SSCI-Q2-ImpactofSocialMediaUse

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/362607055
Impact of social media use on critical thinking ability of university students
Article in Library Hi Tech · August 2022

DOI: 10.1108/LHT-11-2021-0393
CITATIONS READS
7 10,379
5 authors, including:
Gege Fang
Tsinghua University
2 PUBLICATIONS 9 CITATIONS
SEE PROFILE
All content following this page was uploaded by Gege Fang on 24 August 2022.
The user has requested enhancement of the downloaded file.

The current issue and full text archive of this journal is available on Emerald Insight at:
https://www.emerald.com/insight/0737-8831.htm
Impact of social media use on Impact of

social media
critical thinking ability of use
university students
Li Cheng
School of Public Administration and Communication,
Guilin University of Technology, Guilin, China Received 13 November 2021
Revised 18 February 2022
Gege Fang 11 May 2022
24 June 2022
School of Journalism and Communication, Tsinghua University, Beijing, China and Accepted 24 June 2022
Faculty of Arts and Social Sciences, National University of Singapore, Singapore
Xiaoxue Zhang and Yuxiang Lv
School of Journalism and Communication, Tsinghua University, Beijing, China, and
Lingxuan Liu
School of Journalism and Communication, Sichuan International Studies University,
Chongqing, China
Abstract
Purpose – This research aims to discover the relationship between social media usage (SMU) and the critical
thinking ability (CTA) of university students, and to answer the question that whether social media dependence
(SMD) affects the development of CTA, and thus providing a reference for the social media access strategy of
academic libraries from the perspective of media information literacy.
Design/methodology/approach – The research data were collected via 300 valid questionnaires whose
respondents are students from three universities in China. Multistage stratified cluster sampling method was
used to select the respondents, which guarantees statistical representativeness. A pre-test was conducted to
ensure the validity of the questionnaire.
Findings – It is shown that the total score of CTA and the six sub-dimensions are significantly positively
correlated with SMU, but strongly negatively correlated with SMD. Based on the mediating effect testing, it is
discovered that the degree of SMD can affect the promoting relations between the usage intensity of social
media (UISM) and CTA. Clearly, SMU is a double-edged sword. While it narrows the digital gap in terms of
accessibility, it widens the digital gap in terms of usage.
Originality/value – The differences in SMU have a significant impact on the development of CTA of
university students. This inspires us to consider the ability of “using social media in a balanced way” as an
important evaluation and training direction when inquiring media literacy. As social media is becoming a
critical channel in cultivating individual’s thinking skills, it is highly suggested that the amount of time spent
on reading fragmented information on the internet should be controlled.
Keywords Critical thinking ability, University students, Social media usage, Social media dependence,
Digital gap
Paper type Technical paper
1. Introduction
Critical thinking is an important learning and innovative ability that students must develop.
It includes basic skills, such as analysis, reasoning, induction, deduction, judgment, and
evaluation (Lai, 2011). High-level critical thinking ability (CTA) can help students develop
communication and leadership skills, cultivate team spirit, and inspire multi-cultural and
inter-cultural awareness (Manasijevic et al., 2016). CTA is typically considered as a vital
indicator of quality education (Ay et al., 2015). Library Hi Tech
Social media has become a part of daily living, especially for the young generation. People © Emerald Publishing Limited
0737-8831
can now access all types of information from across the globe via social media, such as DOI 10.1108/LHT-11-2021-0393
LHT Facebook, Twitter, Yahoo, Telegram, Imo, Skype, Messenger, WhatsApp, and YouTube.
Social media has also become a research focus in the field of library science. At the beginning
of the 21st century, Han and Liu (2009) examined data on Chinese academic libraries and
found that the overall penetration and diversity of the Web 2.0 network in Chinese libraries
were low. Similarly, Linh (2008) conducted a study using Australian libraries as the
observation objects and found that the overall use of Web 2.0 products was minimal and
underdeveloped. Previous studies related to digital libraries have critically analysed the
value, purpose, and practical use of social media. However, the library science literature lacks
sufficient studies on how social media practices can promote marketing, public relations, and
learning (Young and Rossmann, 2015). So far, most libraries that have adopted social
networks mainly use them for marketing and information announcement, but the extensive
and two-sided impact of social media use (SMU) is yet to be fully explored (Forcier et al., 2013).
SMU can improve the CTA of users. According to Black (2005), online discussions have
the potential to stir students’ inquiry and create a context for collaborative learning,
promoting both reflection and critical thinking (Maurino, 2006-2007). Using Facebook and
WhatsApp can enhance students’ CTA through analysing, evaluating, applying, generating,
and expressing ideas (Kustijono and Zuhri, 2018).
However, students’ academic achievement has a significantly negative correlation with
SMU (Ndaku, 2013) and smartphone usage (Lepp et al., 2015; Samaha and Hawi, 2016).
Millennials are considered to possess low CTA because they are more inclined to passively
accepting information from social media (Menichelli and Braccini, 2020). Bauerlein (2011)
noted that the social media generation has become the least intelligent generation because
they spend far more time on social media than on studying and reading. Consequently, their
knowledge level, reading ability, sense of social responsibility, and international
competitiveness are much inferior to those of previous generations.
Therefore, this research focused on the relationship between SMU (at different levels) and
the CTA of university students. Meanwhile, regarding the construction of digital libraries,
this study also provides empirical data on the possible two-sided impact of social media
introduction as well as specific recommendations for preventing multimedia devices in the
library from affecting learning efficiency.
2. Literature review
2.1 Necessity of critical thinking ability for students
Previous studies have shown that a correlation exists between CTA and students’ academic
achievement. Watson and Glaser (1980) found that CTA is associated with Grade Point
Average (GPA) and the grades that teachers assign to students. Facione (1997) reported a
statistically significant positive correlation between CTA and GPA, ranging between 0.235
and 0.596. A weak but consistent relationship was also found between CTA (dispositions) and
GPA (Bers et al., 1996; Giancarlo and Facione, 2001) among college and university students.
Fong et al. (2017) found that CTA was moderately related to the success of community college
students while the relationship between students’ achievement and CTA (as skills or
dispositions) was constant for grades and individual test outcomes. Students’ CTA was more
strongly associated with long-term than short term outcomes. Ma et al. (2021) examined the
relationship between CTA and student’s academic achievement in science in the context of
Chinese physical education. With the same control, a statistically significant relationship was
found between students’ CTA and their achievement in physical education.
However, some studies have reported that no relationship exists between students’ CTA
and their academic achievements. In a distant-learning class, Puzziferro (2006) successful and
under-achieving students in terms of their course grades and CTA and found no significant
difference between both groups. Similarly, Singleton-Williams (2009) found a non-significant
correlation between students’ CTA and their final grades in an introductory class of computer
application. Shirazi and Heidari (2019) studied the relationship between CTA and the Impact of
academic achievement of nursing students and found no correlation between CTA and social media
academic achievement, whereas Thompson (2009) and Foust (2008) found negative effects.
use
2.2 The positive impact of SMU on CTA
Typically, the usage intensity of social media (UISM) does not emphasise inappropriate SMU
but only measures SMU frequency, time spent on different social media platforms, and the
subjective feelings of social media users. To understand user behavior on social media,
researchers have proposed and validated numerous scales, most of which can be mainly
categorised into two groups. The first group focuses on the intensity of social media use while
the second group focuses on the addictive aspects of social media use (Mieczkowski et al.,
2020). Using different word choice in the questionnaires of these groups may yield different
results (Jenkins-Guarnieri et al., 2013). Therefore, this study differentiated between the
concepts of usage intensity and media dependence.
Previous studies have focused on the relationship between SMU and CTA, and some have
reported that differences in social media activities can result in differences in self-
construction, affecting an individual’s way of thinking. Thus, social media activities influence
individuals’ self-constructions, thereby influencing their relative reliance on cognitive and
affective thinking styles (Xu et al., 2019). Sumanti et al. (2019) investigated the usage of a
geography-related group on Facebook and found that social media can help develop CTA;
Facebook and WhatsApp may also help to improve students’ CTA (Kustijono and Zuhri,
2018). Reciprocal peer tutoring via Facebook can also enhance students’ CTA during the
learning process (Zulkifli et al., 2020).
The differences in CTA development and UISM reflect the gap in digital technology
usage, further pointing to a knowledge gap. From the perspective of the third digital divide,
which is the knowledge gap on the internet, Internet usage influences personal knowledge
acquisition. For example, individuals who use the internet for political purposes tend to be
more knowledgeable about presidential candidates (Wei and Zhang, 2006). Tien and Fu
(2008) used different statistical methods, such as multiple regression and generalised
ordered logit models, to investigate the digital divide among 3,083 freshmen from 12
universities in Taiwan. The study showed that computer literacy is an indicator of
academic achievement, as students who possessed vast computer software knowledge
performed better academically. Kuh and Vesper (2001) investigated the relationship
between the impact of information technology usage and the ideal academic outcomes of
125,000 university students in the US. They found that the more familiar the students were
with computers, the more capable they were to develop important skills and abilities,
including analytical, logical, and comprehensive thinking abilities as well as
conceptualisation and social skills.
2.3 The negative impact of social media dependency on CTA

Nowadays, university students generally feel lonely, and social media tends to relieve this
emotion through social inclusion. Furthermore, because of the need for personal
development, SMU may provide university students with human capital that they cannot
achieve in real life (Kim, 2014). SMU helps university students to find satisfaction through
their social networks, thereby encouraging them to use social media more deeply (Wang et al.,
2012). Social media platforms are designed for hedonic needs, and thus usually distinguish
and absorb users, thereby increasing users’ participation and dependence (Cao et al., 2018).
This is the typical process of SMD and addiction.
When SMU becomes uncontrollable and compulsive, social media addiction can have
adverse consequences (Andreassen, 2015). Addictive SMU by university students is common
LHT (Hou et al., 2019) and has a negative impact on academic performance (Huang, 2014; Nida,
2018). The use of social media for academic purposes is not an important predictor of
academic performance, but the use of social media for other activities (especially for video
games) and social media multitasking for non-academic purposes are significantly negatively
correlated with academic performance (Lau, 2017).
Students’ GPA is negatively correlated with the number of social media sites they
subscribe to. Although previous studies have suggested that social media sites should be
included in the classroom, such incorporation can have a negative impact on undergraduates’
study habits and their academic performance (Leyrer-Jackson and Wilson, 2017).
The impact of social media addiction on CTA has also been investigated. Studies have
shown that dependence on Weibo (a popular social media in China) is widespread among
university students in China, and such dependence is significantly passive and unintentional
(Liu, 2013). A significantly negative correlation also exists between smartphone addiction
and CTA (Lee and Kim, 2017). For example, some studies have reported a negative
relationship between social media addiction and CTA by adopting sense of achievement and
user emotions as mediating variables (Thomas, 2020).
The negative impact of social media addiction on university students’ academic
performance and CTA is related to certain social media deficiencies, such as information
fragmentation and content overload. Weibo and Twitter are typical platforms on which the
information communicated is highly fragmented. Such information exerts a negative impact
on users’ understanding. When we share information, we increase the burden on our
cognitive resources unknowingly. Consequently, our understanding of information is
impaired, and our subsequent learning is hindered (Jiang et al., 2016). The content distribution
and service model in the mobile era have further aggravated information overload. The
excessive connection between people and the internet will further distract users’ attention
and reduce their thinking and concentration skills. People seem to know more peripherally,
but their deep understanding and thinking are decreased accordingly (Peng, 2019).
In the face of information overload, users of social media tend to escape. Such a
phenomenon is known as social media burnout (Bright et al., 2015). A survey on Chinese
youth groups found that information overload and social overload have a significantly
positive impact on social media burnout (Liu and He, 2021). Social media burnout is a
consequence of multi-dimensional user experience due to the overuse of social media, and it
includes fatigue, annoyance, anger, disappointment, caution, loss of interest, decreased needs,
and decreased motivation (Ravindran et al., 2013). During SMU, users tend to withdraw
themselves from social media participation when they are overwhelmed. Social media
burnout is a negative behaviour of SMU (Bright et al., 2015). An empirical study based on
Chinese youth found that negative comparison, interaction disorder, information overload,
social overload, self-efficacy, and impression management all have significant positive effects
on social media burnout (Liu and He, 2021). Strong social media burnout and retreat are also
observed among young people (Xue and Hong, 2020).
Social media burnout can lead to a decline in academic performance. Malik et al. (2020)
found that social media burnout caused by the instant messaging application – MIM –
significantly contributed to students’ academic burnout. Kirschner and Karpinski (2010)
found a significantly negative correlation between Facebook use and academic performance.
The average GPA of Facebook users was lower, and their average learning time per week was
less than that of non-Facebook users. Using Facebook does not increase the time spent online
but may affect the available learning time (Kirschner and Karpinski, 2010).
2.4 Other variables related to CTA

2.4.1 Parental income (PI) and parental education (PE). Previous studies have shown that
CTA is correlated with parental income and educational level. Huang et al. (2019) found that
the socioeconomic status of the family had a positive albeit limited influence on the general Impact of
self-efficacy and CTA of students in Chinese medical studies. General self-efficacy mediates social media
the effect of the family’s socioeconomic status on CTA and plays a significant role. Karahan
_
and Iskifo glu (2020) found that mothers’ educational level may lead to a strong CTA
use
disposition, whereas fathers’ educational level may lead to the smallest amount of variance in
CTA disposition. Kavenuke et al. (2020) measured CTA from three dimensions: systematicity,
self-confidence, and scepticism. In terms of scepticism, while years of study and mothers’
education accounted for a negative proportion in prediction, fathers’ education accounted for
a positive prediction. Additionally, fathers’ education and occupation significantly predicted
prospective teachers’ systematicity.
2.4.2 Sub-dimensions/sub-variables of CTA. The Delphi project, implemented by Facione
(1990) and supported by the American Philosophical Association, investigated the meaning
of critical thinking and the skills and personalities it comprises. A total of 46 academicians
from different disciplines participated in the project. The project proposed that individuals
with good CTA have inquisitiveness, openness, systematicity, cognitive maturity, and the
tendency to find the truth (Facione, 1990). Researchers have identified the following six sub-
dimensions in the process of measuring CTA.
First, self-confidence refers to the degree of trust a person has in his/her reasoning process
(Wangensteen et al., 2010). In other words, self-confidence refers to a student’s ability to trust
his/her usage of rational and reflective thinking (Kavenuke et al., 2020). Kavenuke et al. (2020)
investigated the CTA of 965 prospective teachers and found that their average self-
confidence score was significantly low. Such a lack of self-confidence comes from the vertical
relationships between children and parents, students and teachers, or students and other
authority figures (Kavenuke, 2018). Confidence in CTA is positively correlated with
argument-driven evaluation but negatively correlated with belief-driven response (Stupple
et al., 2017). Chan et al. (2011) studied the relationship between Chinese students’ cognitive
beliefs and CTA and found that beliefs have a unique impact on the performance of CTA,
surpassing the impact of cognitive ability. Thus, a clear evidence was found that those who
believe in the certainty of knowledge show poor two-sided thinking and are more inclined to
underestimating or ignoring opposing views.
Second, inquisitiveness refers to an individual’s desire to reduce the gap between the
current level of knowledge and the required level of knowledge (Ismail et al., 2019).
Inquisitiveness is a complex psychological tendency related to various functions (Perlovsky
et al., 2010). Ulucinar and Aypay (2016) found that inquisitiveness could predict CTA,
indicating that inquisitiveness plays an important role in the development of CTA. The thirst
for knowledge is a factor that stimulates and arouses the reasoning of a phenomenon.
Third, systematicity refers to the habit of striving to approach problems in a disciplined,
organised, focused and systematic manner (Derrick-Telemaque, 2014; Shin et al., 2015).
Similarly, Mwalongo (2014) asserted that systematicity is related to the skills and
dispositions of being organised and focused on thinking.
Fourth, maturity of judgment or cognitive maturity is the tendency to see problems as a
complex (Facione and Facione, 2014).
Fifth, analyticity is the tendency or habit of being alert to situations, always striving to
anticipate both good and bad potential consequences or outcomes of situations, choices,
proposals, and plans (Facione and Facione, 1992).
Finally, truth-seeking is a desire for the best possible understanding of a situation
(Facione and Facione, 2014). Numerous scholars have empirically shown that changes in
teaching methods and specific thinking training could effectively improve students’ CTA in
certain sub-dimensions, such as truth-seeking, analyticity, systematicity, and self-confidence
(Tiwari et al., 2010; Miri et al., 2007), as well as openness, maturity, and inquisitiveness (Zhang
et al., 2017; Wu and Wu, 2020).
LHT In light of the literature discussed above, this research raises basic hypotheses and
mediating effect models:
H1a. There is a significantly positive correlation between UISM and CTA.
H1b. There is a significantly negative correlation between SMD and CTA.
H2a. There is a significantly positive correlation between PI and CTA.
H2b. There is a significantly positive correlation between PE and CTA.
As CTA contains six sub-dimensions, this research also raises some complimentary
hypotheses below:
H3a. Self-confidence is significantly positively correlated with UISM and significantly
negatively correlated with SMD.
H3b. SMD mediates the relationship between self-confidence and UISM.
H4a. Inquisitiveness is significantly positively correlated with UISM and significantly
H4b. SMD mediates the relationship between inquisitiveness and UISM.
H5a. Systematicity is significantly positively correlated with UISM and significantly
H5b. SMD mediates the relationship between systematicity and UISM.
H6a. Cognitive maturity (CM) is significantly positively correlated with UISM and
significantly negatively correlated with SMD.
H6b. SMD mediates the relationship between cognitive maturity and UISM.
H7a. Analyticity is significantly positively correlated with UISM and significantly
H7b. SMD mediates the relationship between analyticity and UISM.
H8a. Truth-seeking ability is significantly positively correlated with UISM and
significantly negatively correlated with SMD.
H8b. SMD mediates the relationship between truth-seeking ability and UISM.
In summary, this research aimed to explore the relationship between SMU and CTA as well as
that between SMU and the six sub-dimensions (self-confidence, inquisitiveness,
systematicity, cognitive maturity, analyticity, truth-seeking ability) of CTA (Figure 1). The
research also analysed how SMD moderates the relationship between SMU and CTA.
3. Methodology
The questionnaire used for this study included three scales: the scales of CTA, UISM, and
SMD. Demographic control variables such as gender, age, and economic status were also
included. The questionnaire mainly explored the basic status and differences in CTA among
undergraduates in different universities in China. It also incorporated two new variables,
UISM and SMD, to further explain the influence of social media exposure and usage on critical
thinking. The scale of critical thinking refers to the California Critical Thinking Disposition
Inventory (CCTDI) (Facione and Facione, 1992), as well as a simplified Chinese version of the
scale adapted by Yu and Yu (2020) and others. The simplified Chinese version, also known as
Critical Thinking Disposition Inventory-Chinese Version (CTDI-CV), is a revised version of
Impact of
social media
use
Figure 1.
Research framework
CCTDI based on the translation and edition made by Peng et al. (2004). The revised version
further corrects factor stability and semantic expression, making it more robust to the
language habits of Chinese in terms of reliability, factor structure, and explanatory power.
The scale of critical thinking contains a total of 28 items in the 6 sub-dimensions. Self-
confidence represents an individual’s trust in his own rational analysis process; inquisitiveness
represents an individual’s intellectual curiosity and desire for learning; systematicity represents
an individual’s organisation, orderliness, concentration, and diligence; cognitive maturity
represents an individual’s degree of wisdom in making decisions; analyticity represents an
individual’s ability to question and use evidence to solve problems; truth-seeking represents an
individual’s ability to solve problems in a specific situation, including his/her eagerness to seek
the best solution, and his/her honesty and objectivity in the inquiry even when the ultimate
findings are inconsistent with his/her interests or existing perceptions.
WeChat, the infrastructural social media platform in China, was selected as the
representative social media given its large userbase. In January 2021, WeChat founder, Zhang
Xiaolong, disclosed in the WeChat Open Class that 1.09 billion users were using WeChat
every day, establishing WeChat as platform with the highest number of user registrations
and daily active rate in China.
For the measurement of the Scale of UISM, we referred to the questionnaire compiled by
Ellison et al. (2007) and translated and revised by Chen (2015) in Chinese. In the original
questionnaire, the first two questions asked for a self-measurement of the total number of friends
on a social networking site and the average time spent on social networking sites each day. The
rest six questions measured the strength of an individual’s emotional connection with social
networking sites and the degree to which social networking sites integrated into the individual’s
life. These eight questions were reduced to six questions to fit the actual demand of this research.
The scale for social media dependence contains three dimensions: time of usage, emotional
connection strength, and degree of integration of social networking sites into the individual’s
life. This research assumes that increasing SMU may contribute to SMD and addiction. The
scale for this part is adapted from the Mobile Phone Addiction Index (Leung, 2008; Bianchi
and Phillips, 2005). We also refer to the Chinese translation of the MPAI proposed by Chinese
scholars (Jiang et al., 2019). The MPAI comprises 11 questions. All the scales mentioned above
are assessed to have acceptable reliability and structural validity.
The respondents of this research were Chinese university students selected via stratified
sampling. According to the 2021 Chinese university ranking on CNUR.com compiled by the
ABC RANKING consulting agency, we divided the top 100 universities in China into three
categories: A (ranked 1–33), B (ranked 34–67), and C (ranked 68–100). For the ranking offered
LHT by the ABC RANKING consulting agency, the “hierarchical classification” and “segmented
analysis” methods were adopted to construct five first-level indicators (including talent
training, social reputation, academic research, social services, and international exchanges),
and 14 second-level indicators. The evaluation data were acquired from public sources
disclosed on the universities’ official websites or the websites of relevant departments,
ensuring the credibility and authority of the evaluation. One university from each of the three
categories was selected, namely, Tsinghua University (ranked 1), Beijing Foreign Studies
University (ranked 34), and Shenzhen University (ranked 76).
This research randomly sampled students of an elective course from the three universities.
The selected public elective course was open to the whole university and thus included
students from different majors and with different grades and genders. Therefore, the
diversity and representativeness of the samples was guaranteed. From March to May 2021,
with permission, researchers came in person to the classroom of the course and requested
students to fill in the questionnaire with the assistance of the teaching assistant during the 10-
min class break. We distributed 163, 128, and 133 questionnaires to students of the selected
elective course respectively and collected 353 (116 þ 113 þ 124) answered questionnaires in
total, making the response rate 83%. Ten questionnaires were eliminated as the respondents
were from non-target institutions. To ensure that respondents carefully made judgements
about their SMU, we set reverse questions in the questionnaire to check the consistency of the
answers. Thirty-four questionnaires were then eliminated because of inconsistencies in the
answers. The remaining questionnaires were completed between 50 s and 2,000 s, and no
same option was selected for all the questions. Therefore, all the remaining questionnaires
were valid. Overall, a total of 53 invalid questionnaires were excluded, and 300 valid
questionnaires were adopted, with an effective recovery rate of 85%. Meanwhile, z-score
standardisation was performed on the questionnaire data. We used AMOS for data analysis.
AMOS is an analytical software for structural equation modelling and measures basic
features of data, such as variance and covariance. The software encompasses the structural
equations to explore the relationships between variables. To a certain extent, the data
analysis of AMOS overcomes the limitation of large sample conditions, and the results of
Bayesian methods are still relatively stable when the sample is below 200 or even below 100.
4. Results
4.1 Reliability analysis
We conducted a reliability test, which is also known as credibility or consistency test. The
reliability of the results is defined and measured by the scores obtained by individuals on
successive independent tests. The reliability test aims to draw the same conclusions through
repeated tests so as to determine the generality and stability between different project
variables (Cronbach, 1947).
Table 1 presents the results of reliability tests that were performed for the scale of overall
critical thinking and the six sub-dimensions. The Cronbach α values were all above 0.9,
indicating both reliability and validity. The Cronbach α values of the scale of WeChat usage
intensity and the scale of WeChat usage dependence were 0.913 and 0.935, respectively,
showing good reliability and validity.
4.2 Validity analysis

Validity (i.e. effectiveness) refers to the degree to which the required variables can be
accurately measured through measurement means and tools; it shows whether the content to
be measured can be measured (Field, 2012). Validity explains how well the collected data
covers the actual area of investigation (Ghauri and Grønhaug, 2010). It can generally be
divided into several types, such as surface validity, content validity, and construction Impact of
validity. We determined the desirable degree of validity through the KMO value; a KMO social media
value greater than 0.8 is regarded as good validity.
We used factor analysis to conduct information enrichment research (see Table 2). First,
use
we analysed whether the research data were suitable for factor analysis. The table above
shows that the KMO values of the three scales were 0.855, 0.911, and 0.961, respectively (all
greater than 0.6), which makes the data suitable for factor analysis. All the data also passed
the Bartlett sphericity test (p < 0.05), also indicating that the research data were suitable for
factor analysis.
4.3 Analysis of confirmatory factors

The AMOS software was used to calculate the parameters of the model and test the
significance of the indicators. As shown in Table 3 above, the values of GFI, AGFI, and NFI of
the analysis model failed to meet the expected value but were still within the acceptable range.
In addition, other adaptation indicators met the requirements; hence, the model had a good fit.
According to Table 4, the factor loading of the explicit variables in the model were all
greater than 0.5, and their significance (p-values) were less than 0.05. These values indicate
that a significant correlation existed between the explicit and latent variables, and the explicit
variables can explain their corresponding latent variables. Meanwhile, the average variance
extracted (AVE) value of each latent variable was greater than 0.5, and the combined
reliability (CR) value was greater than 0.7, indicating that the scale had good convergence
validity.
Variables have discriminative validity when the coefficient between them is lower than
the square root of their AVE. According to Table 5, it is the case for self-confidence,
Variables Items Cronbach α
Self-confidence 7 0.911
Inquisitiveness 5 0.879
Systematicity 5 0.897
Cognitive maturity 5 0.89
Analyticity 3 0.838
Truth-seeking 3 0.844
Overall critical thinking 28 0.89 Table 1.
Overall WeChat usage intensity 6 0.913 Reliability analysis
Overall WeChat dependence 11 0.935 (n 5 300)
CTA KMO value 0.855

Chi-squared approximation 5068.059
df 378
p Value 0
UISM KMO value 0.911
Chi-squared approximation 1141.846
df 15
p Value 0
SMD KMO Value 0.961
Chi-squared approximation 2075.03 Table 2.
df 55 Tests of KMO and
p Value <0.001 Bartlett
LHT Fit index Value of expectation Actual value
Absolute goodness of fit index

CMIN/DF <3 1.393
GFI >0.9 0.844
AGFI >0.9 0.824
RMSEA <0.08 0.036
Comparative fit index
NFI >0.9 0.864
IFI >0.9 0.958
Table 3. TLI >0.9 0.954
Fit index CFI >0.9 0.957
inquisitiveness, systematicity, cognitive maturity, analyticity, truth-seeking, WeChat usage

intensity, and WeChat usage dependence. In other words, the discriminative validity of each
latent variable was good.
4.4 Correlation analysis

Correlation may be described as the degree of association between two variables. Pearson
(1896) developed Pearson correlation coefficient R to measure the correlation between
variables. The linear relationship between two variables can be measured by normative
correlation analysis (Asuero et al., 2006). Table 6 presents the impact of the correlation
analysis on the relationship between the total CTA score and 10 variables, namely, parental
income, parental educational level, self-confidence, inquisitiveness, systematicity, cognitive
maturity, analyticity, truth-seeking, WeChat usage intensity, and WeChat usage dependence.
The Pearson correlation analysis shows that the total CTA score was significantly
correlated with parental income (r 5 0.536, p < 0.001), parental educational level (r 5 0.388,
p < 0.001), self-confidence (r 5 0.672, p < 0.001), inquisitiveness (r 5 0.580, p < 0.001),
systematicity (r 5 0.643, p < 0.001), cognitive maturity (r 5 0.635, p < 0.001), analyticity
(r 5 0.513, p < 0.001), truth-seeking (r 5 0.459, p < 0.001), WeChat usage intensity (r 5 0.524,
p < 0.001), and WeChat usage dependence (r 5 0.584, p < 0.001). Specifically, in addition to
the negative correlation between the total CTA score and WeChat usage dependence, all other
correlations were positive.
4.5 Regression analysis

Regression analysis is a statistical method used for testing the quantitative relationship
between two or more variables. It measures the relationship between the specified and the
corresponding variables (Asuero et al., 2006). For linear regression analysis in the study, we
set parental income, parental educational level, WeChat usage intensity, and WeChat usage
dependence as independent variables, and we set the total CTA score as the dependent
variable (see Table 7).
The formula for the analysis is: total CTA score 5 0.000 þ 0.296 3 parental
income þ 0.110 3 parental educational level þ 0.274 3 usage intensity of WeChat
–0.354 3 WeChat usage dependence. The R value of the model was 0.552, indicating that
parental income, parental educational level, WeChat usage intensity, and WeChat usage
dependence accounted for 55.2% of the total CTA score in the whole process.
The model passed the F test (F 5 90.778, p 5 0.000 < 0.05), indicating that at least one of
the parental income, parental educational level, WeChat usage intensity, and WeChat usage
dependence influenced the total CTA score. In addition, the test for the multicollinearity of the
Latent variables Explicit variable Coef. SE z p Factor loading AVE CR
Impact of
social media
Self-confidence SC1 1 – – – 0.734 0.626 0.918 use
SC2 1.67 0.102 16.31 0 0.93
SC3 0.944 0.078 12.179 0 0.704
SC4 1.048 0.079 13.223 0 0.761
SC5 1.074 0.083 12.961 0 0.747
SC6 1.075 0.08 13.499 0 0.776
SC7 1.028 0.079 13.096 0 0.754
Inquisitiveness CU1 1 – – – 0.719 0.633 0.892
CU2 1.684 0.112 14.972 0 0.923
CU3 1.002 0.084 11.923 0 0.716
CU4 1.107 0.088 12.636 0 0.758
CU5 1.068 0.085 12.522 0 0.751
Systematicity SA1 1 – – – 0.79 0.676 0.909
SA2 1.501 0.083 18.195 0 0.93
SA3 0.909 0.064 14.271 0 0.76
SA4 0.936 0.065 14.404 0 0.766
SA5 0.917 0.064 14.296 0 0.761
Cogitative maturity CM1 1 – – – 0.757 0.666 0.904
CM2 1.652 0.097 17.065 0 0.941
CM3 1.072 0.076 14.114 0 0.783
CM4 1.01 0.074 13.688 0 0.763
CM5 0.94 0.074 12.698 0 0.714
Analyticity SOA1 1 – – – 0.862 0.658 0.849
SOA2 0.714 0.053 13.548 0 0.771
SOA3 0.753 0.056 13.542 0 0.77
Truth–seeking FTT1 1 – – – 0.893 0.685 0.863
FTT2 0.703 0.05 14.052 0 0.767
FTT3 0.653 0.046 14.171 0 0.774
UISM WUI1 1 – – – 0.912 0.656 0.918
WUI2 0.667 0.04 16.58 0 0.753
WUI3 0.732 0.042 17.437 0 0.774
WUI4 0.713 0.039 18.138 0 0.791
WUI5 0.725 0.039 18.546 0 0.801
WUI6 0.691 0.04 17.117 0 0.766
SMD WUD1 1 – – – 0.9 0.577 0.937
WUD10 0.714 0.047 15.114 0 0.709
WUD11 0.645 0.048 13.354 0 0.654
WUD2 0.75 0.041 18.254 0 0.791
WUD3 0.695 0.041 17.08 0 0.762
WUD4 0.736 0.042 17.436 0 0.771
WUD5 0.688 0.04 17.197 0 0.765
WUD6 0.708 0.04 17.741 0 0.779
WUD7 0.734 0.048 15.445 0 0.718 Table 4.
WUD8 0.764 0.046 16.518 0 0.748 Coefficient table for
WUD9 0.701 0.047 15.013 0 0.706 factor loading
model showed that all the VIF values in the model were less than 5, indicating that no
collinearity problem was encountered. The D-W value was approximately 2, indicating that
the model lacked autocorrelation while no correlation existed between the sample data. Thus,
the model was well-established. Further specific analysis results are described below.
Among the control variables, the regression coefficient of parental income was 0.296
(t 5 6.680, p 5 0.000 < 0.01); thus, parental income had a significantly positive influence on
the total CTA score. The regression coefficient of parental educational level was 0.110
LHT
Table 5.
Discriminative validity
Self-confidence Inquisitiveness Systematicity CM Analyticity Truth-seeking UISM SMD
Self-confidence 0.791
Inquisitiveness 0.247 0.796
Systematicity 0.310 0.224 0.822
CM 0.195 0.215 0.247 0.816
Analyticity 0.197 0.163 0.206 0.360 0.811
Truth-seeking 0.117 0.160 0.191 0.285 0.169 0.827
UISM 0.326 0.297 0.325 0.383 0.276 0.235 0.810
SMD 0.456 0.328 0.294 0.396 0.263 0.281 0.382 0.760
Note(s): The diagonal numbers in italic are the AVE square root values
1 2 3 4 5 6 7 8 9 10 11
1 1
2 0.394*** 1
3 0.367*** 0.209*** 1
4 0.233*** 0.250*** 0.247*** 1
5 0.311*** 0.164** 0.310*** 0.224*** 1
6 0.379*** 0.287*** 0.195*** 0.215*** 0.247*** 1
7 0.289*** 0.264*** 0.197*** 0.163** 0.206*** 0.360*** 1
8 0.330*** 0.253*** 0.117* 0.160** 0.191*** 0.285*** 0.169** 1
9 0.536*** 0.388*** 0.672*** 0.580*** 0.643*** 0.635*** 0.513*** 0.459*** 1
10 0.291*** 0.260*** 0.326*** 0.297*** 0.325*** 0.383*** 0.276*** 0.235*** 0.524*** 1
11 0.330*** 0.253*** 0.456*** 0.328*** 0.294*** 0.396*** 0.263*** 0.281*** 0.584*** 0.382*** 1
Note(s): 1 Parental income; 2 Parental educational level; 3–11 Self-confidence, inquisitiveness, systematicity, cognitive maturity, analyticity, truth-seeking, critical
thinking ability, social media usage intensity, and WeChat usage dependence, respectively; *p < 0.05; **p < 0.01; ***p < 0.001
use
social media
Impact of
Table 6.
Pearson correlation
LHT
Table 7.
regression
Results of linear
analysis (n 5 300)
Non-standardised coefficient Standardised coefficient
B SE Beta t p VIF R2 Adjusted R2 F
Constant 0.000 0.039 – 0.000 1.000 – 0.552 0.546 F 5 90.778, p 5 0.000

PI 0.296 0.044 0.296 6.680 0.000*** 1.293
PE 0.110 0.043 0.110 2.555 0.011* 1.228
UISM 0.274 0.043 0.274 6.320 0.000*** 1.235
SMD 0.354 0.044 0.354 8.094 0.000*** 1.259
Note(s): DV: CTA (total)
D-W Value: 1.818
*p < 0.05 **p < 0.01 ***p < 0.001; SE: Standard error
(t 5 2.555, p 5 0.011 < 0.05); thus, parental educational level had a significantly positive Impact of
influence on the total CTA score. Among the independent variables, the regression coefficient social media
of WeChat usage intensity was 0.274 (t 5 6.320, p 5 0.000 < 0.01), which shows that WeChat
usage intensity had a significantly positive impact on the total CTA score. Among the
use
mediating variables, the regression coefficient of WeChat usage dependence was 0.354
(t 5 8.094, p 5 0.000 < 0.01), showing that WeChat usage dependence had a significantly
negative impact on the total CTA score.
The above regression analysis results support hypotheses 1a, 1b, 2a, 2b, 3a, 4a, 5a, 6a, 7a,
and 8a. A higher education level and better economic class status of a family are beneficial for
the development of children’s CTA. SMU, as an essential means of acquiring knowledge in
the current society, also has a positive effect on the development of university students’ CTA.
Nonetheless, dependence on social media would negatively impact the CTA of college
students.
4.6 Moderating effect tests

The moderation model tests whether the prediction of a dependent variable, Y, from an
independent variable, X, differs across levels of a third variable, Z (Fairchild and MacKinnon,
2009, p. 4). A moderator variable specifies when or under what conditions a predictor variable
influences a dependent variable (Baron and Kenny, 1986; Holmbeck, 1997). It may also reduce
or enhance the direction of the relationship between a predictor variable and a dependent
variable (Lindley and Walker, 1993). This research first analysed how WeChat usage
intensity (as an independent variable) and WeChat usage dependence (as a moderating
variable) affect the total CTA score (as a dependent variable). Then, the total CTA score was
divided into six sub-dimensions (self-confidence, systematicity, analyticity, inquisitiveness,
cognitive maturity, truth-seeking), and the moderating effect was tested.
The moderating effect was divided into three models. As shown in Table 8, Model 1
includes one independent variable (WeChat usage intensity) and two control variables
(parental income and parental educational level). For Model 2, the moderating variable is
WeChat usage dependence, which was added on the basis of Model 1. For Model 3, the
interaction terms (product of independent variables and moderating variables) were added on
the basis of Model 2.
The purpose of Model 1 was to explore the influence of the independent variable (UISM) on
the dependent variable (the total CTA score) without considering the interference of the
moderating variable (SMD). The independent variable (UISM) was significant (t 5 8.316,
Model 1 Model 2 Model 3
PI 0.369*** (7.709) 0.296*** (6.680) 0.308*** (7.198)

PE 0.144** (3.026) 0.110* (2.555) 0.087* (2.075)
UISM 0.379*** (8.316) 0.274*** (6.320) 0.274*** (6.563)
SMD 0.354*** (8.094) 0.359*** (8.508)
UISM* SMD 0.168*** (4.796)
Sample 300 300 300
R2 0.452 0.552 0.584
Adjusted R2 0.447 0.546 0.577
F Value F 5 81.446, p 5 0.000 F 5 90.778, p 5 0.000 F 5 82.640, p 5 0.000
ΔR 2
0.452 0.100 0.033
ΔF Value F 5 81.446, p 5 0.000 F 5 65.517, p 5 0.000 F 5 23.004, p 5 0.000 Table 8.
Note(s): *p < 0.05 **p < 0.01 ***p < 0.001; the numbers in the brackets are T values Results of moderating
Dependent variable: CTA (total) effects
LHT p 5 0.000 < 0.05), meaning WeChat usage intensity has a significantly positive effect on the
total CTA score. Meanwhile, there was a significantly negative correlation between the total
CTA score and the interaction of WeChat usage intensity and WeChat usage dependence
(t 5 4.796, p 5 0.000 < 0.05). Thus, when the usage intensity of WeChat influences the total
CTA score, the fluctuation of the moderating variable (WeChat usage dependence) may have
a significant influence. The horizontal axis shows the difference of UISM, and the vertical axis
shows the CTA (Figure 2); the influence of UISM on the CTA is stronger at the high level of
SMD than at the low level of SMD. Thus, the result of the moderating effect test supports
Hypothesis 3b.
As shown in Table 9, when the interference of the moderating variable (WeChat usage
dependence) was excluded, the independent variable (WeChat usage intensity) had a
significant influence (t 5 4.220, p 5 0.000 < 0.05) on the dependent variable (self-confidence).
Meanwhile, WeChat usage intensity and WeChat usage dependence were significantly
correlated (t 5 3.222, p 5 0.001 < 0.05), indicating that when WeChat usage intensity
influences self-confidence, the fluctuation of the adjustment variable (WeChat usage
dependence) may have a significant influence. The horizontal axis shows the difference of
SMD (low)
5
SMD (high)
4
Critical thinking (CTA)
2
Figure 2.
Moderating effect test
results 1
UISM (low) UISM (high)
PI 0.284*** (4.877) 0.216*** (3.829) 0.226*** (4.075)

PE 0.036 (0.625) 0.005 (0.089) 0.015 (0.283)
UISM 0.234*** (4.220) 0.135* (2.457) 0.136* (2.504)
SMD 0.332*** (5.976) 0.337*** (6.150)
UISM 3 SMD 0.147** (3.222)
Sample 300 300 300
R2 0.188 0.276 0.301
Adjusted R2 0.180 0.266 0.289
F Value F 5 22.877, p 5 0.000 F 5 28.099, p 5 0.000 F 5 25.271, p 5 0.000
ΔR 2
0.188 0.088 0.025
Table 9. ΔF Value F 5 22.877, p 5 0.000 F 5 35.717, p 5 0.000 F 5 10.384, p 5 0.001
Results of moderating Note(s): *p < 0.05 **p < 0.01 ***p < 0.001
effect tests DV: Self-confidence
UISM, and the vertical axis shows the difference of self-confidence (Figure 3); the influence of Impact of
UISM on self-confidence is stronger at the high level of SMD than at the low level of SMD. social media
dependence) was excluded, the independent variable (usage intensity of WeChat) had a
use
significant influence (t 5 3.953, p 5 0.000 < 0.05) on the dependent variable (inquisitiveness).
Meanwhile, the interaction of WeChat usage intensity and WeChat usage dependence were
significantly correlated (t 5 2.346, p 5 0.020 < 0.05), indicating that when usage intensity of
WeChat influences inquisitiveness, the fluctuation of the adjustment variable (WeChat usage
dependence) may have a significant influence. The horizontal axis shows the difference of
UISM, and the vertical axis shows inquisitiveness (Figure 4); the influence of UISM on
inquisitiveness is stronger at the low level of SMD than at the high level of SMD. Thus, the
results of the moderating effect test supports Hypothesis 4b.
significant influence (t 5 4.521, p 5 0.000 < 0.05) on the dependent variable (systematicity).
However, the WeChat usage intensity and WeChat usage dependence were not significantly
correlated (t 5 1.159, p 5 0.247 > 0.05). Combining the indication from Model 1 (that X has
an influence on Y), we assumed that when WeChat usage intensity influences systematicity,
SMD (low)
5
SMD (high)
4
Self-confidence
2
Figure 3.
Moderating effect
1 results
UISM (low) UISM (high)
PI 0.109 (1.809) 0.065 (1.082) 0.074 (1.228)

PE 0.147* (2.465) 0.128* (2.164) 0.112 (1.894)
UISM 0.227*** (3.953) 0.164** (2.776) 0.164** (2.803)
SMD 0.212*** (3.552) 0.215*** (3.637)
UISM 3 *SMD 0.116* (2.346)
Sample 300 300 300
R2 0.130 0.165 0.181
Adjusted R2 0.121 0.154 0.167
F Value F 5 14.719, p 5 0.000 F 5 14.626, p 5 0.000 F 5 12.980, p 5 0.000
ΔR 2
0.130 0.036 0.015
Note(s): p < 0.05 **p < 0.01 ***p < 0.001; t values in parentheses Results of moderating
DV: Inquisitiveness effects
LHT SMD(low)
5
SMD(high)
Inquisitiveness 3
2
Figure 4.
Moderating effect
results 1
UISM(low) UISM(high)
PI 0.234*** (3.937) 0.203*** (3.379) 0.207*** (3.444)

PE 0.006 (0.096) 0.008 (0.141) 0.016 (0.275)
UISM 0.256*** (4.521) 0.212*** (3.598) 0.212*** (3.603)
SMD 0.149* (2.504) 0.150* (2.533)
UISM 3 SMD 0.057 (1.159)
Sample 300 300 300
R2 0.157 0.174 0.178
2
Adjusted R 0.148 0.163 0.164
F Value F 5 18.345, p 5 0.000 F 5 15.570, p 5 0.000 F 5 12.740, p 5 0.000
ΔR2 0.157 0.018 0.004
Results of moderating Note(s): *p < 0.05 **p < 0.01 ***p < 0.001; t values in parentheses
effects DV: Systematicity
the fluctuation of the adjustment variable (WeChat usage dependence) may not have a
significant influence. In other words, WeChat usage dependence will not affect the
relationship between WeChat usage intensity and systematicity. Thus, the result of the
moderating effect test did not support Hypothesis 5b.
As shown in Table 12, when the interference of the moderating variable (usage
dependence of WeChat) was excluded, the independent variable (WeChat usage intensity)
had a significant influence (t 5 5.188, p 5 0.000 < 0.05) on the dependent variable
(cognitive maturity). Meanwhile, WeChat usage intensity and WeChat usage dependence were
significantly correlated (t 5 2.916, p 5 0.004 < 0.05), indicating that when usage intensity of
WeChat influences cognitive maturity, the fluctuation of the adjustment variable (WeChat
usage dependence) may have a significant influence. The horizontal axis shows the difference
of UISM, and the vertical axis shows inquisitiveness (Figure 5); the influence of UISM on
inquisitiveness is stronger at the low level of SMD than at the high level of SMD. Thus, the
result of the moderating effect test supports Hypothesis 6b.
significant effect (t 5 3.262, p 5 0.001 < 0.05) on the dependent variable (analyticity).
However, WeChat usage intensity and WeChat usage dependence were not significantly
Impact of
social media
PI 0.252*** (4.460) 0.206*** (3.659) 0.216*** (3.873) use
PE 0.115* (2.054) 0.094 (1.715) 0.076 (1.387)
UISM 0.279*** (5.188) 0.213*** (3.867) 0.213*** (3.923)
SMD 0.223*** (4.012) 0.227*** (4.135)
UISM 3 SMD 0.133** (2.916)
Sample 300 300 300
R2 0.236 0.275 0.296
2
Adjusted R 0.228 0.265 0.284
F Value F 5 30.419, p 5 0.000 F 5 28.002, p 5 0.000 F 5 24.673, p 5 0.000
ΔR2 0.236 0.040 0.020
Note(s): *p < 0.05 **p < 0.01 ***p < 0.001 Results of moderating
DV: Cognitive maturity effects
SMD(low)
5
SMD(high)
CogniƟve maturity CM
2
Figure 5.
Moderating effect
1 results
UISM(low) UISM(high)
PI 0.178** (2.963) 0.153* (2.504) 0.160** (2.628)

PE 0.146* (2.450) 0.134* (2.259) 0.121* (2.029)
UISM 0.186** (3.262) 0.150* (2.519) 0.150* (2.536)
SMD 0.122* (2.022) 0.124* (2.079)
UISM 3 SMD 0.097 (1.946)
Sample 300 300 300
R2 0.141 0.153 0.164
2
Adjusted R 0.132 0.141 0.149
F Value F 5 16.217, p 5 0.000 F 5 13.311, p 5 0.000 F 5 11.507, p 5 0.000
ΔR2 0.141 0.012 0.011
Note(s): * p < 0.05 ** p < 0.01 *** p < 0.001; the numbers in the brackets are T values Results of moderating
DV: Analyticity effects
an influence on Y), we assumed that when WeChat usage intensity influences analyticity, the
fluctuation of the adjustment variable (WeChat usage dependence) may not have a significant
influence. In other words, WeChat usage dependence will not affect the relationship between
LHT WeChat usage intensity and analyticity. Thus, the result of the moderating effect test does
not support Hypothesis 7b.
dependence) was excluded, the independent variable (WeChat usage intensity) had a
significant influence (t 5 2.325, p 5 0.021 < 0.05) on the dependent variable (truth-seeking).
However, WeChat usage intensity and WeChat usage dependence were not significantly
an influence on Y), we assumed that when WeChat usage intensity influences truth-seeking,
the fluctuation of the adjustment variable (WeChat usage dependence) may not have a
significant influence. In other words, WeChat usage dependence will not affect the
relationship between WeChat usage intensity and truth-seeking. Thus, the result of the
moderating effect test does not support Hypothesis 8b.
5. Discussion
5.1 Principal findings
Most hypotheses of this research were verified using the statistical results presented above.
In terms of correlation, the total CTA score was significantly positively correlated with SMU
and significantly correlated with four of the six sub-dimensions of CTA, namely, self-
confidence, inquisitiveness, cognitive maturity, and analyticity. On the contrary, SMD was
significantly negatively correlated with the total CTA score and significantly negatively
correlated with all six sub-dimensions of CTA. According to the moderating effect tests, we
found that the degree of SMD affected the mutually beneficial relationship between UISM and
CTA. In other words, when the degree of social media dependence is high, an increase in
UISM may suppress or slow down the development of CTA.
In the test of the moderating effect of the six sub-dimensions of CTA, we found that the
same moderating effect was valid in three sub-dimensions, namely, self-confidence,
inquisitiveness, and cognitive maturity. The other three sub-dimensions, however, were
not valid in this context. As social media platforms become integrated into infrastructures,
algorithms serve as social intermediaries that influence users’ decision-making. Algorithms
pose a twofold challenge to users’ decision-making and judgment. First, people exhibit
prudent, dialectical, and reflective attitudes towards algorithms and are sceptical about some
decisions and judgments of algorithms. Second, people need the connections, shared
Constant 0.000 (0.000) 0.000 (0.000) 0.005 (0.087)

PI 0.243*** (4.064) 0.213*** (3.504) 0.214*** (3.508)
PE 0.123* (2.071) 0.109 (1.842) 0.107 (1.797)
UISM 0.133* (2.325) 0.088 (1.485) 0.088 (1.483)
SMD 0.150* (2.502) 0.150* (2.503)
UISM 3 SMD 0.013 (0.259)
Sample 300 300 300
R2 0.143 0.161 0.161
Adjusted R2 0.134 0.149 0.147
F Value F 5 16.439, p 5 0.000 F 5 14.112, p 5 0.000 F 5 11.268, p 5 0.000
ΔR 2
0.143 0.018 0.000
Results of moderating Note(s): *p < 0.05 **p < 0.01 ***p < 0.001
effects DV: Truth-seeking
information, and big data brought by social media to make decisions and observe the world Impact of
broadly. social media
Three sub-dimensions of CTA, namely, analyticity, systematicity, and truth-seeking are
more objective compared with the other three sub-dimensions (self-confidence,
use
inquisitiveness, and cognitive maturity). This observation is a possible reason for the
invalidity of the three hypotheses related to the three sub-dimensions of CTA. The key to
improving such objective capability is the accessibility of reliable information. Although
SMD has compressed the space for users to think independently, its interference is often
overlooked when users perform CTA.
Overall, this research confirms that SMU is a twofold phenomenon. First, social media, as
a convenient tool, is an important way for users to obtain information, perform social
interactions, and enrich their spiritual lives. This is particularly the case for university
students who can significantly improve their learning outcomes through both professional
and non-professional knowledge on social media platforms.
Nevertheless, balanced, appropriate, and speculative SMU is highlighted here. The
research results show that excessive dependence on social media may hinder the
development of CTA, especially in the case of today’s unoptimistic ecological Internet
environment. Serious issues, such as massive spam, rumours, and fake news, consume the
time and energy of young people. Moreover, all information provided by social media follows
algorithmic logic, and the prejudice and discrimination of algorithms are easy to spread but
difficult to detect. The distorted pseudo-environment is likely to be misleading for users in the
decision-making process.
When media literacy education is conducted for the generation of the internet, algorithm
and data literacies are often mentioned. Coincidentally, such emphasis shares the same
original intention of CTA. They both lead to the “usage divide” in the digital divide. “Usage
divide” refers to the gap among people in the usage of information and communication
technology (it is also known as the “second digital divide”). Scholars divide the digital divide
into three layers: access divide, usage divide, and knowledge gap (Wei and Zhang, 2006; Van
and Hacker, 2003).
The digital divide is cyclical and dynamic. It may be closed at one stage but re-opened at
another stage. The difference in the access and usage divides results in the difference in the
knowledge gap. Scholars in this field also face a paradox regarding how to narrow the digital
divide. As Qi (2019) commented, the algorithmic society pushes the tension between freedom
and restrictions to the extreme. SMU can narrow the digital divide from the access divide and
also widen the digital divide from the usage divide. Unfortunately, most studies have failed to
thoroughly distinguish between digital divide and measurement of the specific SMU.
Scholars have shown that the usage of mobile apps not only reflects social class habits, but
also contributes to social capital reproduction. Some people increase their social capital
through online education, information browsing, and social networking on the internet,
whereas some only enjoy leisure and entertainment activities on the internet (Liu and Yu,
2020). Future studies on the measurement of the influential factors of CTA and the
exploration of the relationship between CTA and the SMU may prioritise the differences
brought by mobile apps, social media accounts, and user habits. We highly recommend a
more in-depth measurement and exploration of the second digital divide, namely, usage
divide.
5.2 Conclusion
We conducted an exploratory research on the relationship between SMU and CTA. We found
that the usage intensity of social media (UISM) can promote the development of the CTA of
college students. We also verified the moderating effect of SMD. Our results have
enlightening value for library science. Some scholars have discussed the advantages and
LHT disadvantages of adopting social media applications in academic libraries (Collins and Quan-
Haase, 2014). On the one hand, linking library resources with public relations has become “the
most remarkable achievement of many libraries using social media” (Young and Rossmann,
2015). The learning community constructed by social media is based on the interest points of
the group and improves the learning effect through e-reading or multi-media intervention. Xu
et al. (2009) found that as an extension of traditional libraries, Web 2.0 can deliver news about
the library. Meanwhile, sites such as Facebook can foster norms of reciprocity and trust and,
therefore, create opportunities for collective action (Valenzuela et al., 2009). Cooper and May
(2009) argue that academic library blogs can encourage students to participate in campus
activities and gain library experience. On the other hand, although libraries must remain
current, proactive, and aware of emerging technologies to retain user interest, they are also
faced with the trap of misallocating personnel and technical resources, especially when
emerging technologies are used for non-academic purposes (Burhanna et al., 2009; Jacobson,
2011). Moreover, the use of social media in libraries may affect learning effect for young
people. These groups may use these tools primarily for social purposes and entertainment but
may be less familiar with information-seeking skills on the Web or the methods for using and
assessing digital resources (Collins and Quan-Haase, 2014). On this controversial issue, our
study demonstrates the positive effect of moderate SMU on students’ CTA and provides data
for vigilance against dependence and distraction caused by social media. Under the
environment of new media, media literacy education will become a new direction for
university libraries to help users improve their information evaluation ability effectively and
cultivate their critical thinking skills. The future online information literacy instruction of
university libraries should be based on information demand, and multi-level media literacy
training should be conducted, which can comprehensively promote the development of media
literacy education.
This study has several implications and merits. First, it provides a deeper exploration of
the controversial topic: What effect does SMU have on the health condition of college
students? Traditionally, more emphasis has been attached to the unhealthy impact of social
media on students. Several studies have reported a negative correlation between social media
and academic performance as social media can lead to behaviours that are potentially
harmful to students’ careers (Lau, 2017; Hasnain et al., 2015). Using social media for non-
academic purposes (video gaming in particular) and social media multitasking significantly
negatively predicted academic performance. However, with the recent commoditisation of e-
learning, many scholars have affirmed the positive effects of social media on online teaching
and learning (Kent, 2013; Selwyn, 2009). Social media is often used as a handy tool for
promoting CTA in learning activities (Cheung et al., 2011; Ekoc, 2014). This study describes
the complex relationships between SMU and CTA.
Second, we paid close attention to the conflict of interest and ethical issues in SMU.
Interesting and informative information on social media can indeed broaden horizons, but
this fragmented and entertaining pattern of information acquisition excludes or even rejects
deep thinking. In terms of promoting media literacy among undergraduates, we expect a
moderated use of social media as e-learning is not a complete substitute for offline classes and
face-to-face activities.
Third, for media literacy education, this study highlighted the importance of preventing
social media addiction and social media dependency. The results can provide a basis for
media literacy education and media contact among college students. CTA is an important and
necessary skill because it can help a person address mental and spiritual questions (Hatcher
and Spencer, 2005). As CTA requires higher-order thinking than simply the ability to recall
information, it is necessary for schools, families, and the society to provide curricula where
CTA can be reinforced in the parenting, teaching, and learning processes. Now is the most
appropriate time for university libraries to perform media information literacy education
(MIL) for students. Basic knowledge training of MIL is conducted for university student to Impact of
lead them in establishing correct media awareness and media ethics. The training also helps social media
the students lay a good foundation for the effective development of new college students’
learning and the overall improvement in their comprehensive quality.
use
5.3 Limitations
One possible deficiency of this research is the sampling method used. It was almost
impossible to obtain the information of all students at each university given the objective
constraints. The multistage stratified cluster sampling method was therefore used. Such a
sample selection method does not fully guarantee the randomness and heterogeneity of the
sample and cannot exclude the impact of different professions which may affect
students’ behaviours on media use. Journalism and communication students, for example,
need to use the internet more frequently to browse the news and observe topical social events
because of the demands of their coursework. However, students in science and engineering
disciplines have fewer needs in this respect. Finally, this study investigated the impact of
SMU on CTA at a meso level. Considering the functional diversity of WeChat, such as reading
news, checking study materials, chatting, and playing online games, the differences arising
from different behaviours should be further investigated. We suggest that, where available,
future studies that use probability sampling should obtain a statistically representative
sample set. It might also be helpful to control factors such as major, gender, and grade while
focusing on the impact of specific media use behaviours on the development of CTA.
References
Andreassen, C.S. (2015), “Online social network site addiction: a comprehensive review”, Current
Addiction Reports, Vol. 2 No. 2, pp. 175-184.
Asuero, A.G., Sayago, A. and Gonzalez, A.G. (2006), “The correlation coefficient: an overview”, Critical
Reviews in Analytical Chemistry, Vol. 36 No. 1, pp. 41-59, doi: 10.1080/10408340500526766.
Ay, F.A., Karakaya, A. and Yilmaz, K. (2015), “Relations between self-leadership and critical thinking
skills”, Procedia - Social and Behavioral Sciences, Vol. 207, pp. 29-41.
Baron, R.M. and Kenny, D.A. (1986), “The moderator–mediator variable distinction in social
psychological research: conceptual, strategic, and statistical considerations”, Journal of
Personality and Social Psychology, Vol. 51 No. 6, pp. 1173-1182.
Bauerlein, M. (2011), The Digital Divide: Arguments for and Against Facebook, Google, Texting, and
the Age of Social Networking, Penguin, New York, NY.
Bers, T.H., McGowan, M. and Rubin, A. (1996), “The disposition to think critically among community
college students: the California critical thinking dispositions inventory”, The Journal of General
Education, Vol. 45 No. 3, pp. 197-223.
Bianchi, A. and Phillips, J.G. (2005), “Psychological predictors of problem mobile phone use”,
CyberPsychology and Behavior, Vol. 8 No. 1, pp. 39-51.
Black, A. (2005), “The use of asynchronous discussion: creating a text of talk”, Contemporary Issues in
Technology and Teacher Education, Vol. 5 No. 1, pp. 5-24.
Bright, L.F., Kleiser, S.B. and Grau, S.L. (2015), “Too much facebook? An exploratory examination of
social media fatigue”, Computers in Human Behavior, Vol. 44, mar, pp. 148-155.
Burhanna, K.J., Seeholzer, J. and Salem, J. (2009), “No natives here: a focus group study of student
perceptions of Web 2.0 and the academic library”, The Journal of Academic Librarianship,
Vol. 35 No. 6, pp. 523-532, doi: 10.1016/j.acalib.2009.08.003.
Cao, X., Masood, A., Luqman, A. and Ali, A. (2018), “Excessive use of mobile social networking sites
and poor academic performance: antecedents and consequences from stressor-strain-outcome
perspective”, Computers in Human Behavior, Vol. 85, pp. 163-174.
LHT Chan, N.M., Ho, I.T. and Ku, K. (2011), “Epistemic beliefs and critical thinking of Chinese students”,
Learning and Individual Differences, Vol. 21 No. 1, pp. 67-77.
Chen, Y.Q. (2015), The Relationship Bwtween Use of Socialnetwork Sites and Individual Social Capital
(Unpublished Master Dissertation), Huazhong University Of Science And Technology,
Wuhan, Hubei.
Cheung, C.M., Chiu, P. and Lee, M. (2011), “Online social networks: why do students use Facebook?”,
Computers in Human Behavior, Vol. 27, pp. 1337-1343.
Collins, G. and Quan-Haase, A. (2014), “Are social media ubiquitous in academic libraries? A
longitudinal study of adoption and usage patterns”, Journal of Web Librarianship, Vol. 8 No. 1,
pp. 48-68.
Cooper, J.D. and May, A. (2009), “Library 2.0 at a small campus library”, Technical Services Quarterly,
Vol. 26 No. 2, pp. 89-95.
Cronbach, L.J. (1947), “Test “reliability”: its meaning and determination”, Psychometrika, Vol. 12
No. 1, pp. 1-16.
Derrick-Telemaque, E. (2014), “The relationship of academic achievement between critical thinking
and associated dispositions: a canonical correlation analysis”, Doctoral Dissertation, Andrews
University, MI.
Ekoc, A. (2014), “Facebook groups as a supporting tool for language classrooms”, TurkishOnline
Journal of Distance Education, Vol. 15 No. 3, pp. 18-26.
Ellison, N., Steinfield, C. and Lampe, C. (2007), “The benefits of Facebook “friends”: exploring the
relationship between college students’ use of online social networks and social capital”, Journal
of Computer-Mediated Communication, Vol. 12 No. 3, article 1.
Facione, P.A. (1990), “The California critical thinking skills test–college level”, Technical Report# 1.
Experimental Validation and Content Validity.
Facione, N.C. (1997), Critical Thinking Assessment in Nursing Education Programs: AnAggregate
Data-Analysis, California Academic Press, Millbrae, CA.
Facione, P.A. and Facione, N.C. (1992), The California Critical Thinking Disposition inventory (CCTDI),
Test Administration manual [M], California Academic Press, Millbrae.
Facione, P.A. and Facione, N. (2014), California Critical Thinking Disposition Inventory User Manual,
The California Academic Press: Insight Assessment.
Fairchild, A.J. and MacKinnon, D.P. (2009), “A general model for testing mediation and moderation
effects”, Prevention Science, Vol. 10 No. 2, pp. 87-99, doi: 10.1007/s11121-008-0109-6.
Field, A.P. (2012), Discovering Statistics Using SPSS (And Sex and Drugs and Rock “N” Roll), Sage,
London.
Fong, C.J., Kim, Y., Davis, C.W., Hoang, T. and Kim, Y.W. (2017), “A meta-analysis on critical
thinking and community college student achievement”, Thinking Skills and Creativity, Vol. 26,
pp. 71-83.
Forcier, E., Rathi, D. and Given, L.M. (2013), “Knowledge management and social media: a case study
of two public libraries in Canada”, Journal of Information and Knowledge Management, Vol. 12
No. 4, 1350039.
Foust, R.A. (2008), “Learning strategies, motivation, and self-reported academic outcomes of students
enrolled in web-based coursework”, (Unpublished doctoral dissertation), Wayne State
University, Detroit, MI.
Ghauri, P.N. and Grønhaug, K. (2010), Research Methods in Business Studies : A Practical Guide,
Pearson Education.
Giancarlo, C.A. and Facione, P.A. (2001), “A look across four years at the disposition toward critical
thinking among undergraduate students”, The Journal of General Education, Vol. 50 No. 1,
pp. 29-55.
Han, Z. and Liu, Y.Q. (2009), “Web 2.0 applications in top Chinese university libraries”, Library Hi Impact of
Tech, Vol. 28 No. 1, pp. 41-62.
social media
Hasnain, H., Nasreen, A. and Ijaz, H. (2015), “Impact of social media usage on academic performance of
university students”, 2nd International Research Management and Innovation Conference
use
(IRMIC), Langkawi, Malaysia.
Hatcher, D.L. and Spencer, L.A. (2005), Reasoning and Writing: From Critical Thinking to Composition,
3rd., American Press, Boston.
Holmbeck, G.N. (1997), “Toward terminological, conceptual, and statisticalclarity in the study of
mediators and moderators: examples from the child-clinical and pediatric psychology
literature”, Journal of Consulting and Clinical Psychology, Vol. 4, pp. 599-610.
Hou, Y., Xiong, D., Jiang, T., Song, L. and Wang, Q. (2019), “Social media addiction: its impact,
mediation, and intervention”, Cyberpsychology: Journal of Psychosocial Research on Cyberspace,
Vol. 13 No. 1, article 4.
Huang, H. (2014), Social Media Generation in Urban China, Springer, Berlin.
Huang, L., Liang, Y.L., Hou, J.J., Thai, J., Huang, Y.J., Li, J.X., Zeng, Y. and Zhao, X.D. (2019), “General
self-efficacy mediates the effect of family socioeconomic status on critical thinking in Chinese
medical students”, Frontiers in Psychology, Vol. 9 No. 7, p. 2578.
Ismail, I., Azizan, S.N. and Gunasegaran, T. (2019), “The use of SMS in promoting inquisitiveness
among distance”, Learners, Vol. 21 No. 1, pp. 57-81.
Jacobson, T.B. (2011), “Facebook as a library tool: perceived vs Actual use”, College and Research
Libraries, Vol. 72 No. 1, pp. 79-90.
Jenkins-Guarnieri, M.A., Wright, S.L. and Johnson, B. (2013), “Development and validation of a social
media use integration scale”, Psychology of Popular Media Culture, Vol. 2 No. 1, pp. 38-50.
Jiang, T., Hou, Y. and Wang, Q. (2016), “Does micro-blogging make us “shallow”? Sharing information
online interferes with information comprehension”, Computers in Human Behavior, Vol. 59,
pp. 210-214.
Jiang, Q.L., Hao, X.M. and Lin, C.J. (2019), “Psychological traits and behaviors regarding smartphone
use from media dependency perspective: a comparative study of college students in China and
Singapore”, Journalism Research No. 3, pp. 101-120.
_
Karahan, A. and Iskifo glu, G. (2020), “Using demographic variables to predict graduating university
students’ critical thinking disposition”, Social Behavior and Personality: An International
Journal, Vol. 48 No. 10, pp. 1-13.
Kavenuke, P.S. (2018), “Developing critical thinking skills in students: perceptions and practices of
lecturers from a Tanzanian University”, Unpublished Doctoral Dissertation, Beijing Normal
University, Beijing.
Kavenuke, P.S., Kinyota, M. and Kayombo, J.J. (2020), “The critical thinking skills of prospective
teachers: investigating their systematicity, self-confidence and scepticism”, Thinking Skills and
Creativity, Vol. 37, 100677.
Kent, M. (2013), “Changing the conversation: facebook as a venue for online class discussion in higher
education”, MERLOT Journal of Online Learning and Teaching, Vol. 9 No. 4, pp. 546-565.
Kim, H. (2014), “Enacted social support on social media and subjective well-being”, International
Journal of Communication, Vol. 8, pp. 2340-2342.
Kirschner, P.A. and Karpinski, A.C. (2010), “Facebook® and academic performance”, Computers in
Human Behavior, Vol. 26 No. 6, pp. 1237-1245.
Kuh, G.D. and Vesper, N. (2001), “Do computers enhance or detract from student learning?”, Research
in Higher Education, Vol. 42 No. 1, pp. 87-102.
Kustijono, R. and Zuhri, F. (2018), “The use of Facebook and WhatsApp application in learning
process of physics to train students’ critical thinking skills”, IOP Conference Series: Materials
Science and Engineering, IOP Publishing, Vol. 296 No. 1, p. 012025.
LHT Lai, E.R. (2011), “Critical thinking: a literature review”, Pearson’s Research Reports, Vol. 6 No. 1,
pp. 40-41.
Lau, W.W.F. (2017), “Effects of social media usage and social media multitasking on the academic
performance of university students”, Computers in Human Behavior, Vol. 68, pp. 286-291.
Lee, O.S. and Kim, M.J (2017), “The relationship among smartphone addiction, emotional intelligence,
critical thinking disposition and communication”, Skill for Nursing Students, Vol. 15 No. 7,
pp. 319-328.
Lepp, A., Barkley, J.E. and Karpinski, A.C. (2015), “The relationship between cell phone use and
academic performance in a sample of US college students”, Sage Open, Vol. 5 No. 1,
2158244015573169.
Leung, L. (2008), “Leisure boredom, sensation seeking, self-esteem, addiction: symptoms and patterns
of cell phone use”, in Mediated Interpersonal Communications, Lawrence Erlbaum Associates,
Mahwah, NJ, pp. 359-381.
Leyrer-Jackson, J.M. and Wilson, A.K. (2017), “The associations between social-media use and
academic performance among undergraduate students in biology”, Journal of Biological
Education, Vol. 1-10, pp. 221-230.
Lindley, P. and Walker, S.N. (1993), “Theoretical and methodological differentiation of moderation and
mediation”, Nursing Research, Vol. 42 No. 5, pp. 276-279.
Linh, N.C. (2008), “A survey of the application of Web 2.0 in Australasian academic libraries”, Library
Hi Tech, Vol. 26 No. 4, pp. 630-653.
Liu, Z.S. (2013), “Social media dependence and media demand: a case study of college students’
microblog dependence”, Journalism Bimonthly No. 1, pp. 124-134.
Liu, Y. and He, J. (2021), “Why are you running away from social media?” Analysis of the factors
influencing social media fatigue: an empirical data study based on Chinese youth”, Frontiers in
Psychology, Vol. 12.
Liu, M. and Yu, G.M. (2020), “The second digital divide faced by China: a study of influencing factors:
an empirical analysis from the perspective of social capital”, Modern Communication (Journal
of Communication University of China), Vol. 42 No. 12, pp. 1-8.
Ma, X., Zhang, Y. and Luo, X. (2021), “Students’ and teachers’ critical thinking in science education:
are they related to each other and with physics achievement?”, Research in Science and
Technological Education, Vol. 6, pp. 1-25.
Malik, A., Dhir, A., Kaur, P. and Johri, A. (2020), “Correlates of social media fatigue and academic
performance decrement: a large cross-sectional study”, Information Technology and People.
c, D., Arsic, S. and Milosevic, I. (2016), “Exploring students’ purposes of usage
Manasijevic, D., Zivkovi
and educational usage of Facebook”, Computers in Human Behavior, Vol. 60, pp. 441-450, doi:
10.1016/j.chb.2016.02.087.
Maurino, P.S.M. (2006-2007), “Looking for critical thinking in online threaded discussions”, Journal of
Educational Technology Systems, Vol. 35 No. 3, pp. 241-260.
Menichelli, M. and Braccini, A.M. (2020), “Millennials, information assessment, and social media: an
exploratory study on the assessment of critical thinking habits”, in Exploring Digital
Ecosystems, Springer, Cham, pp. 85-97.
Mieczkowski, H., Lee, A.Y. and Hancock, J.T. (2020), “Priming effects of social media use scales on
well-being outcomes: the influence of intensity and addiction scales on self-reported
depression”, Social Mediaþ Society, Vol. 6 No. 4, 2056305120961784.
Miri, B., David, B.C. and Uri, Z. (2007), “Purposely teaching for the promotion of higher-order thinking
skills: a case of critical thinking”, Research in Ence Education, Vol. 37 No. 4, pp. 353-369.
Mwalongo, A. I. (2014), “Student teacher and lecturer perceptions of the use of asynchronous
discussion forums, quizzes and uploaded resources for promoting critical thinking”, Thesis,
Doctor of Philosophy (PhD), University of Waikato, Hamilton, New Zealand.
Ndaku, A.J. (2013), “Impact of social media on students’ academic performance. A study of students of Impact of
university of Abuja”, Unpublished B. Sc. Dissertation, Department of Mass Communication
Management and Social Sciences. Caritas University, Enugu. social media
Nida, T.K. (2018), “Facebook addiction and its association with academic performance”, Biomedical
use
Journal of Scientific and Technical Research, Vol. 3 No. 5, pp. 1-3.
Pearson, K. (1896), “Mathematical contributions to the theory of evolution, III: regression, heredity and
panmixia”, Philosophical Transaction of the Royal Society, Vol. 187, pp. 253-318, London (A).
Peng, L. (2019), “Connection and anti-connection: the swing of Internet rules”, Chinese Journal of
Journalism and Communication, Vol. 41 No. 2, pp. 20-37.
Peng, M.C., Wang, G.C. and Chen, J.L. (2004), “A study on reliability and validity of critical thinking
ability measurement scale”, Chinese Journal of Nursing, Vol. 39 No. 9, pp. 644-647.
Perlovsky, L.I., Bonniot-Cabanac, M.-C. and Cabanac, M. (2010), “Curiosity and pleasure”, The 2010
International Joint Conference on Neural Networks (IJCNN), pp. 1-3.
Puzziferro, M. (2006), Online Technologies Self-Efficacy, Self-Regulated Learning, and Experiential
Variables as Predictors of Final Grade and Satisfaction in College-Level Online Courses, New
York University, New York.
Qi, Y.P. (2019), “The governance logic of algorithmic society”, ECUPL Journal, Vol. 22 No. 6, p. 5.
Ravindran, T., Chua, A.Y.K. and Hoe-Lian, D. (2013), “Characteristics of social network fatigue”, 10th
International Conference on Information Technology New Generations, IEEE, pp. 431-438.
Samaha, M. and Hawi, N.S. (2016), “Relationships among smartphone addiction, stress, academic
performance, and satisfaction with life”, Computers in Human Behavior, Vol. 57, pp. 321-325.
Selwyn, N. (2009), “Faceworking: exploring students’ education-related use of Facebook”, Learning,
Media and Technology, Vol. 34, pp. 157-174.
Shin, H., Park, C.G. and Kim, H. (2015), “Validation of Yoon’s critical thinking disposition instrument”,
Asian Nursing Research, Vol. 9, pp. 342-348.
Shirazi, F. and Heidari, S. (2019), “The relationship between critical thinking skills and learning styles
and academic achievement of nursing students”, The Journal of Nursing Research, Vol. 27 No. 4,
p. e38.
Singleton-Williams, S.D. (2009), “Motivational and cognitive learning strategies as predictors of
academic success in economically disadvantaged community college students”, Unpublished
doctoral dissertation, Capella University, Minneapolis, MN.
Stupple, E., Maratos, F.A., Elander, J., Hunt, T.E. and Aubeeluck, A.V. (2017), “Development of the
critical thinking toolkit (critt): a measure of student attitudes and beliefs about critical
thinking”, Thinking Skills and Creativity, Vol. 23, pp. 91-100.
Sumanti, S., Efendi, Z.M. and Ridwan, R. (2019), “Online group investigation by using facebook to
improve students critical thinking on geography subjects”, Proceedings of the 1st International
Conference on Innovation in Education (ICoIE 2018), Atlantis Press, pp. 130-131.
Thomas, D (2020), “Social media addiction, critical thinking and achievement emotions among EFL
students in Thailand”, Asia Pacific Journal of Educators and Education, Vol. 35 No. 1, pp.
157-171.
Thompson, J. (2009), “To question or not to question: the effects of two teaching approaches on
students’ thinking dispositions, critical thinking skills, and course grades in a critical thinking
course”, Unpublished doctoral dissertation, Capella University, Minneapolis,MN.
Tien, F.F. and Fu, T.T. (2008), “The correlates of the digital divide and their impact on college student
learning”, Computers and Education, Vol. 50 No. 1, pp. 421-436.
Tiwari, A., Lai, P., So, M. and Yuen, K. (2010), “A comparison of the effects of problem-based learning
and lecturing on the development of students’ critical thinking”, Medical Education, Vol. 40
No. 6, pp. 547-554.
LHT Ulucinar, U. and Aypay, A. (2016), “A model of decision-making based on critical thinking”, Egitim Ve
Bilim, Vol. 41 No. 185, pp. 251-268.
Valenzuela, S., Park, N. and Kee, K.F. (2009), “Is there social capital in a social network site?: facebook
use and college students’ life satisfaction, trust, and participation”, Journal of Computer-
Mediated Communication, Vol. 14 No. 4, pp. 875-901, doi: 10.1111/j.1083-6101.2009.01474.x.
Van, D.J. and Hacker, K. (2003), “The digital divide as a complex and dynamic phenomenon”, The
Information Society, Vol. 19 No. 4, pp. 315-326.
Wang, Z., Tchernev, J.M. and Solloway, T. (2012), “A dynamic longitudinal examination of social
media use, needs, and gratifications among college students”, Computers in Human Behavior,
Vol. 28 No. 5, pp. 1829-1839.
Wangensteen, S., Johansson, I.S., Bj€orkstr€om, M.E. and Nordstr€om, G. (2010), “Critical thinking
dispositions among newly graduated nurses”, Journal of Advanced Nursing, Vol. 66 No. 10,
pp. 2170-2181.
Watson, G. and Glaser, E.M. (1980), The Watson-Glaser Critical Thinking Appraisal, Psychological
Corporation, New York.
Wei, L. and Zhang, M.X. (2006), “The third digital divide: the knowledge gap on the Internet”,
Journalism and Communication, Vol. 13 No. 4, pp. 43-53.
Wu, H.Z. and Wu, Q.T. (2020), “Impact of mind mapping on the critical thinking ability of clinical
nursing students and teaching application”, The Journal of International Medical Research,
Vol. 48 No. 3, 030006051989322.
Xu, C., Ouyang, F. and Chu, H. (2009), “The academic library meets Web 2.0: applications and
implications”, The Journal of Academic Librarianship, Vol. 35 No. 4, pp. 324-331, doi: 10.1016/j.
acalib.2009.04.003.
Xu, H., Gong, Y., Zhang, Q. and Xie, J. (2019), “Relationship between social media activities and
thinking styles”, Marketing Intelligence and Planning.
Xue, J. and Hong, J.W. (2020), “A study on the influencing factors of social media fatigue in youth
groups from the perspective of role stress: taking wechat moments as an example”, Journalism
and Mass Communication Monthly No. 7, pp. 21-40.
Young, S.W.H. and Rossmann, D. (2015), “Building library community through social media”,
Information Technology and Libraries, Vol. 34 No. 1, doi: 10.6017/ital.v34i1.5625.
Yu, B.B. and Yu, L.Z. (2020), “An instrument for measuring critical thinking for LIS research and
practice: feasibility test of CTDI-CV simplification”, Library and Information, pp. 21-27.
Zhang, C., Msn, H.F., Msn, J.X., Guo, H., Msn, X.J. and Msn, Y.Y. (2017), “The effects of reflective
training on the disposition of critical thinking for nursing students in China: a controlled trial”,
Asian Nursing Research, Vol. 11 No. 3, pp. 194-200.
Zulkifli, N.N., Abd Halim, N.D., Yahaya, N. and Van Der Meijden, H. (2020), “Patterns of critical
thinking processing in online reciprocal peer tutoring through Facebook discussion”, IEEE
Access, Vol. 8, pp. 24269-24283.
Corresponding author
Li Cheng can be contacted at: [email protected]
For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: [email protected]
View publication stats

SSCI-Q2-ImpactofSocialMediaUse

Uploaded by

Copyright:

Available Formats

SSCI-Q2-ImpactofSocialMediaUse

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

SSCI-Q2-ImpactofSocialMediaUse

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Impact of social media use on critical thinking ability of university students

Article in Library Hi Tech · August 2022

The user has requested enhancement of the downloaded file.

Impact of social media use on Impact of

2.3 The negative impact of social media dependency on CTA

2.4 Other variables related to CTA

4.2 Validity analysis

4.3 Analysis of confirmatory factors

Variables Items Cronbach α

CTA KMO value 0.855

Absolute goodness of fit index

inquisitiveness, systematicity, cognitive maturity, analyticity, truth-seeking, WeChat usage

4.4 Correlation analysis

4.5 Regression analysis

Constant 0.000 0.039 – 0.000 1.000 – 0.552 0.546 F 5 90.778, p 5 0.000

4.6 Moderating effect tests

Model 1 Model 2 Model 3

PI 0.369*** (7.709) 0.296*** (6.680) 0.308*** (7.198)

Model 1 Model 2 Model 3

PI 0.284*** (4.877) 0.216*** (3.829) 0.226*** (4.075)

Model 1 Model 2 Model 3

PI 0.109 (1.809) 0.065 (1.082) 0.074 (1.228)

Model 1 Model 2 Model 3

PI 0.234*** (3.937) 0.203*** (3.379) 0.207*** (3.444)

Model 1 Model 2 Model 3

PI 0.178** (2.963) 0.153* (2.504) 0.160** (2.628)

Model 1 Model 2 Model 3

Constant 0.000 (0.000) 0.000 (0.000) 0.005 (0.087)

View publication stats

You might also like

PI 0.369* (7.709) 0.296* (6.680) 0.308*** (7.198)

PI 0.284* (4.877) 0.216* (3.829) 0.226*** (4.075)

PI 0.234* (3.937) 0.203* (3.379) 0.207*** (3.444)