The Journal of Academic Librarianship 44 (2018) 724–737

Contents lists available at ScienceDirect

The Journal of Academic Librarianship

journal homepage: [Link]/locate/jacalib

Impacts of Library Space on Learning Satisfaction – An Empirical Study of T

University Library Design in Guangzhou, China☆
Ling Hin Lia, Fan Wub, , Boya Suc
⁎

a
Department of Real Estate and Construction, HKUrban Lab, Faculty of Architecture, The University of Hong Kong, Pokfulam, Road, Hong Kong, China
b
Department of Construction Management, School of Civil Engineering and Transportation, South China University of Technology, Jiaotung Building, Wushan Road,
Tianhe District, Guangzhou, China, 510640
c
Department of Construction Management, School of Civil Engineering and Transportation, South China University of Technology, China

ARTICLE INFO ABSTRACT

Keywords: The current generation of millennial university students is more accustomed to searching reference information
Library design online than visiting the physical library facility, compared to the previous generations of students. Studies have
Learning satisfaction shown that the role of the physical library facilities as a mere collection point of reading and reference materials
University libraries is being threatened by the availability of free and high-speed online search engines. University libraries have
Learning commons
always been an integral part in higher education learning activities, and they are not exempted from this threat.
SEM analysis
Based on a structural equation modelling framework, we analysed empirically the importance of different library
design features that help enhance students' learning satisfaction, and found that lighting environment, acoustic
environment as well as location of the library building were the main determinants impacting on the use of the
university library by students in a major university in China. We conclude the paper with our suggestions in
modifying library design to accommodate students' learning needs, and more importantly in recon uring the
spatial and functional role of university libraries in this age of digital information from a mere provider of
reference materials to a physical space of learning commons on campus.

Introduction the only source of reference information for university students, espe-
cially the undergraduate students who basically grew up in this era.
Students studying at universities are more and more reliant on on- Questions such as “do we still need libraries” or “do libraries still need
line library services as well as other online reference sites such as books” (Buss, 2016; Carlson, 2002; Gray & Tracy, 2011) start to emerge
“Wikipedia” when searching for information to complete their course- in some research studies to reflect the current state of challenges phy-
work, or conducting general learning activities (Thornton-Verma, 2012; sical library facilities face, namely, the spatial and functional role of a
Van Scoyoc & Cason, 2006). This can be evident from the growing physical library facility nowadays. While we do not completely agree
number of references in students' coursework or dissertations ending with the existential crisis of library facilities, we do think that there is a
with the phrase: “retrieved from…”. The ease of downloading academic need for library facilities to evolve with the socio-technological change
papers/documents from any computer or smart phone students have in the university learning mode so that they can still maintain a vital
access to is of course a major reason for this new learning behaviour. role in enhancing students' learning satisfaction on campus. In doing so,
Virtual libraries have become a phenomenal development in the field of we believe that university libraries can maintain the pivotal role of an
information and reference provision around the world when informa- essential part of higher education in the age of new information tech-
tion and reference materials can be delivered to the users without their nology.
going into any physical library facility (Booth, Walton, Fraser, This paper intends to contribute to this debate on the spatial and
Urquhart, & Van Loo, 2002). Moreover, millennial undergraduate stu- functional role of library facilities in our urban environment based on
dents nowadays are used to working in odd hours, such as midnight, an empirical analysis of a university library in China. This will provide
when they cannot find access to a physical library. In this new age of a more scientifically robust foundation for our arguments and re-
online information, library as a confined physical location ceases to be commendations in the conclusion section at the end of this paper. This

☆
Declarations of interest: none.
⁎
Corresponding author.
E-mail addresses: lhli@[Link] (L.H. Li), ctfanwu@[Link] (F. Wu).

[Link]
Received 20 July 2018; Received in revised form 3 September 2018; Accepted 9 October 2018
Available online 25 October 2018
0099-1333/ © 2018 Elsevier Inc. All rights reserved.
L.H. Li et al. The Journal of Academic Librarianship 44 (2018) 724–737

paper is divided into several sections. First of all, we will briefly discuss library users (mainly undergraduate students) were deciphered and
the correlation between university' learning environment and students' significant variables were identified and highlighted in the testing
learning outcomes and satisfaction. Following this, we will discuss model. The outcomes of the analysis will help contribute to the current
previous studies on the impacts of the spatial design of libraries on discussion on how to facilitate libraries, especially university libraries,
university students. In this paper, we examine university libraries as the to become a pivotal physical space for learning in the age of digital
target as opposed to public libraries in general because university li- information for university students.
braries have always been one of the most integral parts in university
education, and university libraries are symbolic to higher education Physical library facilities and learning
learning philosophy (Cantwell, 2013). Built on this background of lit-
erature review on the role of physical library space, we conducted our Libraries are traditionally physical facilities that provide a con-
empirical analysis in a major university in China, South China Uni- centration of book collections for sharing among a specific group of
versity of Technology (SCUT) in Guangzhou. By applying a structural users (such as university libraries) or the general public. However, in
equation modelling framework (SEM), we were able to decipher how this age of digital information, finding references for academic research
university students view the spatial and functional role of university or completing assignment is less and less dependent on the physical
libraries and how different design aspects impact on their learning sa- visits to libraries. Online reference search starts to be viewed as a more
tisfaction on campus. The discussion section details our findings, fol- effective replacement or even as a threat to physical library collections
lowed by our conclusion in this study. (Little, 2011; Odling-Smee, 2007; Rapp, 2011). In essence, library space
Learning environment and learning satisfaction in university edu- works as a central distribution and circulation platform for book-
cation. sharing activities. As it evolves, library as a physical space takes up an
Learning outcomes have almost become an important keyword in addition role of providing a comfortable, quiet and safe environment
higher education curriculum design in the recent years. Learning out- for self-regulated learning activities as well. Although access to space
comes measure how students excel during the period of studies at the within library is open to public or to a specific group of users within a
university and how this will lead to better success for students, which is certain community, library space becomes a sanctuary for book-lovers
one of the main foci of higher educational institutions nowadays in a “free of judgment” environment (Row Farr, 2017). This is feasible
(Jouhari, Haghani, & Changiz, 2015; Verešová, 2014). It fact, it has mainly because of the management of library space within the facilities.
been described that learning outcomes of university students determine “Quietness” is a common denominator in all kinds of libraries, except
the function and objective of higher education (Liu & Chang, 2014). for specific sections in the library, as users are supposed to conduct
The American College Personnel Association (ACPA) and National reading activity only. Hence, verbal discussions are usually forbidden in
Association of Student Personnel Administrators (NASPA) of the U.S. most areas of the library building, and users can enjoy reading or
(2004) depict that in general, students' learning outcomes come under working on their own without being disturbed by others. This makes
seven major categories, namely cognitive complexity, knowledge ac- libraries important as they take up the role of “learning commons” for
quisition, integration and application, humanization/humanity, civic people without the privilege of privacy at home. Hence, there exists a
responsibility, interpersonal relationships and self-understanding strong tie between library facility as a physical space and learning sa-
ability, and continued schooling and academic achievements. Almost tisfaction of the users that has been shown to be of important academic
all of these cognitive and non-cognitive outcomes are developed, fos- interests (Bilandzic & Foth, 2014; Cullen, 2014; Jolly & White, 2016;
tered and enhanced on campus during the study period. University Rudzioniene, 2014). Since libraries, especially university libraries, are
education therefore provides an arena of training for these outcomes specifically built for such purpose, satisfactory collaboration between
through learning and related activities carried out on campus as well as library design features and learning environment allows maximization
outside campus. In any case, these activities, especially those on- of learning satisfaction if such correlation is well-recognised and ap-
campus ones, are conducted within a learning environment specifically plied. In additional to conventional functions, Bilandzic and Foth
designed to maximize such learning outcomes and satisfaction. The (2014) have also shown that architecture characteristics of library
learning environment therefore must be able to fulfil the task of sti- space can play a significant role in highlighting the social attribute of
mulating students' desire to quest for new knowledge. This includes library facilities in amplifying the social interactions among visitors to
providing physical space on campus that allows students to carry out the library. Such informal learning space incorporated as part of the
interactive learning exercises that facilitate “learning by doing” (Dzeng, campus environment within the university library is equally important
Lin, & Wang, 2014). in higher education agenda (Deed & Alterator, 2017). In this way, li-
Nevertheless, since our questionnaire survey covers students braries still maintain an important learning and social role in the age of
(mainly undergraduate students) from various academic programmes digital information, as long as the design and management team of
on the SCUT campus in Guangzhou, China, it becomes difficult to academic libraries understand the needs of the users and appreciate
measure learning outcomes objectively among these students from very that there are aspects which just cannot be replaced by digital tech-
different academic programmes. Consequently, the dependent variable nology, including communications and interactions among users, ap-
has been set to be the learning satisfaction among the users of the preciation of culture and arts, as well as the sense of scholarship and
university library in our SEM analysis, which will be explained in fur- inspiration inside the library facilities (Jolly & White, 2016; Palfrey,
ther below. 2015; Stojanovski, 2013), not to mention the fact that it is still doubtful
Hence, to achieve better university learning satisfaction, it is if digital giants such as Google can actually digitize all collections of
therefore imperative to incorporate the element of motivation in the books in a cost-effective way (Fialkoff, 2011; Palfrey, 2015). More
learning environment. Motivation to study and learn on the other hand importantly, university administrators should also recognize the unique
depends on a number of other exogenous and endogenous factors, in- nature of library space in promoting university learning culture, rather
cluding teaching style, guidance and feedback given by the instructors, than treating library space as a surplus physical space that can be
and more interestingly, gender of the students (Chesbrough, 2011; consumed by other units for non-learning related activities from time to
Özütürk & Hürsen, 2014; Ro & Knight, 2016). time.
The main objective of this paper, therefore, is to empirically ex- Given the importance of design features of library space to the
amine the relative importance of different environmental variables that learning satisfaction of users, researchers start to advocate user-or-
exist in university library space to learning satisfaction among uni- iented design principles that allow multifunctional space to motivate
versity students by means of a robust scientific model of SEM. Through effective learning among library users to be taken into consideration
a large scale questionnaire survey on the SCUT campus, the views of the when designing and building libraries. By incorporating the users' views

725
L.H. Li et al. The Journal of Academic Librarianship 44 (2018) 724–737

through surveys, novel idea such as co-design model can help build us understand more about the impacts of library facilities design and
libraries that can maximize their academic and social functions spatial arrangements. This will also contribute to the current debate on
(Tevaniemi, Poutanen, & Lähdemäki, 2015). Ellison (2016) echoes in whether and how the advent of high speed information in the digital
her studies with this need for incorporating the views of the library age will drastically replace or displace some of our traditional social
users when designing academic libraries that promote better learning space such as libraries in our urban environment.
outcomes. Her studies show very detailed design features such as
lighting, signature and desk space that are influential to students' Methodology
learning activities. In the other words, when designing library facilities,
considerations should be given to the evolving role and functions of Structural Equation Modelling (SEM) is adopted in this paper as the
library space from the users'/visitors' perspective, as they are given core analytical framework. SEM includes a diverse set of mathematical
more choices nowadays in sourcing reference information and mate- models, computer algorithms, and statistical methods that fit networks
rials. For library space to become more functional and valuable to the of constructs to data. Our main explanatory variables are the impacts of
users, design features of library facilities should enhance other role and environmental and design features of university library on students'
function that are not easily replaced by the digital transformation in our learning satisfaction. They are all abstract concepts in nature (latent
society, namely as a place to enrich and enhance learning experience, variables) and we measured them through various indicators in our
academically and socially for the users. In the following, we will con- analysis. SEM allows us to diagnose which observed variables are good
tribute to this aspect of the built environment research with a more indicators of the latent variables. This is the measurement model, one of
scientific method of structural equation modelling (SEM) that will help the two main SEM components. The other main component is the

Fig. 1. Geographical distribution of major students' quarters on SCUT campus.

726
L.H. Li et al. The Journal of Academic Librarianship 44 (2018) 724–737

structural model, which lays out the relationships among latent vari- lot of studies devoted to analysing and evaluating library learning en-
ables from a theory or hypothesis. Basically, SEM combines factor vironment empirically for us to develop an evaluation system on sa-
analysis and multiple regression analysis. Structural model consists of tisfaction level from users' point of view. This may be due to the fact
several latent variables and path relationship between variables. The that traditionally, libraries are regarded as “confined space” holding the
effect of a path can be obtained through using the structural equation collection of books and reference materials in our urban society
model. SEM is suitable for this project because it assesses the multiple (Nicholas, Watkinson, Rowlands, & Jubb, 2011), and hence the spatial
and interrelated dependence among all the variables in a single ana- role of libraries in our urban city tends to be more on the enhancement
lytical framework. of accessibility to these reference materials in different urban districts.
The latent variables of SEM are reflected by multiple-observed In order to identify the relevant environmental factors within the li-
variables, avoiding the multicollinearity problem between independent brary facility that may impact on students' learning satisfaction for our
variables in traditional linear regression. In addition, SEM allows studies, we need to create an evaluation system, based on a general
multiple dependent variables in a model at the same time, which solves literature review on environmental impacts on work efficiency and
the problem that only one dependent variable is allowed in the tradi- performance outcomes of users in a confined physical space. For ex-
tional linear regression. Combined with the actual situation of this ample, Banbury and Berry (1998) find that noise is a major factor af-
study, SEM is used to explore the key design points of university library fecting staff productivity as it reduces memory performance and mental
that affect students' learning satisfaction in our sample. computing capacity of staff. Groth (2007) believes that lighting can
Moreover, SEM is chosen in this research for its versatility that can affect people's behaviour, and consequently productivity, which echoes
be applied in different fields of research. SEM has been widely adopted with the study by Ketutwijaya (2012). In addition, it has been found
in the medical and health care research field for a better understanding that poor indoor air quality (including high carbon dioxide concentra-
of the factors leading to certain illness or those that will help improve tions) can affect students' learning ability (Choi, Guerin, Kim, Brigham,
the patients' problems, as these factors sometimes tend to intertwine & Bauer, 2014; Lee et al., 2012). Furthermore, the general learning
with each other and are difficult to examine individually (Alessandrini environment on campus as well as accessibility of the physical facilities
et al., 2016; Castro, Baltar, Marchioni, & Fisberg, 2016; Cois & Ehrlich, have all been regarded as important factors (Castilla, Llinares, Bravo, &
2014; Lewis et al., 2017; Mitchell, Anderson, & Ferguson, 2017). In Blanca, 2017).
addition, because of its capability to dissect latent variables with mul- Consequently, after reviewing these various studies, we have finally
tiple observations, SEM has also been adopted in other academic fields. arrived at a total of six major categories of environmental factors per-
For instance, SEM has been applied in the examination of the factors taining to our analysis of library spatial design that may impact on
affecting aggressive behaviours (Mckay, Perry, & Harvey, 2016); in the
study of work-family-school conflicts and social support for the medical
care profession (Goong, Xu, & Li, 2016); in the analysis of the factors Table 1
contributing to the sustainability of cities that depend on heavy in- Personal characteristic of respondents.
dustry (Zhang, Zhang, Zhang, Yu, & Zhu, 2016); in the study of planned Personal characteristic Category Quantity % distribution
behaviour in the utilisation of natural resources with supply shortage,
Gender Male 275 59.5%
such as water (Cooper, 2017); in the examination of strategic man-
Female 187 40.5%
agement of public institutions such as universities (Dandagi, Bhushi, Age Below 18 1 0.2%
Bagodi, & Sinha, 2016); as well as in general business production and 18–20 122 26.5%
marketing analysis (Ajayi & Oyedele, 2018; Martínez-López, Gázquez- 21–23 251 54.3%
Abad, & Sousa, 2013; Pal Pandi, Paranitharan, & Jeyathilagar, 2018; 24–27 75 16.2%
Over 28 13 2.8%
Thirupathi & Vinodh, 2016). This paper therefore intends to contribute
Academic status Undergraduate year 93 20.1%
to this rich body of literature on the application of SEM with a specific 2
angle of the changing spatial and functional role of academic libraries Undergraduate year 131 28.4%
in the age of digital information in enhancing students' learning sa- 3
Undergraduate year 117 25.3%
tisfaction.
4
In this paper, data to be utilized in the SEM analysis will be col- MDC year 1 53 11.5%
lected from a major university in China, the South China University of MDC year 2 27 5.8%
Technology (SCUT) in Guangzhou, China. There are a number of rea- MDC year 3 21 4.6%
sons for this choice. South China University of Technology (SCUT) is Doctoral candidate 13 2.8%
Teaching and 3 0.6%
one of the top universities directly under the Ministry of Education in
research staff
China, and student quality is therefore also very high, leading to a Others 4 0.9%
reasonable assumption of higher frequency in library visits at SCUT. Residential location/students' East 127 27.5%
Moreover, the architectural design programme at SCUT is ranked the quarter districts (see West 186 40.3%
Fig. 1 below) North 135 29.2%
5th among all the national higher education institutions in China. In
South 3 0.6%
this research, the SCUT Library was designed and renovated by the Others 11 2.4%
Architectural Design and Research Institute (ADRI), a subsidiary of
SCUT, in 2017. Therefore, this library can represent a combination of Notes
advanced design theories as well as professional practice in archi- MDC- Master Degree Candidate
tecture. Moreover, some of the staff at the ADRI are also teaching staff Year 1 undergraduates are not included in this survey since they study and live
and graduate students of SCUT, the design of the university library in another campus in the Guangzhou University town.
should therefore have got input from the users' point of view from the There are mainly four students' quarter districts (see Fig. 1 below):
East District is about 0.7 km away from the library which needs 9 min of tra-
very beginning. Since the objective of this paper is to examine the
velling on foot.
spatial and functional role of university library in the current age of
West District is about 0.85 km away from the library which needs 11 min of
digital information when reference search is made more easy at home, travelling on foot.
an academic library recently designed and built by a professional team North District is about 2.3 km away from the library which needs 29 min of
with direct and indirect connection to and knowledge of the university's travelling on foot.
learning environment should help illuminate our results. South District is about 0.6 km away from the library which needs 8 min of
According to the literature review elaborated above, there are not a travelling on foot.

727
L.H. Li et al. The Journal of Academic Librarianship 44 (2018) 724–737

Fig. 2. Evaluation system and initial structural equation model.

Table 2 relationship on how they may impact on learning satisfaction of uni-

KMO and Bartlett's test versity students in our model is illustrated in Fig. 1 below. The eva-
Kaiser-Meyer-Olkin measure of sampling adequacy 0.844 luation will be carried out through the Structural Equation Modelling
(SEM) at a later stage.
Bartlett's test of sphericity Approx. chi-square 1907.896 Based on these six categories of latent variables, we then proceed to
df 171
the SEM analysis, with data obtained from a questionnaire survey de-
Sig. 0.000
signed for and distributed to students (mainly undergraduate) and a few
research staff in early 2018 on the campus of SCUT (the full ques-
tionnaire is attached in the Appendix). The questionnaire mainly con-
students' learning satisfaction. These six categories will form the basis sists of four parts, namely, personal details of the respondents, their
of our evaluation system, they include acoustic environment, light en- learning behaviour, evaluation on their satisfaction of individual library
vironment, air quality, learning space, learning facilities, and accessi- learning environment variables and the overall satisfaction of the
bility (“accessibility” in this paper is specifically defined as the ease of learning environment in the SCUT library. Objective answers were
reaching the library by the students either on foot or by school bus from collected based on a Likert 7-scale method. The questionnaires were
their student quarters on the SCUT campus). Each category can be distributed through online social network applications and face-to-face
further sub-divided into individual variables in our questionnaire contacts in the library and various classrooms on the SCUT's main
survey (Fig. 1). In our analysis, we will also include two paths in the campus in early 2018. A total of 529 questionnaires were collected in
analytical framework as major parameters contributing to students' this survey, of which 462 were decided to be valid for analysis and the
learning satisfaction as these two paths seem to be more relevant to effective rate was 87.33%.
learning satisfaction. These two paths are visit time to the university
library and respondents' learning status. Learning status is generally Data analysis and results
defined as two major directions, namely learning efficiency and
learning attention (Everaert, Opdecam, & Maussen, 2017). These two Table 1 shows the personal characteristic of the respondents. The
paths form a structural relationship together as the evaluation system of gender ratio of the whole sample is very near to the gender ratio of
the variables in the six categories outlined above. The overall structural undergraduate students at the SCUT campus. It reflects a balanced

728
L.H. Li et al. The Journal of Academic Librarianship 44 (2018) 724–737

Fig. 3. Scree plot of eigenvalues on all factors.

Table 3 Table 4
Total variance explained Rotated component matrix and grouped variables
Component Initial eigenvalues Categories Factors Factor loading

Total % of variance Cumulative % Light environment Natural lighting 0.773

Electric lighting 0.711
1 2.001 10.534 10.534 Electronic screen glare 0.630
2 1.948 10.252 20.786 Learning supporting facility Collection management 0.776
3 1.932 10.170 30.956 Bookshelf design 0.679
4 1.726 9.086 40.042 Supporting facility convenience 0.599
5 1.705 8.975 49.017 Air quality Temperature 0.824
6 1.666 8.766 57.784 Humidity 0.816
7 1.633 8.595 66.378 Air freshness 0.538
Acoustic environment Typing noise 0.804
Users movement noise 0.769
Equipment and facilities noise 0.533
Accessibility Private traffic conveniencea 0.886
distribution of respondents.
Public traffic convenienceb 0.878
Fig. 2 above illustrates that there are a total of 22 library environ- Self-learning space Adequacy of self-learning spaces 0.861
mental variables listed as potentially relevant to learning satisfaction. Learning space configuration 0.651
Before we carried out the SEM analysis, we needed to further analyse Discussion room design 0.542
whether all these variables are significant or not. To achieve this ob- Desk and chair design Chair design 0.818
Desk design 0.766
jective, we conducted the exploratory factor analysis (EFA).
Notes:
Exploratory factor analysis (EFA) of the independent variables a
Private traffic convenience refers to walking on foot, riding personal bi-
cycles or driving private vehicles(which is mainly applicable to research staff in
Exploratory Factor analysis was applied to extract the common the sample).
factors from the library learning environment variables. The Kaiser- b
Public traffic convenience refers to taking school buses or taking the shared
Meyer-Olkin (KMO) measure of sampling adequacy was calculated at bikes on campus.
0.844, which was considerably higher than the required minimum
(Table 2). The KMO data represents a strong correlation between
variables which are suitable for exploratory factor analysis. According shows the solution accounted for 66.378% of the total variance in the
to the difference of extraction results and research assumptions, un- questionnaire. Variables were categorized and grouped, as shown in
reasonable items were omitted. Then, the common factors were ex- Table 4. Table 4 indicates that from the original total 22 variables, 19 of
tracted again. A scree plot was created which showed that the number them were found to be significant and relevant and they were kept in
of dimensions involved seven significant variables only (Fig. 3). Table 3 the model.

729
L.H. Li et al. The Journal of Academic Librarianship 44 (2018) 724–737

Table 5 Path analysis

KMO and Bartlett's test
Kaiser-Meyer-Olkin measure of sampling adequacy 0.525 Through EFA and CFA, a reasonable structural equation model was
finally obtained. After the insignificant influence paths were deleted,
Bartlett's test of sphericity Approx. chi-square 300.948 the final structural equation model was obtained which had four sig-
df 6
nificant paths (Fig. 5a and b). The basic fitness index is shown in
Sig. 0.000
Table 10. At this stage, the model is proved again to be a good fit.
To finalise our structural equation model, we carried out a stepwise
Table 6 regression to test whether the four paths were significant. Table 11
Total variance explained shows that these four paths identified are statistically significant.

Component Initial eigenvalues

Discussion of results
Total % of variance Cumulative %
Our SEM analysis shows that some internal design criteria for uni-
1 1.682 42.045 42.045 versity libraries are significantly relevant to the learning satisfaction of
2 1.400 35.000 77.045
the students using the university library space. For university libraries
to provide the spatial function of a learning commons, in addition to a
confined space for reference materials, some design features need to be
Table 7 considered. First of all, our study finds that students at SCUT still rely
Rotated component matrix and grouped variables on the physical space provided by their university library to study and
Categories Factors Factor loading to carry out self-learning. As a result, satisfactory lighting design will
affect their degree of learning satisfaction. Lighting facilities affect di-
Learning status Attention 0.916 rectly the comfort of students' eyes when they are reading inside the
Learning efficiency 0.912
library. After conducting on-site analysis, we find that the lighting en-
Visit time Visiting frequency per week 0.841
Average time for every visit 0.826
vironment varies with different sections of the SCUT library building.
For some reasons, there are no curtains hung on the windows of the
library. As a result, the eastern part of the library tends to benefit from
natural lighting in the morning, while such natural light will switch to
EFA of the dependent variables the western part of the facility in the afternoon (Fig. 6a). On the other
hand, some areas in the interior lack both natural and artificial lightings
Exploratory Factor analysis was then applied to extract the common all the day, leading to lower popularity when students have a choice, or
factors from learning behaviours for the dependent variables. The unsatisfactory learning outcomes when they do not have a choice but
Kaiser-Meyer-Olkin measure of sampling adequacy were calculated at have to settle for these areas (Fig. 6b). In terms of lighting environment,
0.525, which was considerably higher than the required minimum there is no gender difference in terms of satisfaction level.
(Table 5). The KMO data represents a strong correlation between Secondly, the satisfaction level of acoustic environment has shown
variables which are suitable for exploratory factor analysis. Table 6 to have a significant impact on the visit time. While most people as-
shows the solution accounted for 77.05% of the total variance in the sociate the word “silence” immediately with libraries, it does not pre-
occupant questionnaire. Variables were categorized and grouped, as clude noise from emerging inside the library facility, even students are
shown in Table 7. not deliberating talking to each other. Most students do respect the
basic rule of using the library, i.e. to be quiet. However, one's normal
learning behaviour may inadvertently impact on others negatively.
Confirmatory factor analysis (CFA) of the independent variables Studying in the age of information means that students will also bring
along their notebook computer with them for working on their as-
CFA was further performed to test the fitness between optimized signments as well as conducting internet search. When all students are
measurement models and data collected. The optimized measurement learning in a confined public space, the noise produced by students
models of exogenous latent variables were drawn in the AMOS. As when typing on their notebook computer will amount to quite a sub-
shown in Fig. 4, the factor loadings for all items were above 0.5, so no stantial level. Such a level of working noise will eventually affect other
item deletion was required at this point. The fitness indices were then students who are reading or trying to concentrate on their work.
assessed and they showed that the model was a good fit. The basic Interestingly enough, male students tend to be more tolerant of such
fitness index is presented in Table 8. working noise generated from typing than female students.
When the significance of Chi-square degree of freedom is < 0.05, it On the other hand, as noticed from Fig. 6a and b, the floor of SCUT
indicates that there is a significant difference between model and actual library is not carpeted for easy maintenance reason. However, it creates
data. In fact, large samples are usually difficult to guarantee that the another noise problem, which is the noise generated by foot-traffic.
value of significance is > 0.05. Thus, it is necessary to test other fitness Noise generated by foot-traffic is a predicament for well-designed li-
indices. Since both GFI and AGFI exceeded 0.9 as shown in Table 8, brary space. Good learning environment attributed to good design
they indicated that the model was an acceptable fit. The value of features contributes to students' higher learning satisfaction level. Good
RMSEA was < 0.05, indicating that the model was not affected by the learning environment therefore attracts more students to utilize the
number of samples and the complexity of the model. In addition, since library space for studying and learning, leading to higher volume of foot
PGFI was larger than 0.5, the model was shown to be a good fit. traffic in most part of the library facilities, which will inevitably gen-
erate negative noise impact (Figs. 7 and 8). However, different from
typing sound generated from using notebook computer, foot-traffic
CFA of the dependent variables noise tends to be transient and sporadic, and not all students feel the
same towards such noise. Students who are more sensitive to such
In the next step, the endogenous latent variables were drawn in the footstep noise will eventually shy away from using the library for
AMOS. The basic fitness index is presented in Table 9. It also suggests learning when they need a more tranquil environment to do so. Our
that the model is a good fit. results show that students who are not satisfied with such noise

730
L.H. Li et al. The Journal of Academic Librarianship 44 (2018) 724–737

Fig. 4. CFA of independent variables.

Table 8 Table 9
Fitness indices (measurement model of independent variables) Fitness indices (measurement model of dependent variables)
Indices Appropriate range Actual value Fit judgement Indices Appropriate range Actual Value Fit judgement

Sig. > 0.05 0.000 No Sig. > 0.05 0.509 Yes

Chi-square degree of freedom < 2.00 1.588 Yes Chi-square degree of < 2.00 0.436 Yes
RMSEA < 0.08 0.04 Yes freedom
GFI > 0.90 0.94 Yes RMSEA < 0.08 0.000 Yes
AGFI > 0.90 0.913 Yes GFI > 0.90 0.999 Yes
PGFI > 0.50 0.648 Yes AGFI > 0.90 0.994 Yes
PGFI > 0.50 0.100 Yes

environment will reduce learning time spent inside the library.

Thirdly, our results also show that the accessibility satisfaction, requires taking vehicular transportation (in particular, North District in
namely the ease of reaching the library, has a significant impact on Fig. 1 above), our results show that students who don't have to spend a
learning satisfaction. Accessibility in our study refers to the travel time lot of time on travelling to and from the library tend to spend more time
for students to get to the library. In Mainland China, almost all students in the library. In addition, they also report better learning satisfaction
are required to live in student quarters on campus. In our study, we compared to those who do not have the same accessibility advantage.
factored the location of respondents' quarter into the analysis. Since Again, there is no gender difference in terms of satisfaction level in this
some student quarters are farther away from the SCUT library which variable.

731
L.H. Li et al. The Journal of Academic Librarianship 44 (2018) 724–737

a Initial structural equation model after CFA b Final structural equation model
Fig. 5. a Initial structural equation model after CFA
b Final structural equation model.

Table 10 materials online anywhere and everywhere they go. While recent stu-
Fitness indices of final structural equation model dies show that the role of library as a confined space holding collection
Indices Appropriate range Actual value Fit judgement
of reading materials may be under threat by the advent of powerful
online reference search engines and platforms, we notice that university
Sig. > 0.05 0.026 No libraries still maintain an irreplaceable role of being a learning com-
Chi-square degree of freedom < 2.00 1.436 Yes mons. Our study in Guangzhou illustrates that this is especially the case
RMSEA < 0.08 0.034 Yes
GFI > 0.90 0.969 Yes
where most students are still living on or near the campus, and student
AGFI > 0.90 0.949 Yes quarters/dormitories may not be ideal for their learning outcomes.
PGFI > 0.50 0.569 Yes Based on a robust SEM analytical framework and reinforced by various
model fitness tests, our study confirms this spatial and functional role
offered by the university library that helps students to enhance their
Table 11 learning satisfaction, as long as a certain design criteria are thoroughly
Parameter Estimates of final structural equation model considered in configuring the physical space of library facilities.
Path Standardized S.E. C.R. P Label Among all the library environmental variables identified as sig-
regression nificant to learning satisfaction in this paper, we find three major areas
weights with significant impact on university students' learning behaviour, as
Learning status ← light 0.256 0.074 3.153 0.002 par_9
well as their willingness to visit the library facilities. From our results,
environment there are three library learning environment categories which have
Learning 0.110 0.059 1.699 0.089 par_10 significant impacts on learning satisfaction, namely, lighting environ-
Status ← accessibility ment, accessibility and acoustic environment. Interestingly enough, our
Visit time ← acoustic 0.129 0.070 1.723 0.085 par_8
respondents seem to enjoy natural lighting and they tend to follow such
environment
Visit time ← learning status 0.149 0.056 2.214 0.027 par_14 natural sunlight for studying inside the library. One possible explana-
tion for this is that artificial lightings in academic buildings on campus
in Mainland China are usually not adequate. Hence, lighting design in
Conclusion the study areas that will help students' learning outcome on university
campus should be enhanced. We notice that in the recent decades,
We set out to examine the spatial and functional role of academic national investment on higher education in China has increased sub-
library within university campus in the age of high-speed information stantially, but mainly in research funding. Investment on university
technology when most students can easily find academic reference teaching and learning facilities, such as university libraries, should be

732
L.H. Li et al. The Journal of Academic Librarianship 44 (2018) 724–737

a : Natural lighting b : Artificial lighting

Fig. 6. a: Natural lighting
b: Artificial lighting.

and e-books are getting more and more common and affordable, and
the existence of academic libraries now depends more on how the en-
vironmental attributes of the confined space can help students' learning
activities to become more satisfactory.
Similarly, acoustic environment is also very important to students
using the library as a learning facility. From our study we recommend,
in terms of architectural design requirements, that the self-study areas
can be divided into purely reading section and working section. The
former can impose stricter rules on the use of electronic equipment with
better design such as carpeted floor, while the latter can allow a higher
degree of students' interaction, including group discussions among
students. In this way, students with different learning needs and re-
quirements can visit the section that accommodates their study needs.
Moreover, entrances, exits, as well as staircases should be far away from
the self-study areas to minimize noise generated by foot traffic. Thirdly,
in terms of location of the university library, where the campus is re-
Fig. 7. Design of learning area and book collection. latively large, and student quarters tend to scatter all over the campus,
we suggest enhancement in transportation arrangement, especially at
night for safety concerns among female students. Where funding and
space are both available, satellite libraries/learning centres could be set
up near remotely-located quarters to increase the accessibility of uni-
versity library.
To conclude, we find that although the current generation of mil-
lennials grew up in the high-speed information age and have been ac-
customed to finding reference materials online rather from visiting the
physical library facility, this does not relegate the spatial and functional
role of libraries to just a simple storage space. University libraries are an
important platform for students' learning activities. University libraries
as a well-designed confined space that accommodate students' learning
requirements are still instrumental in enhancing their learning out-
comes and satisfaction, and library as a learning commons still com-
mands a significant role in higher education development, as well as in
the community in general.

Acknowledgments
Fig. 8. Self-learning area.
The authors would like to express their gratitude for the research
funding support given by the South China University of Technology
strategically allocated as well. There should be a more diversified and undergraduate teaching research and reform project (project number
balanced funding allocation between book collection and physical x2tj/Y1180421), and Teaching Development Grant given by the
space design/library facilities. This is especially true when e-journals University of Hong Kong (Project Number 101000664).

733
L.H. Li et al. The Journal of Academic Librarianship 44 (2018) 724–737

Appendix A

A.1. Questionnaire on learning environment satisfaction of South China University of Technology Library

A.1.1. The first part consists of your personal basic information and learning behaviour
1. Gender of the respondent.
a. Male b. Female.
2. What is your age?
a. Under 18 years b. 18 years c. 19 years d. 20 years e. 21 years f. 22 years g. 23 years h. 24 years i. 25 years j. 26 years k. 27 years l. 28 years or
over.
3. How would you describe your status at SCUT?
a. Undergraduate Year 1 b. Undergraduate Year 2 c. Undergraduate Year 3 d. Undergraduate Year 4 e. MDC year 1 f. MDC year 2 g. MDC year 3 h.
Doctoral candidate. i. Teaching and research staff j. Others.
4. What is your current residential location?
a. East district of SCUT b. West district of SCUT c. North district of SCUT d. South district of SCUT. e. Others.
5. How often do you visit the Library of SCUT?
a. Never b. Once a week c. Twice a week d. 3 times a week e. 4 times a week f. 5 times a week g. 6 times a week h. 7 times a week i. 1–3 times a
mouth j. Others.
6. What is the average time you spend in the library on an average day?
a. 0–2 h b. 2–4 h c. 4–6 h d. 6–8 h e. 8–10 h f. 10–12 h g. 12–14 h h. over 14 h.

A.1.2. Next, you can tick a box that best represents your view in the following questions
7. What do you think of your learning efficiency inside the library?

Low 1 2 3 4 5 6 7 High

8. How easy can you concentrate on learning inside the library?

Easily distracted 1 2 3 4 5 6 7 Highly concentrated

A.1.3. The second part focuses on your satisfaction of the library learning environment
9. Very dissatisfied Very satisfied

Laptop typing noise 1 2 3 4 5 6 7

(mouse, keyboard)

10. Very dissatisfied Very satisfied

Foot-traffic noise 1 2 3 4 5 6 7

11. Very dissatisfied Very satisfied

Equipment and facilities 1 2 3 4 5 6 7

noise (air conditioner,
drinking fountain, etc.)

12. Very dissatisfied Very satisfied

Natural lighting of 1 2 3 4 5 6 7
learning space

13. Very dissatisfied Very satisfied

Artificial lighting of 1 2 3 4 5 6 7
learning space

14. Very dissatisfied Very satisfied

Electronic screen glare 1 2 3 4 5 6 7

(laptop, mobile phone)
*Glare means that extreme brightness contrast which can cause visual discomfort.

734
L.H. Li et al. The Journal of Academic Librarianship 44 (2018) 724–737

15. Very dissatisfied Very satisfied

Indoor temperature 1 2 3 4 5 6 7

16. Very dissatisfied Very satisfied

Indoor humidity 1 2 3 4 5 6 7

17. Very dissatisfied Very satisfied

Indoor air freshness 1 2 3 4 5 6 7

18. Very dissatisfied Very satisfied

Desk design 1 2 3 4 5 6 7

19. Very dissatisfied Very satisfied

Chair design 1 2 3 4 5 6 7

20. Very dissatisfied Very satisfied

Bookshelf design 1 2 3 4 5 6 7

21. Very dissatisfied Very satisfied

Power supply 1 2 3 4 5 6 7
convenience

22. Very dissatisfied Very satisfied

Network/wifi 1 2 3 4 5 6 7
convenience

23. Very dissatisfied Very satisfied

Supporting facility 1 2 3 4 5 6 7
convenience (storage
locker, etc.)

24. Very dissatisfied Very satisfied

Collection management 1 2 3 4 5 6 7
(recovery efficiency,
extent of books
collection, etc)

25. Very dissatisfied Very satisfied

Configuration of learning 1 2 3 4 5 6 7
space
* The learning space configuration reflects the degree of over-
library.

735
L.H. Li et al. The Journal of Academic Librarianship 44 (2018) 724–737

26. Very dissatisfied Very satisfied

Adequacy of self- 1 2 3 4 5 6 7
learning spaces

27. Very dissatisfied Very satisfied

Discussion room design 1 2 3 4 5 6 7

28. Very dissatisfied Very satisfied

Division of acoustic 1 2 3 4 5 6 7
zones

29. Very dissatisfied Very satisfied

Private traffic 1 2 3 4 5 6 7
convenience (walking,
private bikes or cars)

30. Very dissatisfied Very satisfied

Public traffic 1 2 3 4 5 6 7
convenience (shared
bicycles and school
buses)

31. Very dissatisfied Very satisfied

Overall learning 1 2 3 4 5 6 7
environment

References Cooper, B. (2017). What drives compliance? An application of the theory of planned
behaviour to urban water restrictions using structural equation modelling. Applied
Economics, 49(14), 1426–1439.
Ajayi, & Oyedele (2018). Critical design factors for minimising waste in construction Cullen, M. (2014). Collaboration in libraries and learning environments. The Australian
projects: A structural equation modelling approach. Resources, Conservation and Library Journal, 63(4), 325–326.
Recycling, 137, 302–313. Dandagi, S., Bhushi, U., Bagodi, V., & Sinha, D. (2016). Strategic management of technical
Alessandrini, Lançon, Fond, Faget-Agius, Richieri, Faugere, ... Boyer (2016). A structural university: Structural equation modelling approach. Journal of Modelling in
equation modelling approach to explore the determinants of quality of life in schi- Management, 11(1), 75–90.
zophrenia. Schizophrenia Research, 171(1–3), 27–34. Deed, C., & Alterator, S. (2017). Informal learning spaces and their impact on learning in
American College Personnel Association & National Association of Student Personnel higher education: Framing new narratives of participation. Journal of Learning Spaces,
Administrators (ACPA & NASPA) (2004). Learning reconsidered: A campus-Wide focus 6(3), 54–58.
on the student experience. (Washington, DC: author). Dzeng, R., Lin, K., & Wang, P. (2014). Building a construction procurement negotiation
Banbury, S., & Berry, D. (1998). Disruption of office-related tasks by speech and office training game model: Learning experiences and outcomes. British Journal of
noise. British Journal of Psychology, 89, 499–517. Educational Technology, 45(6), 1115–1135.
Bilandzic, M., & Foth, M. (2014). Learning beyond books–strategies for ambient media to Ellison, W. (2016). Designing the learning spaces of a university library. (2016). New
improve libraries and collaboration spaces as interfaces for social learning. Library World, 117(5/6), 294–307.
Multimedia Tools and Applications, 71(1), 77–95. Everaert, P., Opdecam, E., & Maussen, S. (2017). The relationship between motivation,
Booth, A., Walton, G., Fraser, V., Urquhart, C., & Van Loo, J. (2002). Libraries without learning approaches, academic performance and time spent. Accounting Education,
walls still need windows. Health Information and Libraries Journal, 19(3), 181–184. 26(1), 78–107.
Buss, S. (2016). Do we still need reference services in the age of Google and Wikipedia? Fialkoff, F. (2011). We don't need Google. Library Journal, 136(7) (N/a).
The Reference Librarian, 57(4), 1–7. Goong, H., Xu, L., & Li, C. (2016). Effects of work–family–school role conflicts and role-
Cantwell, K. (2013). Collaboration in libraries and learning environments. Australian related social support on burnout in registered nurses: A structural equation model-
Academic and Research Libraries, 44(3), 182–183. ling approach. Journal of Advanced Nursing, 72(11), 2762–2772.
Carlson, S. (2002). Do libraries really need books? The Chronicle of Higher Education, Gray, L., & Tracy, J. (2011). Do school libraries still need printed books? Junior Scholastic,
48(44), A31–A33. 113(10), 9.
Castilla, Llinares, Bravo, & Blanca (2017). Subjective assessment of university classroom Groth, A. (2007). Energy efficiency building design guidelines for Botswana. Bostwana:
environment. Building and Environment, 122, 72–81. Department of Energy, Ministry of Mineral, Energy and Water Resources.
Castro, M., Baltar, V., Marchioni, D., & Fisberg, R. (2016). Examining associations be- Jolly, L., & White, S. (2016). Communication, collaboration, and enhancing the learning
tween dietary patterns and metabolic CVD risk factors: A novel use of structural experience: Developing a collaborative virtual enquiry service in university libraries
equation modelling. British Journal of Nutrition, 115(9), 1586–1597. in the north of England. New Review of Academic Librarianship, 22(2–3), 176–191
Chesbrough, R. (2011). College students and service: A mixed methods exploration of 176-3.
motivations, choices, and learning outcomes. Journal of College Student Development, Jouhari, Z., Haghani, F., & Changiz, T. (2015). Factors affecting self-regulated learning in
52(6), 687–705. medical students: A qualitative study. Medical Education Online, 20(1), 1–8.
Choi, S. M., Guerin, D. A., Kim, H.-Y., Brigham, J. K., & Bauer, T. (2014). Indoor en- Ketutwijaya, I. (2012). Word effect of temperature, the lighting, workload, noise against
vironmental quality of classrooms and student outcomes: A path analysis approach. eye fatigue, general fatigue and stress affect learning outcomes the student computer
Journal of Learning Spaces, 2(2), 2014. users. International Journal of Computer Applications, 58(5), 26–30.
Cois, A., & Ehrlich, R. (2014). Analysing the socioeconomic determinants of hypertension Lee, M. C., Mui, K. W., Wong, L. T., Chan, W. Y., Lee, E. W. M., & Cheung, C. T. (2012).
in South Africa: A structural equation modelling approach. BMC Public Health, 14(1), Student learning performance and indoor environmental quality (IEQ) in air-
414.

736
L.H. Li et al. The Journal of Academic Librarianship 44 (2018) 724–737

conditioned university teaching rooms. Building and Environment, 49(March 2012), New York: Basic Books.
238–244. Rapp, D. (2011). Survey: Online libraries have “greatest impact”. Library Journal,
Lewis, Rowland, Tran, Solomon, Patton, Catalano, & Toumbourou (2017). Adolescent 136(5), 20.
depressive symptoms in India, Australia and USA: Exploratory structural equation Ro, H., & Knight, D. (2016). Gender differences in learning outcomes from the college
modelling of cross-national invariance and predictions by gender and age. Journal of experiences of engineering students. Journal of Engineering Education, 105(3),
Affective Disorders, 212(C), 150–159. 478–507.
Little, G. (2011). The revolution will be streamed online: Academic libraries and video. Row Farr, J. (2017). A place free of judgement – A teenage takeover of libraries (‘Why Are
The Journal of Academic Librarianship, 37(1), 70–72. You Here Teenager?’). Performance Research, 22(1), 91–92.
Liu, R., & Chang, K. (2014). The causal model of the freshman year characteristics, Rudzioniene, J. (2014). Collaboration in libraries and learning environments. (2014). The
campus experiences and learning outcomes for college students. Procedia - Social and Electronic Library, 32(6), 923–924.
Behavioral Sciences, 116, 1383–1388. Stojanovski, J. (2013). Higher education and research libraries: Why do we need them
Martínez-López, F., Gázquez-Abad, J., & Sousa, C. (2013). Structural equation modelling more than ever? Kemija u Industriji, 62(11), 452–455.
in marketing and business research. European Journal of Marketing, 47(1/2), 115–152. Tevaniemi, J., Poutanen, J., & Lähdemäki, R. (2015). Library as a partner in co-designing
Mckay, Perry, & Harvey (2016). The factorial validity and reliability of three versions of learning spaces: A case study at Tampere University of Technology, Finland. New
the aggression questionnaire using confirmatory factor analysis and exploratory Review of Academic Librarianship, 21(3), 304–324.
structural equation modelling. Personality and Individual Differences, 90, 12–15. Thirupathi, R., & Vinodh, S. (2016). Application of interpretive structural modelling and
Mitchell, Anderson, & Ferguson (2017). A predictive model of days from infection to structural equation modelling for analysis of sustainable manufacturing factors in
discharge in patients with healthcare-associated urinary tract infections: A structural Indian automotive component sector. International Journal of Production Research,
equation modelling approach. Journal of Hospital Infection, 97(3), 282–287. 54(22), 6661–6682.
Nicholas, D., Watkinson, A., Rowlands, I., & Jubb, M. (2011). Social media, academic Thornton-Verma, H. (2012). Reaching the Wikipedia Generation. Library Journal, 137(7)
research and the role of university libraries. The Journal of Academic Librarianship, (N/a).
37(5), 373–375. Van Scoyoc, A., & Cason, C. (2006). The electronic academic library: Undergraduate
Odling-Smee, L. (2007). Online resources threaten livelihood of libraries. Nature, research behavior in a library without books. Portal: Libraries and the Academy, 6(1),
446(7139), 958–959. 47–58.
Özütürk, G., & Hürsen, Ç. (2014). Determination of university students motivation in EFL Verešová, M. (2014). The role of learning styles and learning outcomes of university
classroom. Procedia - Social and Behavioral Sciences, 116(C), 7–12. students from the aspect of learning environment in a study subject “primary drug
Pal Pandi, A., Paranitharan, K., & Jeyathilagar, D. (2018). Implementation of IEQMS prevention”. Procedia - Social and Behavioral Sciences, 116, 1501–1508.
model in engineering educational institutions – A structural equation modelling ap- Zhang, Z., Zhang, S., Zhang, F., Yu, J., & Zhu, Y. (2016). Validity of sustainability fra-
proach. Total Quality Management & Business Excellence, 29(1–2), 29–57. mework for China's mining cities – A structural equation modelling approach. Applied
Palfrey, J. (2015). BiblioTech: Why libraries matter more than ever in the age of Google. 2015. Economics, 48(48), 4585–4605.

737