assumptions about the capital that participants could bring to their visits, notably speaking and reading English, understanding scientific terms and concepts, available financial resources, “free” time, and familiarity with ISE institutions’ (p.1004). Instead of focusing on barriers such as income or language, in this contribution we emphasise the importance of the architectural style of the building housing a public institution as both an overlooked and relevant threshold. We will pay special attention to the rise of ‘iconic’ architecture and question whether this specific (rather abstract and artistic) architectural style is a good one for public libraries. Of course, claiming that the architectural style in which a public library is built will have effect on the attendance of lower educated visitors and problematising iconic architecture, raises two questions we must consider:
Does the architectural style of a library building really play a considerable role in the perceived invitingness of a public library?
Is education level connected to the appreciation of iconic architecture?
In section two, we first discuss the use of iconic architecture in public buildings, followed by an overview of sociological theory regarding the nature of aesthetics and its relationship to social background. In section four, we will first answer research question one by discussing several analyses based on a research we conducted ourselves. Next, we will present an exploratory analysis (also based on our own research) which we believe indicates that education level plays a significant role in how visitors experience iconic architecture and interpret the building and invitingness of their library. We consider determining exactly which aesthetic elements or styles repel or attract different social classes as a larger research task we wish to participate in in the future, but the question will not be addressed in this contribution. With this research, we hope to contribute to the field of study regarding class-based dispositions to art and aesthetics, but also add a new perspective to the topic of the inclusiveness of public cultural institutions.
Theory
Iconic architecture
‘Iconic’ architecture has been on the rise ever since Frank Gehry helped revitalise the city of Bilbao with his design for the Guggenheim Museum in 1997: ‘Bilbao posed a big problem for the Basque Government. The city’s century-old steel industry was fast becoming obsolete. No less than thirty per cent of Bilbao’s population was unemployed and terrorism and drugs were ripping the city’s social fabric apart. The city needed to be re-invented. (…) Bilbao became a benchmark for others (…) eager to put themselves on the map, attract tourists and boost the local economy. All that was needed, it seemed, was a special building design by a star architect’. (Moix, 2012, p. 38) Whether the so-called ‘Bilbao effect’ can actually have that much impact on a city’s development has been called in to question increasingly frequent in recent years, Dickson (2017) rightly underlines that before the completion of the Guggenheim, ‘regeneration had been under way for nearly a decade. In 1988, a new metro system was commissioned from Norman Foster, followed by César Pelli's major “master plan for Bilbao” in 1989, which suggested razing industrial areas and redeveloping them into business, residential and leisure hot spots’.
Despite justified doubts regarding its ‘regenerative’ potential, the use of iconic architecture in attempts to try and kickstart a positive development in a neighborhood or even an entire city gained in popularity and did not limit itself to museums. In the last two decades, a considerable amount of cities have constructed ‘iconic’ public libraries. The first to really follow in Bilbao’s example, was the Seattle City Library. Designed by famous Dutch architect Rem Koolhaas, it opened in 2004 and was recently dubbed the ‘most Instagrammable’ library in the world (Davis, 2018). Since then, it seems that the ‘Instagrammability’ of public libraries has become increasingly important. An (incomplete) overview of cities (outside of Belgium) that recently opted for a library in an iconic style (Kowalczyk, 2017) illustrates how widespread the phenomenon has become:
|
|
|
|
The decision-makers spearheading these ambitious projects are often lauded when things come to fruition: a landmark building is seen as a sign that the public library has successfully shifted gears from being a ‘stuffy’, old-fashioned institute to a vibrant and ‘hip’ place ‘for everyone’. We would like to challenge this unbridled optimism somewhat by asking: who are these libraries really for? The Guggenheim in Bilbao was built to ‘attract tourists’, arguably a valid objective for a museum. But what happens when public libraries are built mainly to attract tourists, to be iconic, to be ‘Instagrammable’? As we argued in the introduction, public libraries are (supposed to be) highly approachable, publicly funded cultural institutions carrying out an emancipatory task. Does it really make sense then, to model them after museums for modern art, one of the least approachable public institutions? Data from the Flemish Participation Survey (2014), clearly shows that the underrepresentation of lower educated visitors is greater in museums for modern art than in public libraries. For example: 26.8 per cent of respondents who (only) completed secondary education regularly visit public libraries, while only five per cent of this group visit museums for modern art. Is there then no danger of discouraging people with lower education levels, who already visit public libraries far less and whose attendance culture workers are trying to increase?
Paul Jones discusses the subject of iconic architecture in ‘The Sociology of Architecture’, stating that ‘“political agencies” recent embrace of what has come to be known as “iconic” architecture can be understood as a continuation of longstanding attempts to mobilise major building projects, first, to materialise wider discourses of major social change, and second, to generate surplus value from urban space’ (2011, p. 115). The use of iconic architecture is recognised as a powerful tool in ‘rebranding’ certain sites, neighborhoods and even entire cities. However, Anne-Marie Broudehoux rightly argues that the act of ‘rebranding’ also implies a reduction of ‘several different visions of local culture into a single vision that reflects the aspirations of a powerful elite and the values, lifestyles, and expectations of potential investors and tourists. These practices are therefore highly elitist and exclusionary, and often signify to more disadvantaged segments of the population that they have no place in this revitalised and gentrified urban spectacle’ (2004; in Jones, 2011, p. 126). Again, the Guggenheim in Bilbao is a prime example of this: local residents of the lower and working classes recognise the positive economic impact the museum has had on their city, but generally feel little to no connection to the institution ‘in terms of quality of life, social cohesion, regional identity or governance’ (Jones, 2011, p. 116). Culture-led regeneration projects often display an uncoupling between local sentiments and political aspiration, with ambitious policymakers struggling to ‘sell’ their vision to locals. In an attempt to remedy these reservations, which can turn into roadblocks for the actual completion of the project, the practice of community consultation is deployed, the actual role and impact of which are often very ambiguous. Jones goes so far as to categorise them as ‘legitimation exercises’ for schemes that will be selected ‘regardless of the outcome of the public vote’ (2011, p. 127), because “it is not always clear what ‘consultation’ or ‘community empowerment’ actually mean in practice” (2011, p. 126). In short: iconic architecture often appears after a problematic decision-making process where the voices of (working class) locals are ignored. Dutton and Mann posit that ‘to make architecture is to map the world in some way, to intervene, to signify; it is a political act. Architecture, then, as discourse, practice, and form operates at the intersection of power, relations of production and culture, and representation, and it is instrumental to the construction of our identities’ (2000, in Jones, 2011, p. 166). The fact that iconic building projects are frequently contended, signifies that they tend to lack grassroots support. That iconic architecture has nonetheless become ubiquitous, suggests a problematic and lopsided decision-making process.
One might ask of course: how do we define ‘iconic’ architecture? Categorising and defining aesthetic qualities is invariably a precarious undertaking, especially when operationalising it as a concept for social scientific research. Once again, we turn to Paul Jones for clarification. He defines iconic buildings as ‘distinctive’ and ‘eye-catching’ (2011, p. 121): ‘icons, while maybe or maybe not physically dominating the surrounding landscape, are explicitly positioned relative to a visual consumer – either the visitor in front of the building or more likely the viewer of a mediated image in press, television or film – and, a ‘successful’ building will necessarily develop a strong association to place through an instantly recognisable form designed to be both distinctive and widely disseminated in this mediated form’ (2011, p. 120). Iconic buildings are media-friendly, ‘Instagrammable’ and can ‘survive being shrunk to the size of a TV screen, or smaller, to a letterhead or stamp’ (2011, p. 121). This definition is somewhat problematic for our research, since it also applies to older, historical buildings such as the Eiffel Tower or the Leaning Tower of Pisa. Therefore, we must specify that when we refer to ‘iconic’ buildings in this contribution, we refer to a style that architects would also call ‘contemporary’, meaning it has been around for roughly twenty-five years. The use of the term ‘iconic’ in contemporary architecture often refers to large buildings, characterised by a use of ‘bare’ materials such as glass and concrete, a limited colour palette and a tendency for abstract, stylised shapes.
Aesthetics and class in sociological theory
Pierre Bourdieu is certainly the most influential author in sociology with regards to the link between social class and the interpretation and experience of culture and the aesthetic. Central to Bourdieu’s theories, is the idea that ‘culture (is) an arbitrarily constructed notion, (…) social groups (construct) their own particular contingent notions of ‘culture’ (Webster, 2011, p. 7). This in turn implies that cultural perception is not ‘the result of some inexplicable ‘sense’ or ‘feeling’ (but) in fact the result of the subconscious decoding and reasoning of the ‘learned’ habitus’ (Webster, 2011, p. 99). Bourdieu’s habitus is not the only, but certainly the most common way of conceptualising this connection. Bourdieu (1984) directed his attention to the connection between (interior) architecture and class (albeit in an almost anecdotal way) in Distinction. In a study he asked respondents to describe their ‘ideal home’ from a list of adjectives: ‘The results were correlated against the respondents’ job type and, perhaps not surprisingly, they demonstrated that the working-class respondents expressed preferences for functional interiors, the middle classes wanted ‘cosy’, ‘comfortable’ and ‘neat’ interiors, while the most-educated classes, teachers and professionals, preferred more aesthetic, ‘studied’ and ‘harmonious’ homes’ (Webster, 2011, p. 44).
Several researchers later applied the concept of habitus to studies regarding the experience of space and architecture. Most studies focus on the way employees experience the design and decoration of their office spaces. Kimberly Elsbach conducted a study regarding the role that office décor plays in identifying and marking intra-organisational roles and identities. She underlined the great importance of details, regarding the furniture in the office of one of the managers, one respondent even said: ‘The quality of the furniture, in particular, seemed to indicate a level of separateness’ (2004, p. 116). Wasserman and Frenkel (2015) also studied the relationship between spatial design, gender and social class by conducting a case study in the Israeli ministry of foreign affairs. Interviews with employees of different social backgrounds, managers and architects behind the building (that opened its doors in 2002) formed the basis of their analyses. Their main conclusion was that respondents experienced their (new) office space very differently, depending on their gender and the social class they belonged to: ‘Highly educated middle-class women feel comfortable in the new space and know how to ‘fit in’ and comply with the organisation’s image of the ideal worker, despite experiencing the space as masculine. Women from lower-class backgrounds, however, experience the space as more oppressive and marginalising, and lack the cultural capital to adapt themselves to the newly imposed organisational identity’ (2015, p. 1501). The authors stress the fact that both the architects and the decision makers wanted to convey the sense of a ‘new Israel’, open to Western influences and exhibiting a professional and international image: ‘While the architects’ and managers’ discursive work rarely refers directly to gender or class, it nonetheless lays the foundation for the rejection of markers of femininity, or indeed of anything that diverges from a Western, middle-class, rational aesthetics’ (2015, p. 1486).
Although Marx never really developed a theory regarding the aesthetic (and much less architecture), we could view this last quote as typically Marxist. The aesthetic is regarded as yet another arena for class struggle, or as Richard Shusterman put it: ‘some Marxists would claim that all traditional (including temporary) aesthetics represents an attempt to portray and justify certain socially conditioned and privileged determinations of taste as naturally or ontologically grounded values’ (1989, p. 98). More specifically regarding the experience of spaces, Wasserman and Frenkel also summarise the ideas of neo-Marxist Henri Lefebvre as followed: ‘space is never neutrally conceptualised; rather, it reflects the priorities of the dominant group, and it is affected by the social position of those in power to dictate specific elite tastes’ (2015, p. 1488). Furthermore, it should be noted that several authors suggest that the characteristics of a space and the meaning we ascribe to those spaces can facilitate different types of behavior. The phenomenological occurrence of entering a ‘special’ place and the behaviour that ensues because of it, receives a great deal of attention in the sociology of religion and in anthropology, more specifically regarding the relation between the sacred and the profane: ‘Sacred things are treated with particular respect and preserved from what is thought to defile them. They are treated in this way because they are believed to represent a normative order: (…) distinguishing social roles according to gender or age or marking out territories as proper places for particular groups’ (Dawes, 2017, p. 26). The regulating power of the sacred is emphasized by many authors and is often linked to spatial factors: ‘sacred spaces (…) may be taboo in the sense that a site is so sacred that only certain people can enter’ (Dallen & Olsen, 2006, p. 111).
We would like to argue that social class functions in a manner similar to the spiritual or sacred, with regard to assigning meaning to spaces and buildings. We believe that the uneasy feeling of not belonging somewhere can often be traced back to the sensation that we are venturing beyond our familiar social milieu. Some feel uneasy entering an ‘uptown’ art gallery, others when entering the folksy pub behind a football stadium. The facades of these different buildings contain symbolic signals that communicate a message: for some it will be an invitation, for others a sign to keep on walking. The role of the affective needs to be stressed, as no one consciously thinks, ‘this building and the events that take place in it, are incompatible with the habitus of my social class’. On the contrary, people prefer to ignore class differences, whatever side of the spectrum they are on: ‘recognition of status differences in cross-class interactions is likely to generate anxiety for the interactants’ (Gray & Kish-Gephart, 2013, p. 675). Just like with the spiritual, fear plays an important part in reinforcing social taboo.
Gray and Kish-Gephart see ‘class work’ as the most important coping-mechanism when cross-class interactions take place. They provide the following definition for ‘class work’: ‘interpretive processes and interaction rituals that organisational members individually and collectively take to manage cross-class encounters’ (2013, p. 671). Wasserman and Frenkel (2015) give a good example of this. In their research, they found that ninety per cent of all secretaries used ‘guerilla architecture’ to make the office space, which they deem cold and overly businesslike, more livable: photos and colourful art pieces on their desks play a mediating role. Le Corbusier’s brutalistic projects, most notably his ‘Unité d’Habitation’ in Marseille (constructed in 1952), invoked similar responses from the working class residents who would end up occupying these buildings. The sleek, modern design, completely erected in concrete, did not sit necessarily well with the inhabitants, spurring them to mediate the architects’ vision by adding more traditional elements, such as ornamented hardwood doors. The concept of ‘class work’ can of course be very valuable in researching the relationship between social class and architecture, not in the least because it emphasises the interpretative and creative role that individuals can assume in dealing with class differences. Still, it has to be noted that all research regarding this theme focuses on ‘cross class interactions’ that are almost unavoidable. Methodologically speaking, an office is the ideal setting to research ‘class-salient encounters’: managers and employees of different social backgrounds are forced to share a space. We would argue however, that the potentially most ‘class-salient’ encounters never happen in the first place. If it is true that aesthetics deter people from entering a building or space altogether, it becomes very difficult to conduct a research on the reasons why they do not enter. To put it more plainly: recruiting respondents because they do not do something, generates a lot of practical problems. Moreover, it is not an easy thing to operationalise an individual’s judgement of the aesthetic. Perhaps these are some of the reasons why there is so little attention for the effects of the aesthetic with regards to attendance of public buildings and cultural institutions.
Lastly, we would like to comment on our decision to focus so heavily on education level, as opposed to other indicators of social class, in our own work. This choice should be understood in two ways: firstly, we approach the topic from a distincly Bourdieausian perspective and his work on the attendance of cultural institutions, especially, tends to place great emphasis on the importance of education. Regarding how members of different social classes experience museums, he found that ‘in addition to visiting and its patterns, all visitors’ behavior (…) are directly and almost exclusively related to education’ (Bourdieu, 1991, p. 37). Of course, education level and social class are not interchangeable concepts (the connection between the two is too complex for that) but considering the nature of our data and the emphasis on cultural capital in our research, we do consider it the best indicator of social class at our disposal. We acknowledge that the choice of a different indicator might have led to different results, but education level is the most sound choice with regard to our theoretical assumptions. Secondly, as discussed in the introduction, when looking at the changing composition of the visiting public of public libraries, the most worrying trend seems to pertain to education level (more so than income, gender or age), underscoring the need for more information regarding this topic.
Methods
Study population and data collection methods
The data used for the analyses we present in section four was collected in a large scale user survey we conducted in 105 public libraries in Flanders and Brussels (Belgium) in 2018. The research was an initiative of the Flemish library and archive association (VVBAD) and the goal was to collect data on the state of the Flemish library sector and to provide all Dutch-speaking libraries who participated to the research with a personalised report that gives detailed information on how users perceive and use their library. A total of 45,228 active library users filled out a questionnaire, in which they answered more than 75 questions on how they use their library (for example frequency of visits or typical behaviour) and voiced their opinion regarding their library in a wide variety of subjective items. The distribution of age, gender and education level among the survey’s respondents are presented in the Appendix. These show that respondents tend to be older, female and higher educated. Although this is undoubtedly a reflection of the real-life attendance levels, we would also like to underline that we do not claim that the data we gathered is perfectly representative of the population. The use of a long, written questionnaire tends to favour the participation of higher educated respondents and we also know that women are often more readily persuaded to participate in these kinds of surveys. Furthermore, the participating public libraries insisted that we did not take samples of their user databases, instead allowing every visitor to participate to the research if they so desired. The reason for this was merited by the fact that public libraries also attract a sizable amount of visitors who are not necessarily registered members. While preparing the individual reports for libraries, a number of checks were made to determine whether the overrepresentation of certain groups caused skewed results. When weighing the data so that the distribution of respondents more closely reflects the actual attendance of different age groups and education levels, it was clear that this lead to only marginal changes in results. The weighting coefficients were calculated based on the distributions of respondents that reported to be library visitors in the 2014 Flemish Participation Survey.
Data analysis
The first analysis discussed in section four correlates the level of perceived invitingness (respondents could specify to what degree they agreed with the statement ‘the library is an inviting place’ on a five point Likert-scale) with all items on satisfaction (respondents could say how satisfied they were with a specific aspect of their library, again in a five point Likert-scale). These items cover a lot of ground, from the satisfaction with staff to the quality of infrastructure. This analysis was carried out on respondent (and not library) level. We also conducted a regression analysis with these items. Since many were conceptually very similar (for example ‘Friendliness of staff’ and ‘Helpfulness of staff’), we decided to perform a factor analysis (PCA) first to see if any latent underlying variables could be identified. This approach strongly resembles analyses that were carried out on data collected in another large scale user survey in Flemish public libraries in 2004, although there the percentage of male and lower educated visitors in every library served as dependent variables instead of perceived invitingness (Glorieux et al., 2007). PCA suggested six different components and we identified three items that did not load sufficiently on any of these six:
Satisfaction with opening hours;
Satisfaction with the membership fee;
Satisfaction with fines arrangement.
These were removed from the model and the PCA was run again, resulting again in six components. Our next step was to ascertain whether the suggested components made sense conceptually. We concluded that one did not, PCA suggested that the following four items belonged together:
Satisfaction with copy and printing services in library;
Satisfaction with cultural activities;
Satisfaction with library newsletter;
Satisfaction with social media presence of library.
After removing the items regarding the copy and printing services and the cultural activities because they were conceptually too distinct, PCA suggested a model with five components that was not only statistically sound, but also made sense conceptually. Based on the Kaiser-Meyer-Olkin measure of sampling adequacy (0.908), Bartlett’s test of sphericity (0.000) and a total explained variance of 73.8 per cent, we concluded that this model was valid and usable. The following latent variables were constructed (a detailed overview of which items contributed to which latent variable is given in Table e, in the Appendix):
Satisfaction with building and interior;
Satisfaction with online presence of library;
Satisfaction wth reachability of library;
Satisfaction with system for checking out materials;
Satisfaction with staff.
The data gathered in the user survey was also used to conduct a series of exploratory analyses to identify libraries that were perceived differently by users with varying education levels with regard to the attractiveness of their building and/or their perceived invitingness. This of course means these analyses were conducted on the level of individual libraries. We looked again at the statement ‘the library is an inviting place’ and the item on satisfaction regarding the ‘appeal of the building’, this time linking them to respondents’ education levels. In order to increase the robustness of this indicator, we only selected respondents whose education level is identical to that of their highest educated parent (N = 14,361). The primary way in which we sought to identify libraries that are perceived differently according to education level, was through ANOVA testing, treating education level as the independent variable and the perceived invitingness and attractiveness of the library as dependent variables. Additionally, we also used Spearman’s Rho to find libraries where the attractiveness or invitingness of a library correlates significantly with the education level of its users. All relevant output of these analyses can be found in the Appendix (Tables f–i). While the methods used here might seem somewhat unusual or rudimentary, they nonetheless served the goal of this exploratory analysis very well by being readily interpretable, which allowed us to analyse more than a hundred libraries uniformly and simultaneously.
Discussion
The importance of architecture with regard to perceived invitingness
Starting from the observation that the degree to which a library is perceived to be ‘inviting’ by its users varied significantly between libraries, we set out to ascertain what determines whether a library is inviting or not. The results in table one clearly point to the importance of the building and interior design with regard to the perceived invitingness of libraries. In fact, although all correlations are statistically significant, only items regarding the building and interior design show relevant correlations. It is especially striking to see how seemingly irrelevant the friendliness or availability of the staff is. One would think that these play a major role in creating a welcoming and hospitable environment, perhaps even more so than the building, but the data suggests otherwise (Table 1).
Correlations (Spearman’s Rho) between agreement with statement ‘The library is an inviting place’ and all items in questionnaire regarding satisfaction (sorted from strongest correlation to weakest correlation).
Satisfaction with building of library | 472** |
Satisfaction with interior design of library | 442** |
Satisfaction with reading room | 411** |
Satisfaction with signage in library | 375** |
Satisfaction with how library collection is organised | 347** |
Satisfaction with accessibility of library | 304** |
Satisfaction with cultural activities | 298** |
Satisfaction with reachability of library | 293** |
Satisfaction with library website | 265** |
Satisfaction with library newsletter | 263** |
Satisfaction with social media presence of library | 260** |
Satisfaction with competence of staff | 258** |
Satisfaction with availability of staff | 252** |
Satisfaction with online catalogue of library | 250** |
Satisfaction with how fast staff can provide help | 248** |
Satisfaction with copy and printing services in library | 244** |
Satisfaction with opening hours | 244** |
Satisfaction with parking availability | 241** |
Satisfaction with payment options in library | 238** |
Satisfaction with helpfulness of staff | 237** |
Satisfaction with friendliness of staff | 235** |
Satisfaction with automatised loan-out infrastructure | 234** |
Satisfaction with bicycle accommodations | 228** |
Satisfaction with membership fee | 210** |
Satisfaction with fines arrangement | 205** |
Satisfaction with online service ‘My library’ | 202** |
Satisfaction with building of library | 472** |
Satisfaction with interior design of library | 442** |
Satisfaction with reading room | 411** |
Satisfaction with signage in library | 375** |
Satisfaction with how library collection is organised | 347** |
Satisfaction with accessibility of library | 304** |
Satisfaction with cultural activities | 298** |
Satisfaction with reachability of library | 293** |
Satisfaction with library website | 265** |
Satisfaction with library newsletter | 263** |
Satisfaction with social media presence of library | 260** |
Satisfaction with competence of staff | 258** |
Satisfaction with availability of staff | 252** |
Satisfaction with online catalogue of library | 250** |
Satisfaction with how fast staff can provide help | 248** |
Satisfaction with copy and printing services in library | 244** |
Satisfaction with opening hours | 244** |
Satisfaction with parking availability | 241** |
Satisfaction with payment options in library | 238** |
Satisfaction with helpfulness of staff | 237** |
Satisfaction with friendliness of staff | 235** |
Satisfaction with automatised loan-out infrastructure | 234** |
Satisfaction with bicycle accommodations | 228** |
Satisfaction with membership fee | 210** |
Satisfaction with fines arrangement | 205** |
Satisfaction with online service ‘My library’ | 202** |
Next we conducted a regression analysis, using both the newly constructed variables found through PCA as well as all items that were not included in the final model as independent variables, and the perceived ‘invitingness’ of the library as the dependent variable. The results can be seen in table two and confirm what we saw in table one. All constructed variables are significant, together with most singular items (only the item regarding the copy and print services was insignificant and was therefore omitted from Table 2). When looking at the effect-sizes, both standardised and unstandardised, it becomes very clear that the satisfaction with the building and interior design of a library clearly has the most impact on whether a library will be experienced as being inviting. In preparation of these analyses, we of course suspected that the building or interior design of libraries might be relevant, but even we were very surprised to find how strongly they determine the level of perceived ‘invitingness’. It should perhaps be noted here that the satisfaction with the building and interior design should not be understood as pertaining exclusively to the aesthetic, as an item on how the library collection is organised also loads on this latent variable. This makes sense, as function and design intertwine heavily in architecture, or as Frank Lloyd Wright famously stated: ‘form and function are one’ (Cruz, 2012). The ‘reading room’ (the satisfaction with which also loads on the latent variable) of a library is not something that respondents only ‘look at’, they likely also evaluate how ‘comfortable’ the seating arrangements are for example.
Regression analysis of agreement respondents with statement ‘The library is an inviting place’.
. | Regression coefficient . | Standard error . | Beta . | t-value . | p-value . |
---|---|---|---|---|---|
Intercept | 3,299 | 0,080 | 41,250 | 0,000 | |
Satisfaction with building and interior design (factor) | 0,328 | 0,009 | 0,378 | 36,076 | 0,000 |
Satisfaction with online presence of library (factor) | 0,139 | 0,010 | 0,160 | 13,726 | 0,000 |
Satisfaction with reachability of library (factor) | 0,120 | 0,009 | 0,138 | 13,871 | 0,000 |
Satisfaction with system for checking out materials (factor) | 0,113 | 0,009 | 0,131 | 12,500 | 0,000 |
Satisfaction with staff (item) | 0,100 | 0,009 | 0,115 | 11,060 | 0,000 |
Satisfaction with cultural activities (item) | 0,074 | 0,015 | 0,066 | 4,822 | 0,000 |
Satisfaction with membership fee (item) | 0,056 | 0,013 | 0,048 | 4,151 | 0,000 |
Satisfaction with opening hours (item) | 0,034 | 0,011 | 0,037 | 3,226 | 0,001 |
Satisfaction with system of fines (item) | −0,037 | 0,012 | −0,037 | −3,067 | 0,002 |
. | Regression coefficient . | Standard error . | Beta . | t-value . | p-value . |
---|---|---|---|---|---|
Intercept | 3,299 | 0,080 | 41,250 | 0,000 | |
Satisfaction with building and interior design (factor) | 0,328 | 0,009 | 0,378 | 36,076 | 0,000 |
Satisfaction with online presence of library (factor) | 0,139 | 0,010 | 0,160 | 13,726 | 0,000 |
Satisfaction with reachability of library (factor) | 0,120 | 0,009 | 0,138 | 13,871 | 0,000 |
Satisfaction with system for checking out materials (factor) | 0,113 | 0,009 | 0,131 | 12,500 | 0,000 |
Satisfaction with staff (item) | 0,100 | 0,009 | 0,115 | 11,060 | 0,000 |
Satisfaction with cultural activities (item) | 0,074 | 0,015 | 0,066 | 4,822 | 0,000 |
Satisfaction with membership fee (item) | 0,056 | 0,013 | 0,048 | 4,151 | 0,000 |
Satisfaction with opening hours (item) | 0,034 | 0,011 | 0,037 | 3,226 | 0,001 |
Satisfaction with system of fines (item) | −0,037 | 0,012 | −0,037 | −3,067 | 0,002 |
R2 (explained variance) = 0.26 (N = 45,887).
Exploratory analyses
In our literature review presented above, we hope to have adequately shown that social class membership can considerably determine and shape aesthetic preferences, including (but not limited to) what type of buildings and interiors one feels comfortable in. This leads to the question which specific aesthetic elements can discourage or encourage members of certain social classes to enter/visit a space or building, one we will not conclusively answer in this contribution. Exploratory analysis of the user survey data, however, did help us identify a total of ten libraries where the invitingness and attractiveness of the library is experienced differently by users with different levels of education. Though the size of all found correlations can be interpreted as relatively low, we still think that these merit further research. The primary goal here, after all, was to find cases (public libraries) that can be studied more closely, but also look for a first confirmation of our assumption that certain styles can discourage the attendance of certain types of visitors and that iconic architecure likely appeals more to the higher educated. All libraries have been anonymised, the reason for this being that the degree to which a specific public library is inviting for users with lower levels of education, is something of a controversial matter.
A closer look at the results of the analyses shows us that only three out of these ten libraries were consistently evaluated better on their invitingness and/or attractiveness by respondents with lower educational levels, which seems to confirm that public libraries in general tend to be more tailored to the preferences and tastes of higher educated users. As mentioned before, determining which aesthetic styles repel or attract different social classes is beyond the scope of this contribution. However, while admittedly having no analytic value whatsoever, a subjective assessment of photos depicting the different libraries that were identified can, at the very least, indicate whether libraries that were evaluated better by those with higher levels of education tend to be (more) ‘iconic’. In that sense, it is interesting to note that out of the seven libraries that were judged significantly better by respondents with higher education levels, we considered four as iconic, or at least having iconic characteristics: the buildings themselves, viewed from outside, are not only larger than most libraries, but are characterised by more abstract designs with smooth lines and (often very large) glass walls. The interiors can be described as large, open spaces with high ceilings. The usage of colour is also similar between these buildings, with a clear preference for muted colours. Moreover, two of the four libraries we considered to be iconic were found to be constructed in the last decade and were designed by prestigious architectural firms. The three libraries that were evaluated significantly better by respondents with lower education levels, in turn, seem to adhere more to the classic image of a small-town public library: the buildings are considerably smaller and inconspicuous, the interiors decidedly more homey and less imposing, especially ‘library C’, which features an interior that appears to have remained largely unchanged for decades. It is difficult not to be reminded here of Bourdieu’s findings regarding class-based differences in descriptions of the ‘ideal home’, as discussed in our theory section.
Three libraries were more closely examined since they have a small sample size: libraries A (52 respondents), C (49 respondents) and especially D (32 respondents). These three had the largest effect sizes of all libraries found through ANOVA, which is cause for concern since Eta-squared is influenced by sample size. As a control measure, Omega-squared (which factors in sample size and is therefore more unbiased) was also calculated for these three libraries: effect sizes decreased, but remained high (0,191 for library A; 0,121 for library C and 0,179 for library D). The smaller sample sizes can, at least in part, be explained by the fact that these are simply smaller libraries. As a measure of control, we decided to check whether what was suggested through ANOVA (namely that libraries A and C are more appealing, and library D less appealing, to lower educated visitors) by comparing how well certain educational levels were represented in the sample of each library (and other libraries in that specific cluster). This showed that lower educated visitors are far better represented in libraries A and C than in library D. For instance: respondents who only completed lower secondary education represented 12.0 per cent of the sample in library C and 5.3 per cent in library A. Library D on the other hand, had zero respondents in this category. When looking at respondents who finished higher education (college or university), they represented 75.3 per cent of the sample in library D, 60.0 per cent in library A and merely 54.9 per cent in library C. These additional checks seem to confirm that the suggested connection between educational level and perceived invitingness/attractiveness is, in fact, present in these libraries.
While identifying libraries that are experienced differently according to education level is interesting, we are also confronted with the fact that there are ninety-five libraries in our dataset that do not exhibit significant differences. One could argue that this means that there is no connection, but it is important to stress the exploratory nature of the analyses and the fact that the variables were used in a way for which they were not originally designed. We also suspect that more libraries would have been able to be identified if respondents with lower educational levels would have been better represented in the dataset. Conversely, while it is true that ANOVA and Spearman’s Rho require a sufficient amount of respondents in each category to determine whether a found effect is statistically significant, they are immune to how well certain categories are represented in a sample once those thresholds have been reached. In short: we might have been able to identify more relevant libraries if lower educated visitors were better represented in our sample, but the ones we found are undoubtedly examples of libraries that are evaluated significantly different by lower and higher educated visitors. In any case, the fact that we are dealing with data gathered in a user survey, serves as a considerable limitation with regard to our analyses aimed at determining which libraries are uninviting for lower educated visitors. A last important point to make is that, while ‘iconic’ architecture is on the rise for public libraries, its ‘price tag’ prohibits most small towns from constructing their library in this style and the vast majority of public libraries in our research are not located in urban areas: only eleven libraries in our research are located in what can technically be classified as a ‘city’, of which only two can be considered to be ‘major cities’. Of these eleven libraries, however, three were found to be both more inviting for higher educated respondents, as well as being housed in a building that has iconic characteristics. This puts our findings in a new light by illustrating that, when focusing on libraries that are more likely to have an iconic building, we see that quite a considerable amount have opted for this style and not without consequence for their appeal to lower educated visitors.
Limitations
There are, of course, certain aspects of our approach to this subject one might criticise. An important one being that perhaps we are putting too much emphasis on education level or social class as determinants for whether or not an aesthetic style can encourage or discourage public library attendance. We can only answer that any conflation of identity that serves to predict behavior will reduce complexity and the effects of social class membership will undoubtedly be mediated by individual idiosyncrasies. It is a certainty that not every person with a lower level of education has a deep-rooted aversion for iconic architecture, the question is rather whether it is more prevalent among the lower educated and whether this has can have an impact on library attendance. We think our approach is greatly preferred to a conceptualisation of visitors of public libraries that does not take education (or class membership) into account. It should also be noted that a satisfactory answer to the question of lower library attendance levels of the workingclass has never been given and we believe this is, at least in part, by an overfocus on the practical, rational and economic dimensions of public library attendance. Le Roux, Rouanet, Savage, and Warde (2008) also argue that income or professional status do not suffice in explaining why certain cultural institutions are more ‘high-brow’ than others. We argue that there is a cultural dimension that is being overlooked and perhaps the building that houses a public library is the first and most important cultural message in communicating a sense of openness and belonging. The possibility that exclusion can start with a mere glance at a building, as we have tried to argue here, is worth studying and understanding.
Another limitation of our research pertains to the fact that we did not take the specific surroundings in which public libraries are located (and how this might present an additional exclusionary element for visitors with a working class background) into consideration. Iconic architecture is surely linked to gentrification and while we made the point that ‘iconic’ libraries are predominantly found in (larger) cities, our data does not allow for a more detailed investigation into this subject. While we are certain that the aesthetic characteristics of iconic architecture play a role in and of itself (so outside of the dynamics of urban planning or the processes of gentrification) and can be studied as such, we acknowledge that this aspect perhaps should be given greater consideration than we have been able to do here. Furthermore, while it is suggested here that iconic architecture appeals less to the lower educated, we cannot say so conclusively yet based on this data. The main reason for this is because our sample is composed of mainly higher educated respondents, which reflects the class-based nature of the Flemish library system, as discussed in the introduction.
Therefore, two things are very important for us going forward in this line of research. First of all, we want to be able to ‘open up’ our research to a wider audience, not just library visitors. As noted before, the potentially most class-salient encounters might never happen in the first place. In all likelihood, the lower educated or working class visitors of public libraries also represent a segment of those groups who are already more inclined to have positive attitudes towards cultural institutions and middle class attitudes, values and tastes. Therefore, respondents need to be more widely recruited in future research projects. Secondly, together with a more detailed evaluation of the spaces, we want respondents to be able to assess multiple environments and compare them. How respondents score and describe the different environments, should of course be linked to socio-demographic variables. This will allow us to analyse class-based dispositions towards aesthetics and architecture in a very detailed manner, which is important, since it could be argued that ‘there has not been any sustained attention to the aesthetic elements of Bourdieu’s thinking’ (Hanquinet, Roose, & Savage, 2014, p. 113).
Conclusion
We started this contribution by positing that public libraries are not living up to the emancipatory task that society has bestowed on them, illustrated by low (and decreasing) attendance among the lower educated. Next, through an overview of relevant literature, we argued that it is cause for concern that an iconic brand of architecture is gaining popularity for public buildings, including public libraries, considering that this style of architecture is often associated with very highbrow cultural institutions and therefore might discourage the attendance of lower educated visitors even more. By analysing the data gathered in our large scale user survey, we showed that the architectural style and interior design of a library play a prominent role in its perceived invitingness, more so than any other factor, including the friendliness of the staff or more practical aspects of its infrastructure. Finally, through a series of exploratory analyses, we argue that iconic architecture is likely experienced and evaluated differently according to the education level of visitors and that this might have a negative effect on the library attendance of lower educated visitors. We consider our work here as partaking into an inquiry on the connection between social class membership and appreciation or preference for certain architectural styles and aesthetics. We believe our findings suggest at it, but further research regarding this complex topic is certainly needed.
In ‘Palaces for the People’ (Klinenberg, 2018), Eric Klinenberg argues that we should (re)value public libraries (and invest more in them) because they can help ‘people elevate themselves and improve their situation’ (p. 37). While Klinenberg acknowledges the importance of the ‘look and feel’ of public spaces and institutions, he also states that ‘attractive places (…) are only one element of good social infrastructure; programming, which can be inclusive or exclusive, welcoming or forbidding, is also important’ (p. 155). We agree that the services provided by public libraries (and other institutions) are crucial and deserve attention, but we also think we provide an important additional perspective on the matter by questioning the (often made) assumption that ‘attractiveness’ means the same thing for every type of visitor. If we want to build true ‘palaces for the people’, we need to make sure that everyone feels comfortable in entering them.
The fact that architecture and aesthetics are often experienced and evaluated very differently depending on education level and social class membership, has been underlined and illustrated to a great extent here. This could lead to the assumption that failure is inevitable when it comes to constructing public buildings that should appeal to all layers of society, even if architects and decision-makers should decide to give this aspect of public buildings the importance we believe it deserves. We are more optimistic for two reasons. First of all: not every library that was included in our dataset and had recently opted for a more contemporary architectural style was evaluated significantly different by lower and higher educated respondents. This suggest that modern (maybe even slightly iconic) library-buildings are not necessarily evaluated poorly by members of the working class or visitors with lower incomes or education levels.
Charles Jencks (2006), perhaps the strongest advocate for iconic buildings in the field of architectural theory, not only claimed that they can play a significant symbolic role in our relativistic and postmodern society, he also argued for ‘a code of good practice’ (p. 3) when designing and building iconic architecture, admitting that many iconic buildings miss their mark by not resonating with the general public. He argued that there should be ‘more thought on the iconography behind the buildings’ and ‘more coherence in the use of metaphors’ (p. 16). As social scientists, we are interested in what would constitute this ‘code of good practice’ when taking social stratification, which we put forward as being important with regard to architecture in this contribution, into consideration. What sets ‘good’ iconic buildings apart? Why do some of these buildings (perhaps inadvertently) manage to appeal to a wider range of visitors? Our objective, understanding more about the preferences or interpretation-schemes of members of different social classes, can hopefully contribute to a more inclusive architecture for public buildings.
Disclosure statement
No potential conflict of interest was reported by the author(s).
References
Distribution of age in dataset.
. | Frequency . | Per cent . | Cumulative Per cent . |
---|---|---|---|
18–29 years old | 3702 | 8.2 | 8.2 |
30–44 years old | 9551 | 21.1 | 29.3 |
45–59 years old | 13,963 | 30.9 | 60.2 |
60 years or older | 18,012 | 39.8 | 100.0 |
Total | 45,228 | 100.0 |
. | Frequency . | Per cent . | Cumulative Per cent . |
---|---|---|---|
18–29 years old | 3702 | 8.2 | 8.2 |
30–44 years old | 9551 | 21.1 | 29.3 |
45–59 years old | 13,963 | 30.9 | 60.2 |
60 years or older | 18,012 | 39.8 | 100.0 |
Total | 45,228 | 100.0 |
Distribution of gender in dataset.
. | Frequency . | Per cent . | Valid Per cent . |
---|---|---|---|
Male | 15,405 | 34.1 | 34.4 |
Female | 29,206 | 64.6 | 65.3 |
No answer | 109 | 0.2 | 0.2 |
Total | 44,720 | 98.9 | 100.0 |
. | Frequency . | Per cent . | Valid Per cent . |
---|---|---|---|
Male | 15,405 | 34.1 | 34.4 |
Female | 29,206 | 64.6 | 65.3 |
No answer | 109 | 0.2 | 0.2 |
Total | 44,720 | 98.9 | 100.0 |
Distribution of education levels in dataset.
. | Frequency . | Per cent . | Cumulative Per cent . |
---|---|---|---|
None or lower education | 715 | 1.7 | 1.7 |
Lower secondary education | 2177 | 5.1 | 6.7 |
Higher secondary education | 10,536 | 24.5 | 31.3 |
Higher education | 28,792 | 67.1 | 98.3 |
Other | 709 | 1.7 | 100.0 |
Total | 42,929 | 100.0 |
. | Frequency . | Per cent . | Cumulative Per cent . |
---|---|---|---|
None or lower education | 715 | 1.7 | 1.7 |
Lower secondary education | 2177 | 5.1 | 6.7 |
Higher secondary education | 10,536 | 24.5 | 31.3 |
Higher education | 28,792 | 67.1 | 98.3 |
Other | 709 | 1.7 | 100.0 |
Total | 42,929 | 100.0 |
Principal component analysis, total variance explained.
Component . | Initial Eigenvalues . | Extraction Sums of Squared Loadings . | ||||
---|---|---|---|---|---|---|
. | Total . | Per cent of Variance . | Cumulative Per cent . | Total . | Per cent of Variance . | Cumulative Per cent . |
Satisfaction with online presence of library | 7247 | 42.6 | 42.6 | 7247 | 42.6 | 42.6 |
Satisfaction with building and interior | 1735 | 10.2 | 52.8 | 1735 | 10.2 | 52.8 |
Satisfaction with staff | 1406 | 8.3 | 61.1 | 1406 | 8.3 | 61.1 |
Satisfaction with reachability of library | 1108 | 6.5 | 67.6 | 1108 | 6.5 | 67.6 |
Satisfaction with system for checking out materials | 1044 | 6.1 | 73.8 | 1044 | 6.1 | 73.8 |
Component . | Initial Eigenvalues . | Extraction Sums of Squared Loadings . | ||||
---|---|---|---|---|---|---|
. | Total . | Per cent of Variance . | Cumulative Per cent . | Total . | Per cent of Variance . | Cumulative Per cent . |
Satisfaction with online presence of library | 7247 | 42.6 | 42.6 | 7247 | 42.6 | 42.6 |
Satisfaction with building and interior | 1735 | 10.2 | 52.8 | 1735 | 10.2 | 52.8 |
Satisfaction with staff | 1406 | 8.3 | 61.1 | 1406 | 8.3 | 61.1 |
Satisfaction with reachability of library | 1108 | 6.5 | 67.6 | 1108 | 6.5 | 67.6 |
Satisfaction with system for checking out materials | 1044 | 6.1 | 73.8 | 1044 | 6.1 | 73.8 |
Principal component analysis, rotated factor matrix.
Items . | Factor . | Component . | ||||
---|---|---|---|---|---|---|
. | 1 . | 2 . | 3 . | 4 . | 5 . | . |
Satisfaction with library newsletter | 0.771 | 0.228 | 0.174 | 0.064 | 0.047 | Satisfaction with online presence of library |
Satisfaction with social media presence of library | 0.769 | 0.238 | 0.130 | 0.074 | 0.004 | |
Satisfaction with online catalogue of library | 0.755 | 0.116 | 0.133 | 0.153 | 0.236 | |
Satisfaction with library website | 0.746 | 0.208 | 0.173 | 0.145 | 0.158 | |
Satisfaction with online service ‘My library’ | 0.693 | 0.050 | 0.133 | 0.140 | 0.294 | |
Satisfaction with interior design of library | 0.185 | 0.849 | 0.187 | 0.215 | 0.141 | Satisfaction with building and interior |
Satisfaction with building of library | 0.149 | 0.835 | 0.113 | 0.221 | 0.141 | |
Satisfaction with reading room | 0.221 | 0.800 | 0.159 | 0.192 | 0.114 | |
Satisfaction with how library collection is organised | 0.290 | 0.654 | 0.258 | 0.175 | 0.117 | |
Satisfaction with friendliness of staff | 0.173 | 0.180 | 0.863 | 0.146 | 0.123 | Satisfaction with staff |
Satisfaction with competence of staff | 0.218 | 0.214 | 0.840 | 0.156 | 0.114 | |
Satisfaction with how fast staff can provide help | 0.216 | 0.180 | 0.833 | 0.206 | 0.122 | |
Satisfaction with parking availability | 0.120 | 0.187 | 0.116 | 0.840 | 0.050 | Satisfaction with reachability of library |
Satisfaction with bicycle accommodations | 0.164 | 0.200 | 0.147 | 0.764 | 0.058 | |
Satisfaction with reachability of library | 0.134 | 0.281 | 0.240 | 0.709 | 0.192 | |
Satisfaction with automatised loan-out infrastructure | 0.217 | 0.166 | 0.149 | 0.127 | 0.811 | Satisfaction with system for checking out materials |
Satisfaction with payment options in library | 0.218 | 0.179 | 0.129 | 0.087 | 0.806 |
Items . | Factor . | Component . | ||||
---|---|---|---|---|---|---|
. | 1 . | 2 . | 3 . | 4 . | 5 . | . |
Satisfaction with library newsletter | 0.771 | 0.228 | 0.174 | 0.064 | 0.047 | Satisfaction with online presence of library |
Satisfaction with social media presence of library | 0.769 | 0.238 | 0.130 | 0.074 | 0.004 | |
Satisfaction with online catalogue of library | 0.755 | 0.116 | 0.133 | 0.153 | 0.236 | |
Satisfaction with library website | 0.746 | 0.208 | 0.173 | 0.145 | 0.158 | |
Satisfaction with online service ‘My library’ | 0.693 | 0.050 | 0.133 | 0.140 | 0.294 | |
Satisfaction with interior design of library | 0.185 | 0.849 | 0.187 | 0.215 | 0.141 | Satisfaction with building and interior |
Satisfaction with building of library | 0.149 | 0.835 | 0.113 | 0.221 | 0.141 | |
Satisfaction with reading room | 0.221 | 0.800 | 0.159 | 0.192 | 0.114 | |
Satisfaction with how library collection is organised | 0.290 | 0.654 | 0.258 | 0.175 | 0.117 | |
Satisfaction with friendliness of staff | 0.173 | 0.180 | 0.863 | 0.146 | 0.123 | Satisfaction with staff |
Satisfaction with competence of staff | 0.218 | 0.214 | 0.840 | 0.156 | 0.114 | |
Satisfaction with how fast staff can provide help | 0.216 | 0.180 | 0.833 | 0.206 | 0.122 | |
Satisfaction with parking availability | 0.120 | 0.187 | 0.116 | 0.840 | 0.050 | Satisfaction with reachability of library |
Satisfaction with bicycle accommodations | 0.164 | 0.200 | 0.147 | 0.764 | 0.058 | |
Satisfaction with reachability of library | 0.134 | 0.281 | 0.240 | 0.709 | 0.192 | |
Satisfaction with automatised loan-out infrastructure | 0.217 | 0.166 | 0.149 | 0.127 | 0.811 | Satisfaction with system for checking out materials |
Satisfaction with payment options in library | 0.218 | 0.179 | 0.129 | 0.087 | 0.806 |
Notes: Rotation method; Varimax with Kaiser Normalisation.
Educational level crossed with level of agreement with statement ‘the library is an inviting place’.
. | . | Totally don't agree . | Don't agree . | Neutral . | Agree . | Totally Agree . |
---|---|---|---|---|---|---|
Library E* (N = 192) η2 = 0.064 | None or elementary school | 11.1% | 11.1% | 22.2% | 33.3% | 22.2% |
Lower secondary education | 0.0% | 12.5% | 0.0% | 50.0% | 37.5% | |
Higher secondary education | 1.9% | 3.7% | 11.1% | 57.4% | 25.9% | |
Higher education (college) | 0.0% | 1.2% | 7.0% | 60.5% | 31.4% | |
Higher education (university) | 0.0% | 0.0% | 8.6% | 48.6% | 42.9% | |
Library F** (N = 978) η2 = 0.017 | None or elementary school | 9.5% | 14.3% | 23.8% | 42.9% | 9.5% |
Lower secondary education | 3.4% | 3.4% | 31.0% | 58.6% | 3.4% | |
Higher secondary education | 1.0% | 3.6% | 19.3% | 55.3% | 20.8% | |
Higher education (college) | 4.1% | 6.1% | 20.7% | 45.8% | 23.3% | |
Higher education (university) | 0.8% | 7.5% | 16.8% | 47.4% | 27.6% | |
Library G** (N = 257) η2 = 0.055 | None or elementary school | 15.8% | 15.8% | 36.8% | 26.3% | 5.3% |
Lower secondary education | 0.0% | 6.7% | 26.7% | 46.7% | 20.0% | |
Higher secondary education | 0.0% | 5.3% | 38.9% | 52.6% | 3.2% | |
Higher education (college) | 3.7% | 14.6% | 30.5% | 41.5% | 9.8% | |
Higher education (university) | 2.2% | 8.7% | 26.1% | 43.5% | 19.6% | |
Library A** (N = 52) η2 = 0.258 | None or elementary school | 0.0% | 0.0% | 0.0% | 33.3% | 66.7% |
Lower secondary education | 0.0% | 0.0% | 25.0% | 75.0% | 0.0% | |
Higher secondary education | 0.0% | 0.0% | 22.2% | 61.1% | 16.7% | |
Higher education (college) | 0.0% | 4.5% | 50.0% | 40.9% | 4.5% | |
Higher education (university) | 0.0% | 0.0% | 80.0% | 20.0% | 0.0% | |
Library H* (N = 262) η2 = 0.042 | None or elementary school | 8.3% | 8.3% | 41.7% | 33.3% | 8.3% |
Lower secondary education | 0.0% | 0.0% | 23.1% | 46.2% | 30.8% | |
Higher secondary education | 2.5% | 4.9% | 22.2% | 43.2% | 27.2% | |
Higher education (college) | 2.0% | 1.0% | 11.0% | 61.0% | 25.0% | |
Higher education (university) | 0.0% | 3.6% | 17.9% | 51.8% | 26.8% | |
Library I** (N = 97) η2 = 0.148 | None or elementary school | 14.3% | 28.6% | 28.6% | 28.6% | 0.0% |
Lower secondary education | 0.0% | 0.0% | 30.0% | 70.0% | 0.0% | |
Higher secondary education | 0.0% | 6.9% | 13.8% | 62.1% | 17.2% | |
Higher education (college) | 0.0% | 2.2% | 24.4% | 57.8% | 15.6% | |
Higher education (university) | 0.0% | 0.0% | 33.3% | 50.0% | 16.7% |
. | . | Totally don't agree . | Don't agree . | Neutral . | Agree . | Totally Agree . |
---|---|---|---|---|---|---|
Library E* (N = 192) η2 = 0.064 | None or elementary school | 11.1% | 11.1% | 22.2% | 33.3% | 22.2% |
Lower secondary education | 0.0% | 12.5% | 0.0% | 50.0% | 37.5% | |
Higher secondary education | 1.9% | 3.7% | 11.1% | 57.4% | 25.9% | |
Higher education (college) | 0.0% | 1.2% | 7.0% | 60.5% | 31.4% | |
Higher education (university) | 0.0% | 0.0% | 8.6% | 48.6% | 42.9% | |
Library F** (N = 978) η2 = 0.017 | None or elementary school | 9.5% | 14.3% | 23.8% | 42.9% | 9.5% |
Lower secondary education | 3.4% | 3.4% | 31.0% | 58.6% | 3.4% | |
Higher secondary education | 1.0% | 3.6% | 19.3% | 55.3% | 20.8% | |
Higher education (college) | 4.1% | 6.1% | 20.7% | 45.8% | 23.3% | |
Higher education (university) | 0.8% | 7.5% | 16.8% | 47.4% | 27.6% | |
Library G** (N = 257) η2 = 0.055 | None or elementary school | 15.8% | 15.8% | 36.8% | 26.3% | 5.3% |
Lower secondary education | 0.0% | 6.7% | 26.7% | 46.7% | 20.0% | |
Higher secondary education | 0.0% | 5.3% | 38.9% | 52.6% | 3.2% | |
Higher education (college) | 3.7% | 14.6% | 30.5% | 41.5% | 9.8% | |
Higher education (university) | 2.2% | 8.7% | 26.1% | 43.5% | 19.6% | |
Library A** (N = 52) η2 = 0.258 | None or elementary school | 0.0% | 0.0% | 0.0% | 33.3% | 66.7% |
Lower secondary education | 0.0% | 0.0% | 25.0% | 75.0% | 0.0% | |
Higher secondary education | 0.0% | 0.0% | 22.2% | 61.1% | 16.7% | |
Higher education (college) | 0.0% | 4.5% | 50.0% | 40.9% | 4.5% | |
Higher education (university) | 0.0% | 0.0% | 80.0% | 20.0% | 0.0% | |
Library H* (N = 262) η2 = 0.042 | None or elementary school | 8.3% | 8.3% | 41.7% | 33.3% | 8.3% |
Lower secondary education | 0.0% | 0.0% | 23.1% | 46.2% | 30.8% | |
Higher secondary education | 2.5% | 4.9% | 22.2% | 43.2% | 27.2% | |
Higher education (college) | 2.0% | 1.0% | 11.0% | 61.0% | 25.0% | |
Higher education (university) | 0.0% | 3.6% | 17.9% | 51.8% | 26.8% | |
Library I** (N = 97) η2 = 0.148 | None or elementary school | 14.3% | 28.6% | 28.6% | 28.6% | 0.0% |
Lower secondary education | 0.0% | 0.0% | 30.0% | 70.0% | 0.0% | |
Higher secondary education | 0.0% | 6.9% | 13.8% | 62.1% | 17.2% | |
Higher education (college) | 0.0% | 2.2% | 24.4% | 57.8% | 15.6% | |
Higher education (university) | 0.0% | 0.0% | 33.3% | 50.0% | 16.7% |
Notes. We provide cross tables because they give a more tangible overview of differences between categories than the actual output created by ANOVA. Furthermore, we only included libraries that also yielded significant differences between lower and higher educational levels when looking at the Multiple Comparisons table.
Educational level crossed with evaluation of attractiveness of building.
. | . | Not at all satisfied . | Not satisfied . | Neither satisfied, nor dissatisfied . | Satisfied . | Very satisfied . |
---|---|---|---|---|---|---|
Library E** (N = 195) η2 = 0.068 | None or elementary school | 0.0% | 0.0% | 0.0% | 50.0% | 50.0% |
Lower secondary education | 0.0% | 0.0% | 0.0% | 33.3% | 66.7% | |
Higher secondary education | 0.0% | 1.8% | 7.3% | 40.0% | 50.9% | |
Higher education (college) | 0.0% | 0.0% | 1.1% | 21.6% | 77.3% | |
Higher education (university) | 0.0% | 0.0% | 2.9% | 25.7% | 71.4% | |
Library G* (N = 255) η2 = 0.038 | None or elementary school | 0.0% | 0.0% | 20.0% | 35.0% | 45.0% |
Lower secondary education | 0.0% | 0.0% | 18.8% | 56.3% | 25.0% | |
Higher secondary education | 1.1% | 4.3% | 17.4% | 57.6% | 19.6% | |
Higher education (college) | 4.8% | 6.0% | 20.5% | 48.2% | 20.5% | |
Higher education (university) | 4.5% | 11.4% | 25.0% | 40.9% | 18.2% | |
Library D* (N = 32) η2 = 0.237 | None or elementary school | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% |
Lower secondary education | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | |
Higher secondary education | 0.0% | 0.0% | 0.0% | 55.6% | 44.4% | |
Higher education (college) | 0.0% | 0.0% | 0.0% | 21.4% | 78.6% | |
Higher education (university) | 0.0% | 0.0% | 0.0% | 0.0% | 100.0% | |
Library H* (N = 252) η2 = 0.046 | None or elementary school | 0.0% | 18.2% | 18.2% | 18.2% | 45.5% |
Lower secondary education | 0.0% | 0.0% | 14.3% | 42.9% | 42.9% | |
Higher secondary education | 0.0% | 1.3% | 7.8% | 44.2% | 46.8% | |
Higher education (college) | 0.0% | 0.0% | 5.2% | 34.4% | 60.4% | |
Higher education (university) | 0.0% | 3.7% | 5.6% | 42.6% | 48.1% | |
Library C* (N = 49) η2 = 0.190 | None or elementary school | 0.0% | 0.0% | 0.0% | 25.0% | 75.0% |
Lower secondary education | 0.0% | 0.0% | 0.0% | 33.3% | 66.7% | |
Higher secondary education | 0.0% | 5.6% | 16.7% | 50.0% | 27.8% | |
Higher education (college) | 0.0% | 6.3% | 18.8% | 62.5% | 12.5% | |
Higher education (university) | 0.0% | 0.0% | 20.0% | 80.0% | 0.0% | |
Library J** (N = 207) η2 = 0.086 | None or elementary school | 0.0% | 0.0% | 10.0% | 10.0% | 80.0% |
Lower secondary education | 0.0% | 0.0% | 0.0% | 45.5% | 54.5% | |
Higher secondary education | 0.0% | 6.3% | 12.7% | 52.4% | 28.6% | |
Higher education (college) | 0.0% | 4.2% | 11.5% | 53.1% | 31.3% | |
Higher education (university) | 0.0% | 3.7% | 25.9% | 63.0% | 7.4% |
. | . | Not at all satisfied . | Not satisfied . | Neither satisfied, nor dissatisfied . | Satisfied . | Very satisfied . |
---|---|---|---|---|---|---|
Library E** (N = 195) η2 = 0.068 | None or elementary school | 0.0% | 0.0% | 0.0% | 50.0% | 50.0% |
Lower secondary education | 0.0% | 0.0% | 0.0% | 33.3% | 66.7% | |
Higher secondary education | 0.0% | 1.8% | 7.3% | 40.0% | 50.9% | |
Higher education (college) | 0.0% | 0.0% | 1.1% | 21.6% | 77.3% | |
Higher education (university) | 0.0% | 0.0% | 2.9% | 25.7% | 71.4% | |
Library G* (N = 255) η2 = 0.038 | None or elementary school | 0.0% | 0.0% | 20.0% | 35.0% | 45.0% |
Lower secondary education | 0.0% | 0.0% | 18.8% | 56.3% | 25.0% | |
Higher secondary education | 1.1% | 4.3% | 17.4% | 57.6% | 19.6% | |
Higher education (college) | 4.8% | 6.0% | 20.5% | 48.2% | 20.5% | |
Higher education (university) | 4.5% | 11.4% | 25.0% | 40.9% | 18.2% | |
Library D* (N = 32) η2 = 0.237 | None or elementary school | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% |
Lower secondary education | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | |
Higher secondary education | 0.0% | 0.0% | 0.0% | 55.6% | 44.4% | |
Higher education (college) | 0.0% | 0.0% | 0.0% | 21.4% | 78.6% | |
Higher education (university) | 0.0% | 0.0% | 0.0% | 0.0% | 100.0% | |
Library H* (N = 252) η2 = 0.046 | None or elementary school | 0.0% | 18.2% | 18.2% | 18.2% | 45.5% |
Lower secondary education | 0.0% | 0.0% | 14.3% | 42.9% | 42.9% | |
Higher secondary education | 0.0% | 1.3% | 7.8% | 44.2% | 46.8% | |
Higher education (college) | 0.0% | 0.0% | 5.2% | 34.4% | 60.4% | |
Higher education (university) | 0.0% | 3.7% | 5.6% | 42.6% | 48.1% | |
Library C* (N = 49) η2 = 0.190 | None or elementary school | 0.0% | 0.0% | 0.0% | 25.0% | 75.0% |
Lower secondary education | 0.0% | 0.0% | 0.0% | 33.3% | 66.7% | |
Higher secondary education | 0.0% | 5.6% | 16.7% | 50.0% | 27.8% | |
Higher education (college) | 0.0% | 6.3% | 18.8% | 62.5% | 12.5% | |
Higher education (university) | 0.0% | 0.0% | 20.0% | 80.0% | 0.0% | |
Library J** (N = 207) η2 = 0.086 | None or elementary school | 0.0% | 0.0% | 10.0% | 10.0% | 80.0% |
Lower secondary education | 0.0% | 0.0% | 0.0% | 45.5% | 54.5% | |
Higher secondary education | 0.0% | 6.3% | 12.7% | 52.4% | 28.6% | |
Higher education (college) | 0.0% | 4.2% | 11.5% | 53.1% | 31.3% | |
Higher education (university) | 0.0% | 3.7% | 25.9% | 63.0% | 7.4% |
Education and perceived invitingness of library (Spearman’s Rho, controlled for age).
. | Correlation . |
---|---|
Library A | −0.439** |
Library B | 0.397* |
. | Correlation . |
---|---|
Library A | −0.439** |
Library B | 0.397* |
Education and perceived attractiveness of library (Spearman’s Rho, controlled for age).
. | Correlation . |
---|---|
Library C | −0.386** |
Library B | 0.430** |
Library D | 0.481** |
. | Correlation . |
---|---|
Library C | −0.386** |
Library B | 0.430** |
Library D | 0.481** |
Percentages within education levels that visit public libraries in Flanders, data derived from different editions of Participation Survey (2003/2004, 2009 & 2014).
. | 2003/2004 (N = 2548) . | 2009 (N = 3144) . | 2014 (N = 3927) . | Difference 2004–2014 . |
---|---|---|---|---|
No or lower education | 11.8% | 11.6% | 7.6% | −4.20% |
Lower secondary education | 26.3% | 19.3% | 17.8% | −8.50% |
Higher secondary education | 33.8% | 27.8% | 26.1% | −7.70% |
Higher education | 48.9% | 44.1% | 45.8% | −3.10% |
Total | 36.6% | 33.0% | 30.0% | −6.60% |
. | 2003/2004 (N = 2548) . | 2009 (N = 3144) . | 2014 (N = 3927) . | Difference 2004–2014 . |
---|---|---|---|---|
No or lower education | 11.8% | 11.6% | 7.6% | −4.20% |
Lower secondary education | 26.3% | 19.3% | 17.8% | −8.50% |
Higher secondary education | 33.8% | 27.8% | 26.1% | −7.70% |
Higher education | 48.9% | 44.1% | 45.8% | −3.10% |
Total | 36.6% | 33.0% | 30.0% | −6.60% |