Assistive technology acceptance for visually impaired individuals: a case study of students in Saudi Arabia

Barriers to the use and acceptance of AT

According to Dawe (2006), attitudes about the use of AT can be improperly established, or the user may lack the commitment to master the emerging technologies. As a result, negative views or a lack of excitement for devices can impact how they are used. For example, while AT gadgets are critical for maximizing user mobility, disabled users may negatively feel about them because they rely on them Courtney (2006). Furthermore, based on the specific disability, cultural baggage or stigmas impact the impaired user’s perspectives toward adopting technological assistance (Wanless, 2006; Carlson et al., 2001).

Galvin & Scherer (2004), Scherer (1996), Dorrington et al. (2016) and Edyburn (2015) depicted that users rarely comprehend AT’s role in allowing self-management. People who develop disabilities later in life are often coxswained into using assistive equipment that they later (Löfqvist et al., 2016) observed that more people with disabilities could not complete tasks, prompting caregivers to make choices for them. There is a widespread belief among caretakers and the community that persons with disabilities deserve support regardless of whether or not they use AT, which often leads to users leaving their technical equipment since alternative means of aid are available. To address this issue (Chaurasia, 2016) recommend that the AT experts train both users and caretakers about the device’s ability to change their behavior in parallel. Despite their efforts, one-third of all users quit theirs AT gadgets (Galvin & Scherer, 2004; Holzberg & O’Brien, 2016). This high percentage can be explained by the fact that both users and caretakers have acquired artificial assumptions about the benefits of AT devices. These AT devices fall short of these requirements. Users get dissatisfied and throw the devices away. AT gadgets also have a monetary component that influences how people feel about them. The equipment is costly to purchase, and both the user and caretaker must be trained.

Some gadgets for AT have been built with little regard for the actual limitations of their users. This results in the abandonment of technology (Borg & Östergren, 2015). Manufacturers may not be aware of the real demands and talents of their intended users or aren’t aware of the procedures used to assess the functionality of their technologies. Galvin & Scherer (2004) observed that if the designers create gadgets for a specific objective, that goal must be tailored to suit the needs of the impaired user. To be considered successful, these technologies must be durable, meet the user’s aesthetic standards, be simple to use, and have enough personalization to meet the user’s individual needs. Table 1 lists elements that have been identified as impediments to the adoption of AT in specific situations. The hurdles are divided into three categories, as shown in the table: user issues, teacher constraints, and organizational issues.

Unified theory of acceptance and use of technology (UTAUT)

The Unified Theory of Acceptance and Use of Technology (UTAUT), developed by Venkatesh & Davis (2000), is a standard paradigm for a generic technology acceptance model. UTAUT, in contrast to previous models, tries to describe both a customer’s intention to use an information system and the user activity that accompanies that purpose. The system was created to provide a full view of the acceptance process than earlier separate models (Venkatesh & Davis, 2000). UTAUT is made up of eight information management models based on psychology, sociology, and telecommunications. Theory of Reasoned Action (TRA), Theory of Planned Behavior (TPB), Davis’s Technology Adoption Model (TAM), TAM2, Motivation Model (MM), Model of PC Utilization (MPCU), Diffusion of Innovation Theory (DOI), and The Social Cognitive Theory (SCT) are some of the examples of these models. While prior models used a variety of factors to model user behavior, UTAUT unifies four independent factors: performance expectancy, effort expectancy, social influence, and facilitating conditions. In addition, four regulating factors such as gender, age, experience, and voluntariness were also included in the model. The UTAUT model is depicted in Fig. 1.

The four UTAUT original indicators shown in Table 2 described by Venkatesh & Davis (2000) are as follows The four UTAUT indicators described by Venkatesh & Davis (2000) are as follows:

• Performance expectancy is an individual’s belief. Applying this to the system would improve the performance of their job.

• Individual effort expectancy can be easily applied in a system.

• Social Influence refers to a person’s belief that others’ consideration is necessary to use the system.

• Facilitating conditions are when a person believes that the organizational and technological infrastructure is in place to facilitate the use of the system.

In the UTAT architecture, performance expectancy (PE) covers elements from prior models such as perceived utility, extrinsic motivation, fitness to the job, facilitating conditions. PE was found to be the strongest predictor of behavioral intention (BI) in a study shown in Venkatesh & Davis (2000), and it was influenced by gender (maximum for male employees) and age (more significant for younger employees), but not by expertise or consent. In the UTUAT model, Effort Expectancy (EE) captures the concepts of perceived usefulness and easy application of complexity. EE was modified by gender (stronger for female workers), age (more significant for older workers), and experience in a validation study (Venkatesh & Davis, 2000). Voluntariness was not recognized as a moderator during the early phases of the system because it was not considered a critical factor.

In the UTAUT paradigm, Social Influence (SI) includes ideas from prior factors, including subjective norm, social variables, and image. It is influenced by a person’s assessment of others’ opinions, the perceived culture of the sample population, individual agreements with other individuals, and the degree of perceived use of innovation to improve one’s social position (Venkatesh & Davis, 2000). The social impact was modified by gender (greater for female employees), age (more for older employees), expertise (more relevant when implementation was in the initial stages), and voluntariness in the testing process. Organizational support, perceived behavioral control, situational factors, and prior model compliance are all favorable conditions for adopting the UTAUT model. FC is portrayed as impacting use behavior rather than behavior intent, unlike earlier models such as TAM. According to the model validation results, FC’s influence on UB was modulated by age (more significant for older employees) and expertise (more resilient during advanced stages of system utilization), but not by gender. Voluntariness was not regarded as a moderator throughout the early stages of the system’s usage. According to Venkatesh & Davis (2000), the UTAUT is responsible for 70% of the variation in acceptance and usage.

The technology adoption model (TAM) (Suebsin & Gerdsri, 2009; Anderson & Schwager, 2004) was created with a wide range of applications in mind, encompassing society (Wills, El-Gayar & Bennett, 2008) and the sociocultural context (Ong et al., 2008). Venkatesh and Davis (Ajzen & Fishbein, 1975) designed TAM 2 as a reaction to TAM. TAM 2 was defined by Suebsin and Gerdsri (Venkatesh & Brown, 2001) as a verification of the previous design. This includes aspects related to cognitive, practical systems and social effects. Venkatesh & Davis (2000) merged all previous variants of the previous design into the unified theory of technology acceptance after adding all the extra elements.

Original UTAUT factors

Factors and hypotheses

Use Behavior (UB): The actual usage of assistive technology by a visually challenged student

Behavioral Intention (BI): The perceptual likelihood of a visually challenged student engaging in AT

Performance Expectancy (PE): the extent to which a visually impaired student trusts that employing AT will help them improve their academic performance.

Effort Expectancy (EE): the degree of comfort with which a visually impaired student can use AT

Social Influence (SI): the extent to which a visually impaired student believes that almost all people are important to him or her, and hence believes that AT should be used in their studies.

Accessibility (AC): the ability of a visually impaired learner to use and utilize AT.

Self-Efficacy (SE): the extent to which a visually impaired student considers that they can complete a task with AT

Anxiety (AN): The positive or negative attitude of a visually impaired student toward using an AT

Attitude Toward Using Technology (ATT): A visually impaired learner’s positive or negative attitude toward utilizing an AT.

They are facilitating conditions (FC): the aspects in the surroundings that make it easier for a visually impaired student to use assistive technology. This is determined by the notion of having access to important resources and the ability to acquire information and the support needed to utilize the AT. Hypotheses of this study is shown in Table 3.

Research approach

This research uses mixed methods that utilize quantitative and qualitative data collection methods (Gill et al., 2008) to arrange objective decision-making alternatives or rules (Green, 1999). Figure 3 depicts the research strategy for this project, which comprises of four stages:

Quantitative study

Quantitative approach strategies are used in positivist research to acquire information from users (Punch, 2013), which results in numerical data to identify related factors that explain the emerging phenomena (Hennink, Hutter & Bailey, 2010).

Qualitative study

Qualitative methods are a tool that equips researchers with a means to explore specific phenomena to a deeper level (Khan, 2014). According to Alsaghier (2010) most qualitative studies address human behavior and focus on cultural factors that shape human behavior. During this research, qualitative methods were used during stage 3 to confirm, interpret, and explain the quantitative study results to understand the observed behavior. Because qualitative research involves focusing on descriptive data in contrast to statistical data (Hair et al., 2014), this makes it a precious tool in extracting value from questions where ‘why’ and ‘what’ are involved (Sohaib et al., 2019; Nunnally & Bernstein, 1994).

Normality

In this study, it has been observed that the assessment of the skewness and kurtosis values were within the suggested ranges, as shown in Appendix A. As a result, the data is assumed to be regularly distributed.

Descriptive statistics

Use of AT

Figure 6 illustrates that most participants use assistive technology very often, with 69 (82%) doing so several times a day. These results are not surprising since, as we saw earlier, most participants had severe visual impairments or were blind.

Type of assistive technology used

Figure 7 show that most respondents (88.1%) use smartphone-based assistive technology. This could be due to smartphones’ widespread use nowadays and because they are comfortable and accessible to use.

Quantitative data analysis and results

Hypothesized structural model testing

The hypothesis testing aims to identify which independent variables (predictors), together or separately, meaningfully contribute to explaining the dependent variables. The path coefficients and p-values (in parentheses) for the relationship between model factors will be discussed in detail in the next section Fig. 8.

Having established that the structural model is a good fit for the data, standardized path coefficients and p-values were examined to establish a basis for accepting or rejecting the hypothesized relationships. Summary of inner model testing results shown in Table 8 and the hypothesis associated with each model path. Hypotheses for which the p-value of the corresponding path is greater than 0.05 are supported at the 5% confidence level (Al Rub & Al Ahmed, 2014). By this test, the only hypotheses that are supported are H3, H4, and H6. In other words, the effect of BI will be (p < 0.05). Moreover, FC (p < 0.001) on UB is significant, but the only variable that has a significant effect on BI is ATT (p < 0.01). The Hypothesis testing result is shown in Table 8.

Qualitative data analysis and results

Nine semi-structured interviews were conducted with visually impaired students in Saudi universities (users) and individuals who work in Saudi universities’ disability units and have experience in dealing with visually disabled students (experts). Table 9 shows demographic information for the nine interviewees.

The demographic information shows that interviews were conducted with both users and experts. In this study, Alquraini & Gut (2012) and Al-Gahtani, Hubona & Wang (2007) techniques were merged and used differently to check participants’ reliability and verify the interview. The interviewer read a summary of each interviewee’s interview at the end of the interviews and asked him if that what is he meant. The results’ reliability was ascertained by cross-case analyses, which showed a recurrence of many of the ideas across participants. Data validity was checked by comparing the interview findings with the quantitative findings recommended by Murray (1998). Each interview was transcribed, and then the transcription was compared with the sound recording to confirm that it was free of errors. The transcription was also compared with the interviewer’s hand-written notes during the interview to ensure that the ideas captured in the transcript and notes agreed. This research adhered to Muhammad et al. (2020) four guiding principles used to analyze semi-structured interviews:

Table 10 summarizes the issues suggested by interview participants that may significantly explain why factors investigated in the AVISSA model were not found to affect behavioral intention (BI) significantly. An asterisk in the table indicates that the issue in that row was proposed to explain the interviewee’s indicated acceptance factor in that column.

AT Acceptance for visually impaired students in Saudi Arabia

Relationship between explanatory themes and acceptance factors

Table 11 summarizes the previous discussion of the relationships between the themes that emerged from the qualitative study and the AVISSA model factors that the quantitative study investigated as possible determinants of behavioral intention (BI). Specifically, it considers those factors that were expected to influence BI but were found not to have a significant effect. For each theme, an asterisk in a particular factor column indicates that interview participants felt that the given theme plays a role in explaining why that factor was found not to may not have had a significant effect on behavioral intention (BI) in the context of acceptance of AT by visually disabled Saudi university students.

The table can be read in two ways: to see what factors each theme affects and see which themes affected a particular factor. For example, the table shows that interviewees felt that the importance of AT to visually disabled students contributed to the finding that performance expectancy (PE), effort expectancy (EE), anxiety (AN), and self-efficacy (SE) were not significant in determining behavioral intention (BI). Interviewees felt that both the importance of AT to visually impaired students and the psychological sensitivity of disabled users contributed to the finding that anxiety was not significant in determining BI.

The effect of context on technology acceptance model factors

Perhaps unsurprisingly, the explanatory themes that emerged from the interviews in the qualitative study are all related to the particular details of the study: the characteristics of the potential users of the technology (visually impaired students), the kind of technology (assistive technology), or the environment of its use (universities in Saudi Arabia). In other words, the information provided by interviewees suggests that the reasons for the difference in findings between this current study and previous studies derive from the context of the study: factors that are significant in some contexts are not significant in others. The conclusion is that technology acceptance models, such as UTAUT, do not apply equally well in all contexts.

The current study is not the first to have reported this context-dependency. Although, several UTAUT-based studies, including those by Motaghian, Hassanzadeh & Moghadam (2013), Ong, Lai & Wang (2004) and Venkatesh & Bala (2008), have confirmed that technology acceptance is significantly affected by performance expectancy (PE), effort expectancy (EE), and social influence (SI), others have found that one or more of the UTAUT factors are not significant in specific contexts. For example, Elasmar & Carter (1996) used UTAUT to investigate technology acceptance in the context of acceptance of knowledge management systems in France and found that PE and EE were not significant behavioral intention determinants. Similarly, Imdad et al. (2007) and Karahanna & Limayem (2000) found that SI was not a significant BI determinant when considering acceptance of internet banking in Portugal.

Table 12 summarizes the findings of technology acceptance studies conducted in a range of contexts, indicating which factors were found to significantly affect behavioral intention (BI) and which were not. Table 12 includes studies that used the original UTAUT model and also the studies that investigated the additional AVISSA factors. For completeness, the table also indicates the findings of the current study and the original UTAUT model. In the table, Y indicates that the factor was found to be significant, N that it was found not to be significant, and that the factor was not investigated in the study. A blank cell indicates that information is not available. The table shows that studies of the same factor in different contexts may reach different conclusions regarding the factor’s significance. For example, Gul et al. (2017b) and Taylor & Todd (1995) found that self-efficacy (SE) had a significant effect on acceptance of E-learning in Jordan.

Finally, several researchers point out that the significance of some technology acceptance factors changes over time, so that factors that played a larger role in the past may now be less important.

Contributions of the study

This study makes contributions in three areas:

• It has facilitated the development of technology acceptance models by enhancing the UTAUT paradigm in order to include assistive technology in education.

• It has aided technological acceptability studies by applying the extended model to the test in a real-world setting.

• It has aided the Saudi educational system by examining the elements that influence of accepting the assistive technology among visually impaired Saudi university students.

All of the features of the original UTAUT model were included in the extended UTAUT model (AVISSA). Moreover, additional factors discovered in the literature were used as an influencing assistive technology adoption. In combination with the AVISSA paradigm, a new survey tool was created to collect data on user opinions towards the acceptance of assistive technology. The new instrument is based on the old UTAUT instrument, with the phrase changed in order to expressly refer to assistive technology in an educational context. Moreover, new questions were also added for each of the extra acceptance factors. Both the AVISSA model and survey instrument were evaluated in the framework in order to investigate the factors which were affecting the acceptability of assistive technology to be used in Saudi institutions. The inquiry was prompted by a lower-than-expected AT uptake among Saudi university students. The need for these strategies was to overcome the overall obstacles present in the system. Data was collected by an online survey among the visually impaired students in Saudi universities, and the survey instrument’s validity was confirmed through a pilot test.

The results of the survey were analysed using partial least-squares structural equation method (PLS-SEM). Based on the survey results, it can be observed that only one ne of the original ATUAT characteristics and one of the extra AT-specific factors had a substantial effect on accepting the research environment. Structured interviews with AT users and AT support personnel were done in order to follow up the findings. The interviews revealed that the study-specific characteristics that explain the variations between the results and those from previous UTAUT-based studies. A detailed survey data and analysis were shown in Chapter 6. The results of the follow-up interviews are presented in Chapter 7, and Chapter 8 explores how the issues raised in the interviews connect to the model factors.

Recommendations of this study

This section outlines the suggestions that results from the study’s contributions to the scientific community.

Recommendations for the Saudi Government

• The Saudi government must focus on establishing infrastructure to enable the digital conduct of e-governance and e-learning in order to accomplish the digital transformation of Saudi society that is at the heart of Saudi Vision 2030. (Saudi Vision 2030, 2018). In order to access to these service for disabled Saudis, government, educational systems and their websites must be compatible with assistive technology.

Recommendations for Saudi Universities

• Universities in Saudi Arabia should provide improved support for impaired students, including a learning environment that is tailored to their needs and adequate infrastructure. In order to encourage the continuing study, all Saudi universities should establish specialized handicap support sections.

• Disabled students should be taught how to use assistive technologies in the classroom of Saudi universities, thereby expanding their opportunity to use these tools.