Factors Affecting Information and Communication Technology Use for University Students With and Without Learning Disabilities

Abstract

The purpose of the current study was to develop an appropriate model to represent the relationships among all the significant factors about information and communication technology use for university students, both with and without learning disabilities. This study was conducted in Taiwan with a sample consisting of an international array of research participants. A 40-item researcher-developed questionnaire was used to collect data from 400 students with learning disabilities and 747 students without disabilities. The structural equation modeling approach was a good model fit for both the measurement and structural models. The results indicated that the model proposed can explain the use of information and communication technology by students both with and without learning disabilities in universities. The model can explain 92% of information and communication technology use for students without learning disabilities, and 73% of that for students with learning disabilities. The present study illustrated that a model developed through structural equation modeling could demonstrate structural relations of the essential factors and information and communication technology usage for universities students, both with and without learning disabilities, in a holistic and comprehensive framework.

Keywords

students with learning disabilities information and communication technology structural equation modeling

Introduction

With the popularity of technology in today’s digital society, people depend on technology more heavily in their daily lives, especially in education. Information and communication technology (ICT) has been used in at all aspects of education, including distance teaching and learning, thus enabling a knowledge network that is useful for students, training teachers, providing high-quality teaching materials, and enhancing efficiency and effectiveness of educational administration and policy (Kaware & Sain, 2015).

Information and communication technology is widely used in universities, particularly where a prevalent e-learning environment exists. Students in many universities are required to attend virtual classes, use the Internet to submit homework, take exams, and/or read e-books, etc. (Heiman, Fichten, Olenik-Shemesh, Keshet, & Jorgensen, 2017). ICT is not only used for academic purposes, but also for campus administration functions, such as taking sick leave for a specific class, and selecting and dropping courses (Fichten et al., 2012). ICT changes the behaviors of students and teachers, and ICT capabilities and literacy have now become critical fundamental skills needed for participation in all aspects of life, including education, social, culture, and political. Currently, the coronavirus disease (COVID-19) has inevitably impacted education: universities have been forced to close their campuses and learning activities have been forced to move online. The ICT skills necessary for online learning were challenging to both lecturers and students (Aini, Budiarto, Putra, & Rahardja, 2020). Therefore, university students’ ICT skills play a crucial role both in their learning activities and adaptation to daily life during the pandemic.

Although people with disabilities of all ages have personal experience in using ICT, they often lack sufficient skills and knowledge to participate in today’s digital society (Mavrou, Meletiou-Mavrotheris, Kärki, Sallinen, & Hoogerwerf, 2017). According to a report from the Pew Research Center in 2021, 62% of adults with disabilities own a desktop or laptop computer, whereas 81% of those without disabilities do. The same report revealed that adults with disabilities are significantly less likely go online than adults without disabilities (15% vs. 5%, respectively) (Perrin & Atske, 2021). The digital divide between individuals with and without disabilities is still an important issue for professionals who work with persons with disabilities.

The digital divide issue comprises two aspects: ICT access and ICT competency (DiMaggio, Hargittai, Celeste, & Shafer, 2004; Mossberger, Tolbert, & Stansbury, 2003; Van Dijk, 1999; Van Dijk & Hacker, 2003; Warschauer, 2003). As internet access has become a standard for most Western populations, digital divide research shifted its focus to the tangible outcomes of internet use, which are called the third-level digital divide (Scheerder, Van Deursen, & Van Dijk, 2017; van Deursen & van Dijk, 2014). Elena-Bucea et al. (2021) investigated the influences of social demographic contexts on the digital divide between and within the 28 member-states of the European Union. The results found that e-Service adoption is influenced primarily by the one’s education level. Johansson et al. (2021) also investigated the use of the Internet, smartphones, computers, and tablets among people with disabilities in Sweden. The survey received 771 responses from people with 35 different diagnoses. The authors reported that a larger proportion of the participants than what is the norm for the general Swedish population reported not feeling digitally included. There are differences in digital inclusion between the diagnoses’ sub-groups. Depending on the types of disabilities involved, the impact of the use of ICT may vary. It might be necessary to investigate disability digital divides based on disability sub-groups. Therefore, this study focused on the university students with learning disabilities (LD) and investigated how they actually used their ICT skills at university.

Learning disabilities is a neurological disorder that affects the brain’s ability to receive, process, store, and respond to information (Hallahan, Kauffman, & Pullen, 2015). According to the identification standards for gifted and disabled students in Taiwan, learning disabilities are due to neuropsychological dysfunctions manifested in such areas as attention, memory, comprehension, perception, perceptual action, and reasoning. These factors cause significant difficulties in listening, speaking, reading, writing or arithmetic. Such impairment is not due to sensory, intellectual, emotional and other obstacles or environmental factors such as insufficient cultural stimulation and improper teaching (Identification Standards for Students With Gifted and Disabilities in Taiwan, 2013). A person with LD has difficulties in acquiring basic academic and functional skills (Pullen, Lane, Ashworth, & Lovelace, 2017). Due to these learning difficulties, academic skills and achievements of students with LD are usually worse than those of their peers, commencing at elementary school (Judge & Bell, 2011). In Taiwan, approximately 15,012 students with disabilities attend colleges or universities as of 2020, and the majority are students with LD. Approximately 5197 of these students with LD are enrolled in universities, accounting for about 34.6% of all university students with disabilities in Taiwan (Special Education Transmit Net, 2021).

When students with LD enroll in a university, they will still face various challenges (McGregor et al., 2016). They will encounter more problems than their peers in understanding coursework, completing homework and achieving good performance in exams. They also have more anxiety about school and need to spend more time studying to keep up with their academic coursework (Weis, Dean, & Osborne, 2016). ICT can provide support for students with LD in spelling, planning, organization, editing and calculation (Heiman et al., 2017). Students with LD stated that they are better able to express their opinions and needs when using ICT (Heiman & Shemesh, 2012; Heiman et al., 2017). ICT can provide LD students with the means needed to construct their own learning experiences according to their own learning pace (Adam & Tatnall, 2017). ICT skills are quite important to students with LD and one may thus ask: Do these students have sufficient ICT skills to meet their needs?

Wu et al. (2014) investigated 117 elementary school students with LD and found that those students with LD have equal opportunities to access computers and the Internet as compared to their peers, but they possessed poorer ICT competencies than their peers (Wu et al., 2014). They also found that those students without LD enhanced their ICT competency gradually year by year, but students with LD did not. A digital gap exists between students with LD and their peers in the skills of using graphic software, spreadsheet software and PowerPoint. Wu et al. (2014) concluded that mere provision of ICT access is not sufficient to close the digital gap between students with LD and their peers. A specific designed ICT instruction program should be provided to children with LD to enhance their ICT skills. Wu et al.’s research focused on the ICT skills of students with LD in elementary schools and their peers. Although the ICT skills improved across grades for typical developing students, it is unknown whether students with LD also enhance their ICT skills year by year. The question remains: Is there still a digital gap between students with LD and their peers when they enroll in the university?

Literature Review

ICT and Students With LD

Petretto et al. (2021) reviewed relevant articles from 2010 to 2020 on the positive and negative effects of distance teaching and e-learning on students with LD. The literature indicates that there is a general agreement that e-learning has positive effects for students with LD for learning certain academic skills, such as the ability to read, write, calculate, reason mathematically, as well as in learning certain cognitive and metacognitive abilities (Petretto et al., 2021). The use of online learning platforms can increase the autonomy of students with LD and allow them to develop individualized study plan that can reduce the cognitive load of students with LD, and improve their self-esteem.

Nguyen et al. (2013) investigated the use of ICTs by students with LD at an English college (Dawson College) and at two French colleges (Collège Montmorency and Cégep André-Laurendeau). The researchers interviewed experts and students with LD. Students with LD mentioned fewer specialized ICTs and more general use ICTs than did the experts. ICTs mentioned by LD students included mobile technology, instant messaging, and MP3s used to listen to books. Both students with and without LD reported that they preferred courses using ICTs and felt that ICTs helped them to complete school work. Fichten et al. (2014) also reported similar results. Those ICTs that are generally considered mainstream were used as assistive technology by students with LD.

Hollins & Foley (2013) interviewed 16 university students with LD to find out if they had any difficulties in their current virtual campus environment. The results showed that for each of the campus tasks, one to eight students out of 16 expressed difficulties. The most difficult item was “Identifying a journal article in the online library database.” Half of the students said they had difficulty with this item.

It can be seen from the above literature that ICT has a positive impact on students with LD. Technology is advancing and becoming more and more popular, and can assist LD students in their university studies. However, students with LD reported having more difficulty using ICTs and less knowledge about how to use ICTs than did their non-disabled peers (Fichten et al., 2013). Other studies have also demonstrated that students with LD have poorer ICT competencies than do their peers in elementary and junior high school, even if they have equal opportunities to access computers and the Internet (Wu et al., 2014, 2018). The cause of the digital divide is no longer the presence or absence of hardware, but rather, the difference in ICT competencies (Chen et al., 2014; Wu et al., 2014). This leads us to certain key questions. One may ask: As students with LD enter college, do they continue to experience digital gaps between themselves and their peers?

Theory and Related Concepts

Davis (1986, 1989, 1993) proposed the Technology Acceptance Model (TAM) to investigate whether the technology is accepted or rejected by the user. Research using TAM indicates that users’ motivation to use technology can be explained by three factors: perceived usefulness, perceived ease of use and attitude towards using. Davis hypothesized that the attitude of a user toward the system was a major determinant of whether he or she will accept or reject the use of this system. He defined perceived usefulness as the degree to which the persons believed that using a particular technology will improve their performance, whereas perceived ease of use refers to the belief that using the technology will be effortless (Davis, 1989). During the past three decades, the TAM has been refined by including other variables and modifying the relationships that were initially formulated.

Venkatesh et al. (2003) gathered the literature related to the user’s acceptance of information technology, and proposed the Unified Theory of Acceptance and Use of Technology (UTAUT). UTAUT can be seen as the extension of the TAM model. The authors proposed that in the UTAUT model, four constructs, performance expectancy, effort expectancy, social influence and facilitating condition, can directly affect the individual’s behavioral intention (Venkatesh, Morris, Davis, & Davis, 2003). Performance expectancy is equivalent to perceived usefulness in TAM and is defined as the degree to which individuals believe that using a particular system can help them achieve performance at work. Effort expectancy is similar to perceived ease of use in TAM, and is defined as the degree of ease associated with use of a particular system. Social influence means that the individual feels that important others think he or she should use the new system. Facilitating condition is defined as the objective factors of the environment that the user agrees make the task easy to achieve, for example, the provision of computer support. Behavioral intention means the user’s intention to use particular technology and it has a positive impact on the actual use of particular technology (Venkatesh et al., 2003). Other factors, such as self-efficacy and anxiety, have an indirect impact on a use’s intention. Self-efficacy is defined as judging a person’s ability to use technology to complete a specific job or task. Venkatesh et al. (2003) found that the influence of self-efficacy on behavioral intention was regulated by perceived ease of use. Anxiety does not directly affect a user’s intention, and the influence of anxiety on user intention was also regulated by ease of use (Venkatesh et al., 2003).

Park (2009) used TAM to investigate college students' intentions about e-learning in Korea. The researchers used questionnaires to obtain data from 628 college students who took online courses. The Structural Equation Modeling (SEM) was used to analysis the data. The results indicated that self-efficacy, social influence and attitude have a direct impact on behavioral intention. Self-efficacy and social influence also play an important role in affecting attitude towards e-learning. However, in this model, neither perceived usefulness nor perceived ease of use had a significant direct effect on behavioral intention to use e-learning. The reason may because learning through the Internet is already widely known to college students in Korea. Learning to use the Internet is considered by those college students to be easy and beneficial. Therefore, neither construct could directly affect the university students’ intention to use e-learning.

Salloum et al. (2019) used the TAM model to explore the acceptance of online learning (e-learning) among college students in the Arab United Nations. With respect to TAM constructs, the results pointed out that perceived ease of use has a significant positive impact on perceived usefulness, attitudes, and behavioral intention to use e-learning systems. The results also indicated that perceived usefulness has a significant positive effect on both attitudes and behavioral intention. Moreover, it was clear from the results that students' attitudes have a significant positive effect on behavioral intention, whereas the latter provided a strong positive association with the actual use of e-learning. Among those extrinsic factors, students' computer self-efficacy has a significant positive impact on perceived ease of use of e-learning systems.

Alrajawy et al. (2018) also used TAM as the basic theoretical model to explore the influence of anxiety on college students' intention to use mobile learning. The results indicated that perceived usefulness and ease of use can predict college students' willingness to use mobile learning, and that anxiety has a negative impact on both perceived ease of use and perceived usefulness. This study also found that anxiety of LD students has an influence on the intention of ICT use. However, anxiety has a direct effect on intention of ICT use, rather than indirect through perceived usefulness and ease of use.

Conceptual Framework

The TAM has been widely used in previous studies to investigate acceptance of technology by students, teachers, and stakeholders over the three decades (Granić & Marangunić, 2019; Ibrahim et al., 2017; Tarhini, Hone, & Liu, 2014). Therefore, the authors adopted the TAM model as the theoretical basis, and integrated the constructs from the UTAUT model to propose the theoretical conceptual model of this study. In order to understand the complex phenomena in ICT use of university students with and without LD, a Structural Equation Modeling (SEM) model was proposed to represent the relationships among all the related significant factors about ICT use (shown in Figure 1). Granić & Marangunić (2019) systematically reviewed the literate related to using TAM in educational context and found that the SEM was the most often deployed type of data analysis of TAM studies. SEM enables its users to understand antecedents of technology use, to elaborate on complex phenomena, and to enhance predictive validity of the hypothesized model (Hoyle, 1995).

Figure 1.

A theoretical conceptual model of ICT use of university students with and without learning disabilities.

The authors proposed an initial conceptual model which comprises all the possible factors that were generated from the past studies. Five major constructs were proposed in this theoretical conceptual model: (1) external factors (attitude toward using technology, social influence, self-Efficacy, facilitating condition, and anxiety) (2) performance expectancy (3) effort expectancy (4) behavioral intention and (5) ICT Use. ICT use in this study refers to the current status of ICT use for university students, including the current practices of using ICT to administrate school affairs, using social media, and using ICT to learn. The purpose of this study was to develop a model which could illustrate the ICT use and the essential related factors for university students with and without LD. Therefore, the specific research questions of this study were:

(1) Are there any differences in ICT use’ current status and related factors between college students with and without LD?

(2) Could a SEM model be established which includes the related factors comprising the phenomena of ICT use for college students with and without LD?

(3) Is there a significant difference between the path coefficient of a SEM model for college students with LD and that for students without LD?

Methods

Research Design

In order to fulfill the purpose of this study, several consecutive sub-studies were conducted. The authors developed a theoretical conceptual model (Figure 1) in advance, and used a sample of college students without disabilities to construct the SEM model, which elaborates on the phenomena of ICT use for university students without disabilities. Second, we verified the established SEM model via another sample of the NLD and of the students with LD. Finally, the differences between the path coefficients of the SEM model for college students with both LD and NLD were tested.

Sample

A total of 1154 students from 39 universities were enrolled in this study during the spring semester of academic year 2016–2017. Students without disabilities (NLD) were invited to participate in this study via the online platform, and a convenience sample was also obtained from the universities where the authors and project team members taught.

If a student with LD met the identification criteria of Taiwan, he or she was logged into the Special Education Transmit Net in Taiwan. Research assistants screened potential participants with LD via the Special Education Transmit Net in Taiwan and called the Centers for Students with Disabilities at each university to ask if they would be willing to assist in this study. If the school agreed, the questionnaire was sent, and the counselors of the Centers for Students with Disabilities assisted in the distribution the questionnaires. A total of 13 universities, and 430 questionnaires were sent out, 407 were returned, and seven incomplete answers were deleted. According to data from the Special Education Transmit Net, there were 3310 students with LD attending universities in Taiwan in 2017. This study’s LD sample size was 400, accounting for about 12% of all Taiwanese university students with LD in 2017. Students with documented neurological deficits, intellectual delay or physical impairments were excluded from the study. Students were able to fill in the questionnaire via a printed hard copy or the online platform, depending on their preference.

A valid sample of 1147 was obtained. Among those students included, 747 were students without disabilities, and 400 were students with LD. 347 students without disabilities (Sample 1, students with NLD) were randomly selected from 747 students to serve as the sample for constructing the initial SEM model. The remaining students, 400 students without LD (Sample 2) and 400 students with LD, were used as samples to test the fitness of the model. Table 1 showed the sample’s demographic information.

Table 1.

Demographic Information of University Students With and Without Learning Disabilities.

	Sample 1 NLD (n = 347)		Sample 2 NLD (n = 400)		Sample with LD (n = 400)
	n	%	n	%	n	%
Sex
Male	136	39.2	171	42.8	282	70.5
Female	208	59.9	222	55.5	118	29.5
Missing	3	0.9	7	1.8
Type of school
University	143	41.2	187	46.8	36	9.0
U of Tech	204	58.8	213	53.3	364	91.0
Grade
1	124	35.7	154	38.5	123	30.8
2	97	28.0	74	18.5	110	27.5
3	85	24.5	119	29.8	78	19.5
4 (above)	32	9.2	44	11.0	89	22.3
Missing	9	2.6	9	2.3

Note: Sample 1 NLD is the sample used for Parsimonious Structural Equation Modeling Predicting ICT Use of University Students without Learning Disabilities. Sample 2 NLD is the sample used for Verifying Structural Equation Modeling Predicting ICT Use of University Students without Learning Disabilities. U of Tech, University of Technology.

Instruments

The Contents of the Investigation of ICT Use in Universities Questionnaire

A researcher-developed questionnaire, titled “Investigation of ICT Use for Students in Universities (ICTU)”, was used to collect data. The contents of the questionnaire were reviewed and revised by five experts in information education and special education. The first scale was designed to measure students’ acceptance of using ICT at their universities. It was designed based on the TAM (Davis, 1986, 1989, 1993) and Unified Theory of Acceptance and Use of Technology (Venkatesh et al., 2003). The first part of the ICTU questionnaire included eight constructs: attitude toward using technology, performance expectancy, effort expectancy, social influence, self-efficacy, facilitating condition, anxiety, and behavioral intention. Each construct has three items; a total of 24 items were included (shown in Table 2). A correlation coefficient between each item and the scale scores was performed to test if the item was related to the scale. One item from attitude toward using technology and one item from effort expectancy were deleted, since the correlation coefficient between the item and scale scores were lower than 0.3 (Corcoran & Fisher, 1987). Therefore, a total of 22 item remained for the first part to the questionnaire. An example item for attitude toward using technology is “Using the digital system is interesting for me”. An example item for self-efficacy is “I have the knowledge and skills necessary to use the digital system”. This scale is a Likert five-point scale; the higher the score, the stronger the agreement.

Table 2.

Psychometric Properties of the ICT Use Survey.

Construct and Item	Correlation coefficient (>.3)	Factor loading (>.5)	CR (>.7)	AVE (>.5)	α (>.7)
Attitude toward using technology
ATUT 1	.37	.89	.80	.67	.80
ATUT 2	.38	.76
ATUT 3 (deleted)	.03	−.09

Performance expectancy
PE 1	.61	.76	.81	.58	.80
PE 2	.55	.68
PE 3	.65	.84

Effort expectancy
EE 1 (deleted)	.15	.20	.82	.69	.81
EE 2	.47	.80
EE 3	.53	.86

Social influence
SI 1	.45	.71	.80	.57	.79
SI 2	.60	.74
SI 3	.61	.81

Self-efficacy
SE 1	.65	.85	.89	.72	.88
SE 2	.66	.82
SE 3	.73	.88

Facilitating condition
FC 1	.74	.84	.83	.62	.83
FC 2	.73	.80
FC 3	.66	.72

Anxiety
ANX 1	.65	.75	.84	.73	.88
ANX 2	.82	.97
ANX 3	.72	.83

Behavioral intention
BI 1	.40	.74	.75	.51	.75
BI 2	.61	.76
BI 3	.54	.63

ICT use
ICT 1	.59	.67	.76	.58	.76
ICT 2	.50	.69
ICT 3	.53	.79

Note: ATUT, Attitude toward using technology; EE, Effort expectancy; PE, Performance expectancy; SI, Social influence; SE, Self-efficacy; FC, Facilitating condition; ANX, Anxiety; BI, Behavioral intention; ICT, ICT use.

The second part of the ICTU questionnaire is the current status of ICT use for university students. This included 18 items: the current status of using ICT to administrate school affairs (6 items), using social media (6 items), and using ICT to learn (6 items). The items regarding the current status of using ICT to administrate school affairs (ICT 1–6) include the use of the school website to administrate school affairs. These are online course selection, taking sick leave, application for dormitories, application for scholarship, reading announcements and affairs, and developing a personal e-portfolio. An example is: I rely on the school’s e-campus system’s application for school affairs, such as applying for scholarships (ICT 1). This scale is a Likert five-point scale; the higher the score, the higher the frequency. We used the average of items one to six to represent status of the ICT use in administrating school affairs.

The current status of using social media also has six items (ICT 7–12), including behaviors through digital devices for online dating, chatting, communication and discussion activities. Social networking sites include Facebook, Twitter, and social software, including Line, WeChat, BeeTalk and others. An example is: I often use social media to interact with my friends and classmates (ICT 7). We used the average of the scores from items 7–12 to represent the current status of ICT use in social media for university students.

The current status of using ICT to learn also consisted of six items (ICT 13–18), which include online/offline learning activities through digital devices, such as online courses, YouTube videos, learning via school digital learning platforms, and university open courses. Items include, for example, I often use the school’s online resources to study (ICT 13). We used the average of the scores from items 13–18 to represent the current status of ICT use in learning activities for university students.

Psychometric Properties of the ICT Use for Students in Universities Questionnaire

The reliability and validity of the ICTU were established based on a sample of 347 students without disabilities. AMOS 23 software was also used to perform a confirmatory factor analysis, and to test the fit of the structural model. The test of the variance of the ICTU found that none the estimated parameters had negative error variance, since factor loading was more than .5 (Bagozzi & Yi, 1988). The values of all standardized factors loading ranged from .63 to 97, which indicated a good basic fit between the theoretical model and the empirical data in this study.

The indicators of overall fit of this model were as follows: the standardized root mean residuals (SRMR) = .046, root mean square error of approximation (RMSEA) = .067, goodness-of-fit index (GFI) = .877, adjusted goodness-of-fit index (AGFI) = .829; normed fit index (NFI) = .892, non-normed fit index (NNFI) = .914, relative fix index (RFI) = .862, incremental fix index (IFI) =.931, and comparative fit index (CFI) = .930. Based on the above indicators, the overall fit of the model was satisfied.

To test the fit of the inner structure of the questionnaire, two indicators, composite reliability (CR) and average variance extracted (AVE), were used. The values of CR that were more than .70, and of AVE more than .50, were indicative of a good inner structure. All nine constructs in Table 2 met the CR and AVE criteria, thus indicating that the ICTU has good constructive validity. The Cronbach’s α of this scale ranges from .75 to .88, indicating good internal consistency. In summary, this ICTU has satisfactory validity and reliability.

Data Analysis

The initial measurement model, in which all the coefficients between factors were open for estimating, was tested and analyzed by SEM via the AMOS 23 software. The preliminary fit criteria, overall model fit, and fit of internal structure of model were used to test the model fit. The preliminary fit criteria include: no negative error variances, all error variances are significant, all correlations are smaller than one, and the factor loading is between .5 to .95 (Bagozzi & Yi, 1988). The recommended values of these indices were: GFI > 0.90, AGFI > 0.90, RMSEA < 0.08, SRMR < 0.05, NFI > 0.90, and PNFI > 0.50. To test the inner structures of the SEM models, two indicators, composite reliability (CR) and average variance extracted (AVE), were used (Bagozzi & Yi, 1988). The values of CR that were more than .60 and an AVE of more than .50 indicated good inner structures.

The SEM multi-group analysis can be used to determine whether a certain model that is suitable for a sample group is also suitable for other samples. This can evaluate whether the theoretical model proposed by the researcher is equivalent among different sample groups or if the parameter is invariant (Wang & Wang, 2012). Therefore, in this study, the multi-group analysis and verification function of the AMOS software was used to test the two groups – university students with LD and without LD (NLD). In addition, the path coefficients of these two groups, LD and NLD, were compared in the proposed model. If the path coefficients of the two groups of students reach a significant level, it is recommended that follow-up researchers explore the reasons in depth.

Results

ICT Use by University Students With and Without LD

As shown in Table 3, differences in ICT use by the two groups of students, LD and NLD, were shown in the sections describing current status of using social media, and using ICT to learn. Among students with NLD, the frequency of using social media and using ICT to learn is significantly higher than that of students with LD with a medium effect size (η² = 0.013; η² = 0.053). There is no significant different in using ICT for school administration between students with LD and NLD.

Table 3.

Scale Scores of the ICT Use Survey for University Students With and Without Learning Disabilities Using ICT (n = 800).

Construct	Item	NLD (n = 400)	LD (n = 400)	t	p	η ²
Attitude toward using technology	2	3.493 ± 0.644	3.545 ± 0.625	−1.171	.242	0.002
Performance expectancy	3	3.811 ± 0.747	3.722 ± 0.666	1.782	.075	0.004
Effort expectancy	2	3.339 ± 0.580	3.364 ± 0.578	−0.611	.542	<0.001
Social influence	3	3.780 ± 0.687	3.564 ± 0.683	4.455***	<.001	0.024
Self-efficacy	3	3.866 ± 0.744	3.734 ± 0.669	2.631**	.009	0.009
Facilitating condition	3	3.675 ± 0.832	3.677 ± 0.732	−0.030	.976	<0.001
Anxiety	3	2.406 ± 0.989	2.770 ± 0.977	−5.232***	<.001	0.033
Behavioral intention	3	3.620 ± 0.755	3.637 ± 0.745	−0.310	.756	<.001
ICT use in school administration	6	3.461 ± 0.881	3.460 ± 0.780	0.020	.984	<.001
ICT use in social media	6	4.003 ± 0.762	3.831 ± 0.760	3.188**	.001	0.013
ICT use in learning activities	6	3.804 ± 0.735	3.471 ± 0.680	6.656***	<.001	0.053

Factors Related to ICT Use of University Students With and Without LD

As shown in Table 3, there is no significant difference in the averages of scores in attitude toward using technology, performance expectancy, effort expectancy, facilitating, and behavioral intention between the students with LD and NLD. However, the average of scores in self-efficacy for students with NLD is significantly higher than that of students with LD with a low effect size (η² = 0.009), and the average of scores of social influence for students with NLD is significantly higher than that of students with LD with a medium effect size (η² = 0.024). The mean scores of anxiety for students with LD are significantly higher than NLD students with a medium effect size (η² = 0.033).

Constructing a SEM Model of ICT Use

Three hundred forty-seven college students without disabilities were used to construct the initial SEM model. The path coefficients between all variables were performed and those non-significant ones were deleted to form a parsimonious model.

Social influence was deleted in the parsimonious model since the path coefficients from social influence to effort expectancy, performance expectancy, behavioral intention, and ICT use were not significant, and the path coefficient was 0. Those path coefficients, including from anxiety to performance expectancy (p = .673), from anxiety to effort expectancy (p = .518), from facilitating condition to performance expectancy (p = .964), from facilitating condition to effort expectancy (p = .689), from attitude toward using technology to behavioral intention (p = .866), and from performance expectancy to ICT use (p = .732), that did not reach a significant level were also deleted in the parsimonious model. In addition, the correlation coefficients between anxiety and facilitating condition (p = .610), between anxiety and attitude toward using technology (p = .912), as well as between anxiety and self-efficacy (p = .099) that did not reach a significant level were also deleted. Figure 2 depicts the parsimonious model. The results obtained by the analysis indicated a good overall fit of this parsimonious model: χ2 = 461.303, p < .001; GFI = .887; AGFI = .852; CFI = .942, RMSEA = .063; SRMR = .057; NFI = .905; PNFI = .752.

Figure 2.

A parsimonious structural equation model for predicting ICT use of university students without learning disabilities (N = 347).

Verifying the Parsimonious SEM Model

After the parsimonious model was established, a cross-group verification was conducted to evaluate whether the model established in this study has invariant parameters between the other college students. This constructed SEM model was verified with two other groups: 400 university students without LD (Sample 2 NLD) and 400 university students with LD. In this cross-validation model, all the indexes indicated a good overall fit of the revised model (χ2 = 1290.625, p < .001; GFI = .863; AGFI = .821; CFI = .904, RMSEA = .054; SRMR = .0626; NFI = .870; PNFI = .727). The composite reliability (CR = .743–.865) and average variance extracted (AVE = .492–.682) of the eight latent variables in this model were also good, thus indicating the good internal structure of this model.

The path coefficient of 400 students without LD in this model was shown in Figure 3 and Table 4. The results indicated that self-efficacy had a direct effect on effort expectancy (.81, p < .001), as well as that attitude toward using technology and effort expectancy had a direct effect on performance expectancy (.71, p < .001; .30, p < .001). In this model, performance expectancy and self-efficacy had powerful direct positive effects on behavioral intention (.49, p < .001; .43, p = .004). These two constructs, combined with effort expectancy, facilitating condition, and anxiety, explained 85% of the total variance of behavioral intention. The most powerful direct positive effect on ICT use was effort expectancy (.39, p < .001), followed by attitude toward using technology (.35, p = .002) and then facilitating condition (.19, p = .005). Overall, this model explained 92% of the total variance of ICT use.

Figure 3.

Validation of a structural equation model for predicting ICT use of university students without learning disabilities (N = 400).

Table 4.

The Path Coefficient of a Structural Equation Model for Predicting ICT Use Among University Students With and Without Learning Disabilities (n = 800).

Path	NLD	Group LD	Z test (>1.96)	% of variance
ATUT→ EE	.01 (p = .902)	.24*** (p < .001)	2.220*	NLD: 67%; LD: 63%
SE → EE	.81*** (p < .001)	.63*** (p < .001)	1.449	NLD: 67%; LD: 63%
ATUT→ PE	.71*** (p < .001)	.63*** (p < .001)	1.280	NLD: 87%; LD: 82%
EE → PE	.30*** (p < .001)	.37*** (p < .001)	0.186	NLD: 87%; LD: 82%
PE → BI	.49*** (p < .001)	.55*** (p < .001)	0.113	NLD: 85%; LD: 51%
EE → BI	−.01 (p = .885)	−.32* (p = .022)	1.490
SE → BI	.43** (p = .004)	.38* (p = .041)	0.730
FC → BI	.08 (p = .371)	.03 (p = .840)	0.369
ANX → BI	.09 (p = .058)	.33*** (p < .001)	2.185*
BI → ICT	.16 (p = .267)	.00 (p = .979)	0.918	NLD: 92%; LD: 73%
EE → ICT	.39***(p < .001)	.17 (p = .095)	2.013*
ATUT→ ICT	.35** (p = .002)	.41*** (p < .001)	0.434
FC → ICT	.19** (p = .005)	.42*** (p < .001)	2.183*

*p < .05, **p < .01, ***p < .001.

The path coefficient of 400 students with LD in this model is shown in Figure 4 and Table 4. The results indicated that self-efficacy and attitude toward using technology had a direct effect on effort expectancy (.63, p < .001; .24, p < .001), and can also explain the 63% variance of effort expectancy. Attitude toward using technology and effort expectancy had a direct effect on performance expectancy (.63, p < .001; .37, p < .001), and can also explain the 82% variance of performance expectancy. In this model, performance expectancy, self-efficacy, and anxiety had a powerful direct positive effect on behavioral intention (.55, p < .001; .38, p = .041; .33, p < .001). Effort expectancy had a direct negative effect on behavioral intention (−.32, p = .022). These four constructs, combined with facilitating condition, explained 51% of the total variance of behavioral intention. The most powerful direct positive effect on ICT use was facilitating condition (.42, p < .001), followed by attitude toward using technology (.41, p < .001). Overall, this model explained 73% of the total variance of the ICT use for university students with LD.

Figure 4.

Validation of a structural equation model for predicting ICT use of university students with learning disabilities (N = 400).

Based on the above results, this SEM Model can be applied to college students both with LD and NLD, and that all the indicators of the models were good. However, the path coefficients might be different from these samples with LD and NLD.

Comparing the Path Coefficients of Models for University Students With and Without LD

In order to compare the path coefficients of the NLD students and the LD students, a Z test was performed. The results indicated that most of the path coefficients of these 2 samples did not have significant differences, except those from attitude toward using technology to effort expectancy, from anxiety to behavioral intention, from effort expectancy to ICT use, and from facilitating condition to ICT use.

In Table 4, the path coefficient from attitude toward using technology to effort expectancy for students with LD (.24**) is higher than that for students without disabilities (.01); the path coefficient from anxiety to behavioral intention for students with LD (.33**) is higher than that for students without disabilities (.09); and the path coefficient from facilitating condition to ICT use for students with LD (.42**) is higher than that for students without disabilities (.19**). On the other hand, the path coefficient from effort expectancy to ICT use for students without LD (.39**) is higher than that for students with LD (.17).

Discussion

ICT Use and Factors Related to Its Use by University Students With and Without LD

This study was designed to understand whether university students with LD use ICT differently than their peers. In our research, we found that students with LD used less ICT than their peers in social media and in learning activities. However, university students with LD did not differ from their peers in using ICT in school administration affairs. Disability might have an impact on ICT use. Previous studies demonstrated that students with LD reported worse ICT skills than did their typical peers (Vicente & López, 2010; Wu et al., 2014). These studies regard disability as the important factor which might generate a digital divide. Past studies have shown that the ICT skills of students with LD are less than that of their peers at the primary and secondary level (Wu et al., 2014). However, we did not investigate ICT skills in this current study. The reason as to whether LD students use ICT less frequently in these two areas is because of insufficient equipment or gaps in their abilities cannot be answered.

In our survey of factors on ICT use, we found that students with LD have lower self-efficacy, social influence and higher anxiety than their peers. Self-efficacy is derived from social cognitive theory and refers to judging one’s own ability to use technology to complete a specific task. It is uncertain whether LD students have low self-efficacy and high computer anxiety due to insufficient ICT ability or for other reasons. We can only observe this phenomenon through the use of descriptive statistics. Further research can explore possible reasons for variations in self-efficacy and high computer anxiety among students with LD.

A SEM Model Illustrating ICT Use by University Students was Established

In order to answer the second research question, a SEM model was established to illustrating the ICT use by university students both with and without LD. Based on the results, the model built in this research roughly conforms to the theoretical framework we proposed based on TAM and UTAUT theories, except social influence. Social influence has been deleted because the path coefficients from social influence to effort expectancy, performance expectancy, behavioral intention, and ICT use were not significant. Thus, social influence was not included in the subsequent model.

From a cross-sample analysis, it is found that the model proposed in this study can explain the use of ICT by both students with LD and NLD in universities. That is to say, these factors affect university students’ ICT use regardless of whether they have LD or not, are similar. The model explained the overall 92% variation of ICT use for NLD students, and the 73% variation of ICT use for LD students. The overall model explains the 85% variation of behavioral intention for NLD, and 51% variation for LD.

The results of this study are consistent with what is reflected in the TAM model. In the TAM model proposed by Davis (1986), the user’s intention to use ICT is an important factor in determining whether or not to use ICT. The intention to use ICT is mainly affected by two factors: perceived usefulness (performance expectancy) and perceived ease of use (effort expectancy) (Davis, 1986; Marangunić & Granić, 2015). In our model, performance expectancy has a direct impact on behavioral intention in the both groups, LD and NLD. However, the effect of effort expectancy on behavioral intention was only shown in the LD group, but not in the NLD group. Furthermore, the paths from effort expectancy to performance expectancy, as well as from performance expectancy to behavior intention, all reach a significant level in both groups, which reconfirm the hypothesis in the TAM model.

Self-Efficacy has a Direct Impact on Effort Expectancy and on Behavioral Intention

In our model, we found that self-efficacy is an important factor in ICT use intention. After Davis' original TAM model appeared, other scholars continued to expand and modify the TAM model (Marangunić & Granić, 2015). Among these external factors, self-efficacy showed a direct effect on effort expectancy (Marangunić & Granić, 2015).

In our model, self-efficacy not only has a direct impact on behavioral intention, but it is also the strongest predictor of effort expectancy. Self-efficacy has an indirect effect on ICT use through effort expectancy. The result of our study is similar to that of several previous studies which indicated that college students’ self-efficacy of technology has a significant impact on perceived ease of use of e-learning system (Salloum, Alhamad, Al-Emran, Monem, & Shaalan, 2019; Venkatesh et al., 2003). The authors of these studies suggested that when students have sufficient computer skills and expose a positive tendency to interact spontaneously with the e-learning system, their perceived usefulness of the system would definitely increase (Salloum et al., 2019).

In addition, our study also found that self-efficacy had a direct effect on behavioral intention in both groups. Some studies, such as Grandon et al. (2005), insisted that e-learning self-efficacy was found to have an indirect effect on students' intentions through perceived ease of use. However, the other studies indicated that self-efficacy has a direct effect on users’ intention (Park, 2009; Park, Nam, & Cha, 2012). Whether self-efficacy has a direct or indirect effect on intention to use, it obviously has a large impact on ICT use intention. In the future, how to enhance students' computer self-efficacy should be considered when developing an ICT curriculum.

Path Coefficients of a SEM Model

In order to answer Research Question Three, the path coefficients of the students with and without LD were compared. The results indicated that most of the path coefficients of these two samples did not have significant differences, except for the paths from attitude toward using technology to effort expectancy, from anxiety to behavioral intention, from facilitating condition to ICT use, and from effort expectancy to ICT use. For those path coefficients that did not differ between the students with and without LD, this represents the predictive power of the variables that did not differ significantly between these two groups.

Among those paths of the SEM Model, our model also shows that facilitating condition, attitude toward using technology also has a direct effect on ICT use for both LD students and NLD students. Improving their attitude toward using technology may be a possible way to increase ICT use on campus for college students both with and without LD.

The results indicated that path coefficients from performance expectancy to behavioral intention and from self-efficacy to behavioral intention for both NLD and LD students were significant. According to the research by Venkatesh et al. (2003), self-efficacy has no direct effect on use intention, but indirectly affects use intention through the moderation of ease of use, differs slightly from the results of this study. In addition to performance expectancy and self-efficacy, the previous study (Salloum et al., 2019) have also shown that effort expectancy was also able to predict behavioral intention.

There are several paths that are significantly different between the two groups of students, LD and NLD. They are the paths from attitude toward using technology to effort expectancy, from anxiety to behavioral intention, from facilitating condition to ICT use, and from effort expectancy to ICT use. For those path coefficients that were significantly different between the students with and without LD, this shows that the predictive powers between two variables in these two groups are different.

Both the path coefficients from attitude toward using technology to effort expectancy, and from facilitating condition to ICT use for students with LD are significantly higher than those for students without LD. This means that changing attitudes towards using technology affect LD students more than NLD students. Similarly, providing external technical technology related supports (facilitating condition) has a greater impact on students with LD than on students without LD.

In addition, we found that the path coefficient from anxiety to behavioral intention is also significant in the LD population, thus indicating that anxiety may affect the behavioral intention in the LD students, but not in their NLD peers. Similar results can also be found in Alrajawy et al. (2018)’s study. They found that, for LD students, anxiety will directly affect their intention to use ICT. How to reduce the anxiety of LD students about the use of technology to increase the use of ICT may be a topic that can be given significant attention in the future.

Summing up the above, the current study illustrated that a SEM model could demonstrate structural relations of the essential factors and ICT use for universities students with and without LD in a holistic and comprehensive framework. The model only reflects the current situation. According to our research, the ICT usage patterns of these LD students in university will turn out to be similar to those of students without disabilities.

Practical Recommendations and Implications

The results of our research are consistent with the theory of TAM, whether for students with or without LD. Ease of use of a system affects usefulness, and usability affects intent to use. The ease of use and usability of the system are very important as to whether users are willing to use the system. Therefore, an e-system must be designed to be easy to use, convenient to use, and useful to students in order to increase the acceptance of its use.

The results show that self-efficacy is a powerful predictive variable for behavior intention, whether in students with LD or NLD. The higher the self-efficacy, the higher the use intention. Therefore, how to improve self-efficacy to promote technology use intention is worthy of follow-up research.

The technology anxiety of students with LD is significantly higher than that of students without LD. In addition, this anxiety can affect students’ intention to use technology. How to reduce the technology anxiety of students with LD and improve their intention to use technology is an important consideration. Reducing students’ anxiety about technology through peer mentoring or segmenting system tools could possibly increase their technology acceptance.

Limitations and Further Study

We only investigated the current status of ICT use for university student with LD, but not their ICT competency. It is unclear whether the ICT competency of university students with LD is the same as that of university students without disabilities. Future studies should explore whether the gap in ICT competency between students with and without LD at primary schools has been closed after they have reached adulthood.

The data obtained in this study are not random sampling, since LD is a heterogeneous group. Therefore, the results of this study may not be over-generalized to all individuals with LD. In addition, the current study only recruited students with LD. Although the highest percentage of students with disabilities of school age is students with LD (Hallahan et al., 2015), the situations of different types of disabilities might have variations. Therefore, future studies could recruit students with other types of disabilities to determine if the model still can explain the ICT use of college students with other types of disabilities.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Ministry of Science and Technology (MOST 104-2511-S-003-010-MY2).

ORCID iDs

Ting-Fang Wu

Cheng-Ming Chen

Author Biographies

Ting-Fang Wu earned her B.S. in Occupational Therapy from National Taiwan University, Taipei, Taiwan in 1991, and received her M.A. from Department of Occupational Therapy, New York University, in U.S. in 1995. She pursued her Ph. D in Special Educational from National Taiwan Normal University, Taipei, Taiwan in 2002. Currently, she is a professor of Department of Special Education and Graduated Institute of Rehabilitation Counseling in National Taiwan Normal University. Her research interests are in applying assistive technology for persons with disabilities, vocational evaluation and rehabilitation for persons with disabilities.

Cheng-Ming Chen is an associate professor of special education at National Kaohsiung Normal University in Taiwan. He earned his B.S. in Department of Physics from National Changhua University of Education, Taiwan in 1993, and received his M.A. from Graduate Institute of Science Education and Environmental Education from National Kaohsiung Normal University, Taiwan in 1997. He pursued his Ph. D in Special Educational from National Kaohsiung Normal University, Taiwan in 2004. His research interest is making science more accessible and advanced to students who have special learning needs. Particularly, he explores how teachers can make adjustments and modifications to written materials and procedures so students who have specific needs can successfully engage in doing and inquiring science.

References

Adam

Tatnall

(2017). The value of using ICT in the education of school students with learning difficulties. Education and Information Technologies, 22(6), 2711–2726. https://doi.org/10.1007/s10639-017-9605-2

Aini

Budiarto

Putra

P. O. H.

Rahardja

(2020). Exploring e-learning challenges during the global COVID-19 pandemic: A review. Journal of Information System, 16(2), 57–65. https://doi.org/10.21609/jsi.v16i2.1011

Alrajawy

Isaac

Ghosh

Nusari

Al-Shibami

A. H.

Ameen

A. A.

(2018). Determinants of student’s intention to use mobile learning in Yemeni public universities: Extending the technology acceptance model (TAM) with anxiety. International Journal of Management and Human Science, 2(2), 1–9. https://ejournal.lucp.net/index.php/ijmhs/article/view/819

Bagozzi

R. P.

(1988). On the evaluation of structural equation models. Journal of the Academy of Marketing Science, 16(1), 74–94. https://doi.org/10.1007/BF02723327

Chen

M. C.

Chen

C. M.

C. C.

T. F.

Yeh

Y. M.

Y. P.

(2014). Different ICT competency but similar pattern between students with/without learning disabilities? Results from structural equation modeling testing. Lecture Notes in Computer Science, 8548, 528-531. https://doi.org/10.1007/978-3-319-08599-9_89.

Corcoran

K. J.

Fischer

(1987). Measures for clinical practice: A sourcebook. New York, NY: Simon & Schuster.

Davis

F. D.

(1986). A technology acceptance model for empirically testing new end-user information systems: Theory and results. (Doctoral dissertation). Cambridge, MA: MIT Sloan School of Management.

Davis

F. D.

(1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319–340. https://doi.org/10.2307/249008

Davis

F. D.

(1993). User acceptance of information technology: system characteristics, user perceptions and behavioral impacts. International Journal of Man-Machine Studies, 38(3), 475–487. https://doi.org/10.1006/imms.1993.1022

10.

DiMaggio

Hargittai

Celeste

Shafer

(2004). Digital inequality: From unequal access to differentiated use. In Neckerman

(Ed.), Social inequality (pp. 355–400). New York: Russell Sage.

11.

Elena-Bucea

Cruz-Jesus

Oliveira

Coelho

P. S.

(2021). Assessing the role of age, education, gender and income on the digital divide: Evidence for the European Union. Information Systems Frontiers, 23(4), 1007–1021. https://doi.org/10.1007/s10796-020-10012-9

12.

Fichten

S. C.

Asuncion

J. V.

Scapin

(2014). Digital technology, learning, and postsecondary students with disabilities: Where we’ve been and where we’re going. Journal of Postsecondary Education and Disability, 27(4), 369–379.

13.

Fichten

S. C.

Asuncion

J. V.

Wolforth

Barile

Budd

Martiniello

Amsel

(2012). Information and communication technology related needs of college and university students with disabilities. Research in Learning Technology, 20(4), 323–344. https://doi.org/10.3402/rlt.v20i0.18646

14.

Fichten

C. S.

Nguyen

King

Barile

Havel

Mimouni

Asuncion

(2013). Information and communication technology profiles of college students with disabilities and adequate and very poor readers. Journal of Education and Learning, 2(1), 176–188. https://doi.org/10.5539/jel.v2n1p176

15.

Grandon

Alshare

Kwan

(2005). Factors influencing student intention to adopt online classes: A cross-cultural study. Journal of Computing Sciences in Colleges, 20(4), 46–56. https://doi.org/10.5555/1047846.1047853

16.

Granić

Marangunić

(2019). Technology acceptance model in educational context: A systematic literature review. British Journal of Educational Technology, 50(5), 2572–2593. https://doi.org/10.1111/bjet.12864

17.

Hallahan

D. P.

Kauffman

J. M.

Pullen

P. C.

(2015). Exceptional learners: An introduction to special education (13th ed.). Upper Saddle River, NJ: Pearson.

18.

Heiman

Fichten

Olenik-Shemesh

Keshet

N. S.

Jorgensen

(2017). Access and perceived ICT usability among students with disabilities attending higher education institutions. Education and Information Technologies, 22(6), 2727–2740. https://doi.org/10.1007/s10639-017-9623-0

19.

Heiman

Shemesh

D. O.

(2012). Students with LD in higher education: Use and contribution of assistive technology and website courses and their correlation to students’ hope and well-being. Journal of Learning Disabilities, 45(4), 308–318. https://doi.org/10.1177/0022219410392047

20.

Hollins

Foley

A. R.

(2013). The experiences of students with learning disabilities in a higher education virtual campus. Educational Technology Research and Development, 61(4), 607–624. https://doi.org/10.1007/s11423-013-9302-9

21.

Hoyle

R. H.

(1995). The structural equation modeling approach: Basic concepts and fundamental issues. In Hoyel

R. H.

(Ed.), Structural equation modeling: Concepts, issues, and applications (pp. 1–15). Thousand Oaks, CA: Sage.

22.

Ibrahim

Leng

N. S.

Yusoff

R. C. M.

Samy

G. N.

Masrom

Rizman

Z. I.

(2017). E-learning acceptance based on technology acceptance model (TAM). Journal of Fundamental and Applied Sciences, 9(4S), 871–889. https://doi.org/10.4314/jfas.v9i4s.50

23.

Identification Standards for Students With Gifted and Disabilities in Taiwan (2013). https://law.moj.gov.tw/LawClass/LawAll.aspx?pcode=h0080065

24.

Johansson

Gulliksen

Gustavsson

(2021). Disability digital divide: The use of the internet, smartphones, computers and tablets among people with disabilities in Sweden. Universal Access in the Information Society, 20(1), 105–120. https://doi.org/10.1007/s10209-020-00714-x

25.

Judge

Bell

S. M.

(2011). Reading achievement trajectories for students with learning disabilities during the elementary school years. Reading & Writing Quarterly, 27(1–2), 153–178. https://doi.org/10.1080/10573569.2011.532722

26.

Kaware

S. S.

Sain

S. K.

(2015). ICT application in education: an overview. International Journal of Multidisciplinary Approach & Studies, 2(1), 25–32.

27.

Marangunić

Granić

(2015). Technology acceptance model: A literature review from 1986 to 2013. Universal Access in the Information Society, 14(1), 81–95. https://doi.org/10.1007/s10209-014-0348-1

28.

Mavrou

Meletiou-Mavrotheris

Kärki

Sallinen

Hoogerwerf

E. J.

(2017). Opportunities and challenges related to ICT and ICT-AT use by people with disabilities: An explorative study into factors that impact on the digital divide. Technology and Disability, 29(1–2), 63–75. https://doi.org/10.3233/TAD-170174

29.

McGregor

K. K.

Langenfeld

Van Horne

Oleson

Anson

Jacobson

(2016). The university experiences of students with learning disabilities. Learning Disabilities Research & Practice, 31(2), 90–102. https://doi.org/10.1111/ldrp.12102

30.

Mossberger

Tolbert

C. J.

Stansbury

(2003). Virtual inequality: Beyond the digital divide. Washington, DC: Georgetown University Press.

31.

Nguyen

M. N.

Fichten

King

Barile

Mimouni

Havel

Asuncion

(2013). Junior/community college students with learning disabilities and their use of information and communication technologies (ICTs). http://dc160.dawsoncollege.qc.ca/cfichten/AdaptechEnglishLDtechReport.pdf

32.

Park

S. Y.

(2009). An analysis of the technology acceptance model in understanding university students’ behavioral intention to use e-learning. Journal of Educational Technology & Society, 12(3), 150–162. https://www-jstor-org-443.web.bisu.edu.cn/stable/10.2307/jeductechsoci.12.3.150

33.

Park

S. Y.

Nam

M. W.

Cha

S. B.

(2012). University students’ behavioral intention to use mobile learning: Evaluating the technology acceptance model. British Journal of Educational Technology, 43(4), 592–605. https://doi.org/10.1111/j.1467-8535.2011.01229.x

34.

Perrin

Atske

(2021). Americans with disabilities less likely than those without to own some digital devices. https://www.pewresearch.org/fact-tank/2021/09/10/americans-with-disabilities-less-likely-than-those-without-to-own-some-digital-devices/

35.

Petretto

D. R.

Carta

S. M.

Cataudella

Masala

Mascia

M. L.

Penna

M. P.

Masala

(2021). The use of distance learning and E-learning in dtudents with learning disabilities: A review on the effects and some hint of analysis on the use during COVID-19 outbreak. Clinical Practice and Epidemiology in Mental Health: CP & EMH, 17(1), 92–102. https://doi.org/10.2174/1745017902117010092

36.

Pullen

P.C.

Lane

H. B.

Ashworth

K. E.

Lovelace

S. P.

(2017). Learning disabilities. In Kauffman

J. M.

Hallahan

D. P.

Cullen

(Eds.), Handbook of special education (2nd ed., pp. 286–299). New York: Taylor & Francis.

37.

Salloum

S. A.

Alhamad

A. Q. M.

Al-Emran

Monem

A. A.

Shaalan

(2019). Exploring students’ acceptance of e-learning through the development of a comprehensive technology acceptance model. IEEE Access, 7(99), 128445–128462. https://doi.org/10.1109/ACCESS.2019.2939467

38.

Scheerder

Van Deursen

Van Dijk

(2017). Determinants of internet skills, uses and outcomes: A systematic review of the second-and third-level digital divide. Telematics and Informatics, 34(8), 1607–1624. https://doi.org/10.1016/j.tele.2017.07.007

39.

Special Education Transmit Net (2021). Annual special education statistics. https://www.set.edu.tw/Stastic_WEB/sta2/stuF_city_spckind.asp

40.

Tarhini

Hone

Liu

(2014). The effects of individual differences on e-learning users’ behaviour in developing countries: A structural equation model. Computers in Human Behavior, 41, 153-163. https://doi.org/10.1016/j.chb.2014.09.020.

41.

van Deursen

A. J. A. M.

van Dijk

J. A. G. M.

(2014). The digital divide shifts to differences in usage. New Media & Society, 16(3), 507–526. https://doi.org/10.1177/1461444813487959

42.

van Dijk

(1999). The network society: Social aspects of new media. London: Sage.

43.

van Dijk

Hacker

(2003). The digital divide as a complex and dynamic phenomenon. Information Society: An International Journal, 19(4), 315–326. https://doi.org/10.1080/01972240309487

44.

Venkatesh

Morris

M. G.

Davis

G. B.

Davis

F. D.

(2003). User acceptance of information technology: Toward a unified view. MIS Quarterly, 27(3), 425–478. https://doi.org/10.2307/30036540

45.

Vicente

M. R.

López

A. J.

(2010). A multidimensional analysis of the disability digital divide: Some evidence for internet use. Information Society: An International Journal, 26(1), 48–64. https://doi.org/10.1080/01615440903423245

46.

Wang

J. C.

Wang

X. Q.

(2012). Structural equation modeling: Applications using Mplus. Chichester, UK: Wiley. https://doi.org/10.1002/9781118356258

47.

Warschauer

(2003). Technology and social inclusion: Rethinking the digital divide. Cambridge, MA: MIT Press.

48.

Weis

Dean

E. L.

Osborne

K. J.

(2016). Accommodation decision making for postsecondary students with learning disabilities: Individually tailored or one size fits all? Journal of Learning Disabilities, 49(5), 484–498. https://doi.org/10.1177/0022219414559648

49.

T. F.

Chen

C. M.

H. S.

Yeh

Y. M

Chen

M. C.

(2018). Factors related to ICT competencies for students with learning disabilities. Educational Technology & Society, 21(4), 76–88. https://www-jstor-org-443.web.bisu.edu.cn/stable/26511539

50.

T. F.

Chen

M. C.

Yeh

Y. M.

Wang

H. P.

Chang

C. H.

(2014). Is digital divide an issue for students with learning disabilities? Computers in Human Behavior, 39, 112-117. https://doi.org/10.1016/j.chb.2014.06.024.