Digital Inequality and Internet Self‐Efficacy in Adults With Intellectual Disability: Motivational, Demographic and Contextual Predictors of Online Participation Authors: Sarah Glencross, Jonathan Mason Published: 2026 ABSTRACT ABSTRACTBackgroundAdults with intellectual disability face multiple barriers to digital participation, yet limited research has examined how motivational and demographic factors interact to influence internet self‐efficacy.MethodStructured interviews were conducted with 96 adults with mild or moderate intellectual disability and 49 adults without intellectual disability across urban and regional locations in Australia. Internet self‐efficacy, attitudes and behavioural beliefs were assessed using adapted measures informed by the Theory of Planned Behaviour.ResultsParticipants with intellectual disability had significantly lower internet self‐efficacy than their peers, particularly those who were older, lived in regional areas or lacked professional support. Motivational variables predicted a small proportion of self‐efficacy variance for participants with intellectual disability and a larger proportion for the comparison group.ConclusionsFindings suggest digital inequality among people with intellectual disability is shaped by contextual and motivational factors. Support programmes informed by behavioural theory may help increase digital confidence and online engagement. Introduction The internet is now the primary access point for a growing number of services and civic activities. Businesses and government agencies increasingly use digital platforms to improve efficiency (Ravi Kumar and Bose 2016; van Holstein et al. 2021), enhance competitiveness (Betzing et al. 2018) and respond to public health imperatives, including those introduced during the COVID‐19 pandemic (Chang et al. 2020; Williams 2020). As a result, individuals without reliable internet access face a heightened risk of exclusion from educational, social and economic opportunities, a phenomenon commonly referred to as the digital divide (Blanck 2014; Lussier‐Desrochers et al. 2017). People with intellectual disability experience disproportionately low rates of internet access (Johansson et al. 2021; Lussier‐Desrochers et al. 2017; Patrick et al. 2020), and those who do access the internet may face barriers such as the cognitive demands of navigating complex online environments (Blanck 2014), concerns about online safety (Anderson et al. 2023; Seale 2014; Seale et al. 2012) and the risk of emotional distress or cyberbullying (Chadwick and Fullwood 2018; Jenaro, Flores, Vega, et al. 2018). Recent cross‐national studies confirm that people with intellectual disabilities are significantly less likely to use the internet than other disability subgroups, and those who do use it tend to engage in narrower online activities (Johansson et al. 2021). During the COVID‐19 pandemic, these disparities were exacerbated, as evidenced by research from Spain showing reduced access, inconsistent support and increased digital dependency for young adults with intellectual disability (Delgado et al. 2024). Structural and societal factors compound these challenges, including limited availability of accessible technologies, inadequate digital support or training, and gaps in policy and practice that fail to meaningfully address digital exclusion. Recent work by Chadwick (2022) highlights how these systemic barriers continue to restrict access and participation, underscoring the need for inclusive strategies co‐designed with people with intellectual disabilities. To promote digital equity and the valued lifestyles of people with intellectual disability, it is essential to understand not only who is excluded, but how and why such exclusion persists. Reducing digital inequality has the potential to improve autonomy, participation and quality of life. A review (Glencross et al. 2021) identified nine key benefits of internet use as reported by people with intellectual disability and their supporters. These included opportunities for friendship, romantic relationships, access to information, increased independence, self‐advocacy and identity development. More recent research has emphasised that mobile technology supports not only autonomy and self‐expression but also active social inclusion, navigation and access to employment and community (Anderson et al. 2023; Caton et al. 2024; Danker et al. 2023; Hebblewhite et al. 2022; Jacob and Pillay 2023). International research has further highlighted the connection between digital participation and wellbeing. Emerging research highlights that internet use can improve access to essential services such as healthcare, though this requires adequate support for digital search skills and evaluation of online content (Kuruppu Arachchi et al. 2023). A recent multi‐country study by Murphy et al. (2024) found that access to the internet was associated with improved physical and mental health, while pandemic‐related service digitisation intensified existing inequities. For some individuals, online spaces also facilitate exploration of romantic and sexual identity; however, these experiences are often shaped by social and technological constraints, including exposure to judgement and digital gatekeeping (Martino and Kinitz 2022). Despite such barriers, for many people with intellectual disability, the internet remains a critical tool for enhancing inclusion and exercising choice and control in everyday life. However, having internet access does not necessarily equate to digital inclusion. Many people with intellectual disability report using the internet for a limited range of purposes, often focused on entertainment or social media, rather than tasks such as information seeking, learning or accessing services (Jenaro, Flores, Cruz, et al. 2018; Normand and Sallafranque‐St‐Louis 2016). This phenomenon has been described as the second‐level digital divide, which describes disparities in types of internet use (van Deursen and van Dijk 2014). Recent findings suggest that people with intellectual disability engage less frequently in internet activities related to education, e‐government and financial services than people without intellectual disability, reinforcing concerns around second‐level digital inequality (Anrijs et al. 2022). In addition, teacher or caregiver concerns about online risks and assumptions of limited competence can lead to overprotection and restricted access to digital technologies, further compounding digital inequality (Borgström 2023, 2023; Chiner et al. 2021; Gómez‐Puerta and Chiner 2022). Support workers also play a central role in shaping online participation, yet many report limited confidence and training in managing online risks (Gómez‐Puerta and Chiner 2022), which can lead to restrictive practices and missed opportunities for digital learning and autonomy. Moreover, in addition to experiencing a second‐level digital divide, people with intellectual disability may gain fewer offline benefits from their internet use, such as improved employment prospects or civic engagement, a further layer of exclusion known as the third‐level digital divide (van Deursen and Helsper 2015). Additional factors such as older age and geographic remoteness are also associated with lower digital participation in the general population (Afshar Ali et al. 2020; Park 2017; Wilson et al. 2019). People with intersecting experiences, such as intellectual disability combined with ageing (Watchman 2019), cultural marginalisation (Hassiotis 2020) or indigeneity (Collings et al. 2017), may face compounded disadvantage. While research has begun to examine digital inclusion among older adults with intellectual disability (e.g., McCausland et al. 2023), few studies have jointly explored how age and other contextual factors such as regionality affect digital participation for this group as a whole (Gell et al. 2015; Wilson et al. 2019). In addition to demographic and environmental barriers, motivation is an important but underexplored component of digital inclusion. Behavioural research indicates that attitudes, beliefs and intentions shape engagement with technology (Peng et al. 2012; van Deursen and van Dijk 2014). The Theory of Planned Behaviour (TPB; Ajzen 1991) is a well‐established framework for understanding how attitudes, subjective norms and perceived behavioural control influence intentions and behaviour, and has been widely used to predict internet adoption and online engagement in the general population, including use for learning, banking and service access (Bidin et al. 2011; George 2004; Lee and Choi 2009; Yousafzai et al. 2010). In the context of digital inequality, the TPB offers a useful lens for examining psychological mechanisms that may contribute to second‐ and third‐level digital divides. Attitudes towards the internet, perceived social expectations and beliefs about one's ability to use digital technologies may shape patterns of use, while low perceived behavioural control and limited positive beliefs may restrict both online engagement and the translation of online activity into meaningful offline outcomes. Accordingly, this study applies the TPB as a guiding framework to examine whether motivational factors are associated with internet self‐efficacy and patterns of use among adults with intellectual disability, alongside demographic and contextual predictors. Despite its potential, the TPB has rarely been applied in studies directly involving people with intellectual disability, particularly as a framework to explore their own attitudes, norms and perceived behavioural control in relation to internet use. Previous studies have used the theory to predict support workers' intentions to promote physical activity (Martin et al. 2011) and healthy eating (Jenkins and Mckenzie 2011) among adults with intellectual disability, but no known studies have directly measured the attitudes, norms or perceived control of people with intellectual disability in relation to their own internet use. This limits the field's understanding of how personal motivation interacts with structural barriers to shape digital participation. Although digital inequality among people with intellectual disability has been documented, there remains a limited body of research specifically examining second‐ and third‐level digital divides in this population. While second‐level digital inequality has received some attention, third‐level inequality, which relates to the offline benefits of internet use, remains especially underexplored. Research investigating the predictors of internet self‐efficacy in this group is also sparse, despite its potential to inform targeted interventions. Comparative studies involving both people with and without intellectual disability are rare. Jenaro, Flores, Cruz, et al. (2018) and Jenaro, Flores, Vega, et al. (2018) found that young adults with intellectual disability used the internet more frequently for social and entertainment purposes and were more likely to experience problematic use. Alfredsson Ågren et al. (2020) compared adolescents with intellectual disability to a nationally representative sample and reported significantly lower access and less diverse use of online services among participants with intellectual disability. Despite these contributions, few studies to date have directly measured psychological predictors such as attitudes, norms or perceived control, highlighting an important gap in the literature on digital inclusion. Building on limited comparative research, this study includes adults across a wider age range and uses a consistent, accessible interviewer‐administered protocol to enhance comparability between groups. The study also incorporates the TPB to examine motivational, alongside demographic and contextual, predictors of internet self‐efficacy. While existing research has documented digital inequality among people with intellectual disability, much of this evidence comes from European contexts and focuses on younger populations. Research involving Australian adults with intellectual disability across diverse urban and regional settings remains limited. Given differences in service systems, digital infrastructure and regional disadvantage, confirmatory research is needed to assess whether previously identified patterns of internet use and digital exclusion are evident in Australia. Accordingly, this study aimed to compare patterns of internet use and internet self‐efficacy between Australian adults with and without intellectual disability, and to examine the demographic, contextual and TPB‐informed motivational factors associated with internet self‐efficacy. In this study, internet use behaviours refer to the frequency and types of online activities undertaken (e.g., messaging, information seeking, streaming, shopping and service access). Internet self‐efficacy refers to perceived confidence in using the internet, operationalised using the confidence component of the Internet Self‐Efficacy Scale. Motivation is operationalised as TPB‐informed beliefs, including behavioural beliefs, normative beliefs and perceived behavioural control. Digital inequality refers to disparities in access, patterns of use and the ability to benefit from online engagement, while digital equity refers to conditions that support fair and inclusive digital participation. The study aims to answer four research questions (RQs): What is the frequency and type of internet use by people with intellectual disability compared to people without intellectual disability? What is the frequency and type of internet use by people with intellectual disability compared to people without intellectual disability? To what extent can internet use behaviours be explained by personal and contextual characteristics? To what extent can internet use behaviours be explained by personal and contextual characteristics? Which variables are most strongly associated with differences in internet self‐efficacy, and can these be leveraged to reduce digital inequality? Which variables are most strongly associated with differences in internet self‐efficacy, and can these be leveraged to reduce digital inequality? To what extent can internet use behaviours be explained by motivational beliefs, over and above demographic and contextual factors? To what extent can internet use behaviours be explained by motivational beliefs, over and above demographic and contextual factors? Method Participants Participants with intellectual disability were aged 18 years or older and engaged with disability service providers across 14 urban and regional locations in Queensland, Australia. Intellectual disability status was determined by the participating service providers based on their existing client records and eligibility criteria for disability services. Individuals who met the study inclusion criteria were referred to the research team by the service providers. Inclusion criteria were mild or moderate intellectual disability, provision of informed consent (including consent from a formal or informal guardian where applicable) and ability to understand and respond to questions administered orally in a structured, interviewer‐administered assessment session with support or assistive technology if required. Recruitment Recruitment occurred via six disability service providers, which were contacted in writing by the first author, who also met with senior staff to explain study procedures. Services distributed ethics‐approved information materials to eligible clients. The first author attended each site to discuss participation and complete informed consent and subsequently returned to administer the structured assessment protocol. As recruitment was mediated by service providers, information about individuals who declined participation was not collected; a small number of individuals who initially expressed interest did not complete the assessment session and reasons were not systematically recorded. Participants without intellectual disability were aged 18 years or older and were recruited through snowball sampling. Inclusion criteria were: (i) no diagnosis of intellectual disability, (ii) informed consent provided and (iii) ability to understand and respond to questions administered orally in the same structured, interviewer‐administered assessment session. A total of 145 participants took part in the study: 96 participants with intellectual disability and 49 participants without intellectual disability. To facilitate comparison, the two groups were matched by age, t(143) = −1.749, p = 0.082. The group of participants with intellectual disability had an age range of 19–64 years (M = 36.25, SD = 12.53), and the group without intellectual disability had a range of 18–58 years (M = 40.10, SD = 12.56). Participants' gender, geographic location, accommodation type and employment status are shown in Table 1. TABLE 1Participants' gender, location, accommodation type and employment status. Participants with intellectual disability (n = 96), n (%) Participants without intellectual disability (n = 49), n (%)Total (N = 145), n (%)GenderMale60 (62.5%)17 (34.7%)77 (53.1%)Female36 (37.5%)32 (65.3%)68 (46.9%)Geographic location a Major city42 (43.8)36 (73.5%)78 (53.8%)Inner regional16 (16.7%)7 (14.3%)23 (15.9%)Outer regional38 (39.6%)6 (12.2%)44 (30.3%)Accommodation typeIndependent6 (6.3%)11 (22.4%)17 (11.7%)With family44 (45.8%)34 (69.4%)78 (53.8%)With friends0 (0%)2 (4.1%)2 (1.4%)Supported accommodation46 (47.9%)2 (4.1%)48 (33.1%)Employment statusFull‐time10 (10.4%)18 (36.7%)28 (19.3%)Part‐time52 (45.7%)17 (34.7%)69 (47.6%)Student0 (0%)6 (12.2%)6 (4.1%)Not employed34 (35.4%)8 (16.3%)42 (29.0%) a Geographic location as designated by Australian Statistical Geography Standard 2016 Remoteness Area (Australian Bureau of Statistics 2016). Participants' gender, location, accommodation type and employment status. Participants with intellectual disability (n = 96), n (%) Geographic location as designated by Australian Statistical Geography Standard 2016 Remoteness Area (Australian Bureau of Statistics 2016). The study included a measure of cognitive profile (described below). An independent samples t‐test confirmed there was a significant difference in the cognitive profile scores for the group with intellectual disability (M = 50.30, SD = 11.32) and the group without intellectual disability (M = 109.93, SD = 19.43); t(37.253) = −15.736, p < 0.001. Measures The assessment instruments used in this study were selected to capture internet attitudes, self‐efficacy and motivational beliefs relevant to digital participation, while allowing for adaption to ensure accessibility for adults with mild or moderate intellectual disability. For the analyses reported, the primary outcome variable was the Internet Self‐Efficacy confidence score (summed confidence across items). Internet attitudes were operationalised using the total attitude score (after reversing negatively worded items so higher scores indicated more positive attitudes). Motivational variables were operationalised using TPB subscale scores (Behavioural Beliefs, Normative Beliefs, Perceived Behavioural Control), with higher scores indicating more positive beliefs/greater perceived control. These derived scores were used in the correlation and regression analyses. The self‐report instruments were administered as structured, interviewer‐delivered questionnaires. All items were read verbatim and participants selected responses using a 5‐point Likert scale supported by a visual response aid. All participants, with and without intellectual disability, completed the same face‐to‐face structured interview protocol. To facilitate comparison between the responses from individuals with intellectual disability and individuals without intellectual disability, all participants were interviewed in person by the first author using the adapted versions of the measures and were provided with the visual analogue response scales. An overview of the research questions, variables, measures and statistical analyses is provided in Table 3. TABLE 3Overview of research questions, variables, measures and statistical analyses.Research questionOutcome variable(s)Predictor(s)/grouping variablesMeasure/derived score usedStatistical tests1. Frequency and type of internet useFrequency of internet use, type of internet use, internet access and device variables, internet skills and support variablesIntellectual disability group compared to no intellectual disability groupCategorical items from structured, interviewer administered internet use questions, e.g., frequency categories, types of use, types of device, support sourceChi‐square tests of association, Fisher's exact tests when expected cell counts were small2. Personal and contextual characteristicsInternet self‐efficacy (primary), plus descriptive links to use patterns where relevantCognitive profile (continuous), age (continuous), geographic location (categorical), type of internet support person (categorical)Cognitive profile: WAIS‐IV General Ability Index (from Verbal Comprehension and Perceptual Reasoning subtests); Internet self‐efficacy scale score; Location (ABS remoteness categories); type of internet support personLinear regression; ANOVA with Tukey post hoc; Kruskal–Wallis with Dunn post hoc and Bonferroni correction3. Internet self‐efficacyInternet self‐efficacy, internet attitude (as the paired outcome in the multivariate group comparison)Intellectual disability group compared to no intellectual disability groupInternet self‐efficacy scale score; General Internet Attitude Scale total score (after adjustment for reverse scoring)Hotelling's T 2/MANOVA; Pearson correlations4. MotivationInternet self‐efficacyTPB subscales: Behavioural Beliefs; Normative Beliefs; Perceived Behavioural ControlTPB subscale scores; Internet Self‐Efficacy Scale score (Confidence total score; after adjustment for reverse scoring)Multiple regressionAbbreviation: ABS, Australian Bureau of Statistics. Overview of research questions, variables, measures and statistical analyses. Abbreviation: ABS, Australian Bureau of Statistics. The study utilised the Wechsler Adult Intelligence Scale—Fourth Edition (WAIS‐IV; Wechsler 2008), a standardised assessment of adult cognitive ability. Cognitive profile was assessed for all participants using the Verbal Comprehension and Perceptual Reasoning subtests of the WAIS‐IV. Prorated subtest scores were used to calculate the General Ability Index, which served as the cognitive profile measure in subsequent analyses. The study also utilised adapted versions of the following measures: The General Internet Attitude Scale (Joyce and Kirakowski 2015), which assesses participants' overall positive and negative attitudes towards internet use. The General Internet Attitude Scale (Joyce and Kirakowski 2015), which assesses participants' overall positive and negative attitudes towards internet use. The Internet Self‐Efficacy Scale (Joyce and Kirakowski 2015) which measures participants' confidence in their ability to use the internet. The Internet Self‐Efficacy Scale (Joyce and Kirakowski 2015) which measures participants' confidence in their ability to use the internet. TPB Questionnaire (Ajzen 2019) which assesses motivational beliefs, including expected benefits, perceived social expectations and perceived control. TPB Questionnaire (Ajzen 2019) which assesses motivational beliefs, including expected benefits, perceived social expectations and perceived control. The original versions of the adapted measures were designed for the general population and included wording which could be a barrier to accessibility for people with mild or moderate intellectual disability. To address this, the items were reviewed by a panel of eight people experienced in working with individuals with intellectual disability. The panel recommended changes to the wording of several of the items; for example, the following item from the General Internet Attitudes Scale: ‘the internet does not threaten me’ was reworded to say, ‘the internet does not scare me’. The panel also sought to ensure the measures were concise to reduce the possibility of the duration of the structured interview being a barrier to participation or a cause of attrition. For example, the statement ‘I worry about using the internet’ was added to the adapted measure to replace the following items: ‘I feel bewildered by the internet’; ‘I feel intimidated by the internet’; ‘I feel overwhelmed by the internet’; ‘the internet makes me feel anxious’ and ‘the internet makes me feel uncomfortable’. To obtain feedback on the adapted items from individuals with intellectual disability, the first five structured interviews with participants with intellectual disability were carried out as pilot interviews. Once these were completed, the participants provided feedback about the accessibility of the items and their responses were reviewed so that, if required, further amendments could be made to the questions. Following this process, no further amendments were required so the responses from the pilot structured interviews were included with the data which were subsequently collected. Cognitive Profile The Verbal Comprehension subtests and the Perceptual Reasoning subtests of the WAIS‐IV were used to provide an indication of participants' cognitive profile. Prorated subtest scores were used to calculate the General Ability Index score. The General Internet Attitude Scale The General Internet Attitude Scale (Joyce and Kirakowski 2015) includes four subscales: Internet Affect; Internet Exhilaration; Social Benefit of the Internet and Internet Detriment. Items are rated on a 5‐point Likert scale from 1 (never) to 5 (always). The original version of the scale has 21 items; the version adapted by the research team using the process outlined above has 14 items across the four subscales. Scores can be combined to produce a total score indicating the degree to which an individual has a positive attitude towards the internet. Negatively framed items (e.g., Internet Detriment) were reverse scored prior to computing total and subscale scores so that higher values reflected more positive attitudes towards the internet. The four subscales capture affective responses (e.g., feeling comfortable vs. worried online), positive engagement (e.g., enjoyment/excitement), perceived social benefits (e.g., helping with connection/communication) and perceived harms/downsides (e.g., concerns about negative impacts). Items were administered orally with a 5‐point frequency agreement format, where 1 indicated the statement was never true for the participant and 5 indicated it was always true. The Internet Self‐Efficacy Scale The Internet Self‐Efficacy Scale (Joyce and Kirakowski 2015) measures both frequency of use and confidence. Frequency was rated on a 5‐point scale from 1 (Less than once a week) to 5 (More than once a day) and confidence on a 5‐point scale from 1 (Not confident) to 5 (Very confident). TPB Questionnaire The items for the TPB Questionnaire were developed following Ajzen's (1991, 2019) guides and reviewed by the panel and the pilot group as described above to check the accessibility of the wording and the face validity of the items. The final version of the Behavioural Beliefs subscale comprised four items; the subscale measuring Normative Beliefs had four items and the Perceived Behavioural Control subscale had five items. All items used 5‐point Likert scales. The TPB items are intended to capture theory‐consistent motivational beliefs rather than serve as a standalone validation of the TPB in this population. Reliability of the Measures Internal consistency (Cronbach's alpha) was calculated separately for each adapted scale within each group (with intellectual disability and without intellectual disability). Alpha coefficients for the adapted scales ranged from 0.735 to 0.889 in the intellectual disability group and from 0.755 to 0.840 in the comparison group (full values reported in Table 2). This indicates that the adapted instruments retained acceptable internal consistency when administered via structured interview to adults with and without intellectual disability. TABLE 2Reliability (Cronbach's alpha) of the adapted measures used in the study.Participants with intellectual disabilityParticipants without intellectual disabilityGeneral Internet Attitude scale0.7430.822Internet Self‐Efficacy scale0.8890.838Theory of Planned Behaviour Questionnaire0.7400.840Self‐determination scales0.8190.748 Reliability (Cronbach's alpha) of the adapted measures used in the study. Procedure Informed Consent Process for Participants With Intellectual Disability Following ethical approval (University of the Sunshine Coast Human Research Ethics Committee S171089 and S171141), service providers were contacted with information about the study. A two‐stage informed consent process was followed to allow potential participants and their informal/formal guardians sufficient time to consider what had been discussed and raise any further questions. An information sheet was provided after the structured interview for future reference. Informed Consent Process for Participants Without Intellectual Disability Participants without intellectual disability were recruited via snowball sampling. A time lag was introduced so their demographic characteristics (gender, age and location) could be matched to those of the intellectual disability group. Data Collection and Analysis Data were collected via in‐person structured interviews conducted by the first author with data collection being completed in 2021. The structured interview included demographic questions, frequency/type of internet use, the four adapted scales and WAIS‐IV subtests. Structured interviews took no more than 60 min; participants could take breaks or withdraw at any time. No financial incentives were offered. Analyses were performed using IBM SPSS Statistics (Version 26.0). Table 3 summarises the research questions, variables, measures and statistical analyses used in the study. Results The results are reported in four sections aligned with the research questions. Group differences in frequency and type of internet use (RQ1) are presented first, followed by group differences in internet attitudes and internet self‐efficacy (RQ2). Personal and contextual correlates of internet self‐efficacy within the intellectual disability group (RQ3) are reported next, and finally motivational correlates and multivariate prediction of self‐efficacy using the TPB subscales in each group (RQ4). Frequency and Type of Internet Use Differences in patterns of internet use between participants with and without intellectual disability are presented in Tables 4, 5, 6. Participants with intellectual disability had begun using the internet more recently than the participants without intellectual disability and were more likely to have been taught internet skills by, and receive internet support from, a friend or family member; they were much less likely to own a mobile device with data and tended to use the internet less frequently (Table 4). Participants with intellectual disability were less likely to have a smartphone, a laptop, internet access or to use the internet at home (Table 5). There were significant differences in all types of internet use; participants without intellectual disability reported higher usage in all types apart from playing games (Table 6). Personal and contextual predictors of internet use behaviours and internet self‐efficacy. TABLE 4Internet skills and support for participants with intellectual disability and participants without intellectual disability.Internet useParticipants with intellectual disability, n (%)Participants without intellectual disability, n (%)Chi‐squareFisher's exact χ 2 p p Time since commencing internet use16.762< 0.001< 2 years9 (11%)1 (2%)2–5 years26 (31%)3 (6%)> 5 years49 (58%)45 (92%)Has never used the internet6 (6%)0 (0%)Not known6 (6%)0 (0%)How they learnt to use the internet< 0.001At school26 (27%)12 (25%)At work0 (0%)2 (4%)Taught by friend/family member34 (35%)4 (8%)Taught by support worker12 (13%)2 (4%)Community service (e.g., library)4 (4%)0 (0%)Self‐taught8 (8%)29 (59%)Has never used the internet6 (6%)0 (0%)Not known/does not recall6 (6%)0 (0%)Person currently providing internet support most often73.749< 0.001Friend/family member32 (33%)11 (22%)Support worker36 (38%)2 (4%)Professional IT worker (e.g., at workplace or service provider)4 (4%)0 (0%)No‐one is available to provide internet support17 (18%)0 (0%)Does not require internet support7 (7%)36 (74%)Satisfaction with current internet skills3.7700.052Satisfied with current skills76 (79%)45 (92%)Would like to learn new skills20 (21%)4 (8%) Internet skills and support for participants with intellectual disability and participants without intellectual disability. TABLE 5Where and how the internet is accessed by participants with intellectual disability and participants without intellectual disability.Internet useParticipants with intellectual disability, n (%)Participants without intellectual disability, n (%)Chi‐squareFisher's exact χ 2 p p Where the internet is usedHas a mobile device with data so can access the internet in most locations7 (7%)39 (80%)78.293< 0.001Does not have a mobile device with data, so can only the internet in the following location(s)Home68 (71%)10 (20%)33.186< 0.001Work6 (6%)0 (0%)0.097Service provider27 (28%)0 (0%)16.935< 0.001Free Wi‐Fi in the community, e.g., library or café19 (20%)0 (0%)11.1600.001Does not use internet at present (may have done so previously)8 (8%)0 (0%)0.052Devices used to access the internet (owned by participant)Smartphone48 (50%)49 (100%)36.624< 0.001Tablet50 (52%)20 (41%)1.6490.199Laptop15 (33%)31 (63%)33.993< 0.001PC24 (25%)25 (51%)9.8170.002Entertainment (e.g., Smart TV, Xbox)8 (8%)14 (29%)10.3240.001 Where and how the internet is accessed by participants with intellectual disability and participants without intellectual disability. TABLE 6Frequency and type of internet use by participants with intellectual disability and participants without intellectual disability.Internet useParticipants with intellectual disability, n (%)Participants without intellectual disability, n (%)Chi‐squareFisher's exact χ 2 p p Frequency of internet use< 0.001Less than once a week2 (2%)0 (0%)Once a week2 (2%)0 (0%)A few times a week21 (22%)2 (4%)Once a day14 (15%)4 (8%)More than once a day49 (51%)43 (87.8%)Does not use internet at present (may have done so previously)8 (8%)0 (0%)Types of internet useSending people messages (e.g., email, WhatsApp)45 (49%)46 (94%)27.512< 0.001Social media (e.g., Facebook)54 (61%)40 (82%)6.3920.011Getting information (e.g., Google)69 (78%)49 (100%)12.878< 0.001Watching TV or movies (e.g., YouTube)63 (71%)48 (98%)14.827< 0.001Playing games (e.g., phone apps, Xbox)56 (63%)13 (27%)16.740< 0.001Shopping12 (14%)42 (86%)69.222< 0.001Banking25 (28%)47 (96%)58.266< 0.001Buying tickets (e.g., movies, transport)10 (11%)41 (84%)71.171< 0.001Learning things (e.g., online course, researching a topic)40 (45%)44 (90%)26.691< 0.001Accessing Government Services (e.g., social security, tax)20 (23%)47 (96%)68.245< 0.001Problematic internet use indicated49 (55.1%)23 (46.9%)0.8340.361 Frequency and type of internet use by participants with intellectual disability and participants without intellectual disability. Cognitive Profile A linear regression analysis examined whether cognitive profile (WAIS‐IV General Ability Index score) predicted internet self‐efficacy. There was a significant, strong positive correlation, r(92) = 0.66, p < 0.001, and cognitive profile significantly predicted internet self‐efficacy, F(1, 92) = 71.563, p = 0.001, R 2 adj = 0.43. Age A linear regression analysis examined whether age predicted internet self‐efficacy in the group with intellectual disability. There was a significant, moderate negative correlation, r(79) = −0.35, p < 0.001. Age significantly predicted internet self‐efficacy, F(1, 79) = 11.234, p = 0.001, R 2 adj = 0.11. Geographic Location Differences in internet self‐efficacy based on geographic location were analysed for participants with intellectual disability. Locations were classified using the Australian Bureau of Statistics' (2016) Remoteness Area categories: major city (n = 38), inner regional (n = 13) and outer regional (n = 30). Two outliers in the major city group were identified by boxplot and winsorized by replacing their values with the next highest value. There were significant differences in internet self‐efficacy across locations, F(2, 78) = 3.988, p < 0.022. Mean scores increased from outer regional (M = 25.40, SD = 17.17), to inner regional (M = 31.92, SD = 10.48), to major city (M = 36.95, SD = 18.00). Tukey's post hoc test revealed a significant difference between outer regional to major city groups (M difference = 11.55, 95% CI [1.78, 21.32], p = 0.016). In practical terms, participants with intellectual disability living in major city areas reported higher internet self‐efficacy than those living in outer regional areas, while inner regional scores fell between these groups and did not differ significantly from either. Internet Support Person Differences in internet self‐efficacy for people with intellectual disability who have different types of internet support person (the person who provides internet support most often) were examined. Participants with intellectual disability were classified into four groups according to type of support person: no support person (n = 8), support workers (n = 33), friends and family members (n = 31) and professional IT workers (n = 4). A Kruskal–Wallis test (Kruskal and Wallis 1952) identified the mean ranks of internet self‐efficacy varied between the groups, χ 2(3) = 8.726, p = 0.033. Subsequently, pairwise comparisons were performed using Dunn's (1964) procedure with a Bonferroni correction for multiple comparisons. This post hoc analysis revealed a statistically significant difference in internet self‐efficacy between the no support person (Mdn = 10.00) and professional IT worker (Mdn = 69.00) groups (p = 0.036). Participants who relied primarily on professional or paid support reported lower internet self‐efficacy than those who received support from friends or family, while participants who did not require support reported the highest self‐efficacy. Internet Attitude and Internet Self‐Efficacy Hotelling's T 2 was used to determine the effect of intellectual disability on internet‐related outcomes. Two variables were assessed: internet attitude and internet self‐efficacy. Participants with intellectual disability obtained higher scores on internet attitude (indicating a more positive view of the internet) than participants without intellectual disability (M = 52.93, SD = 7.35 and M = 48.59, SD = 6.67, respectively). In contrast, the participants with intellectual disability indicated lower levels of internet self‐efficacy than participants without intellectual disability (M = 33.10, SD = 20.45 and M = 62.41, SD = 13.52, respectively). The differences between the two groups of participants on the combined dependent variables was significant, F(2, 115) = 51.253, p < 0.001; Wilks' λ = 0.529; partial η 2 = 0.471. The estimated mean difference in internet attitude was 4.34, 95% CI [1.34 to 7.34], p = < 0.001. For internet self‐efficacy, the estimated mean difference was 29.31, 95% CI [−36.91 to −21.71] p = < 0.001. Pearson's product–moment correlation coefficients were calculated to assess the relationships between internet self‐efficacy and the three subscales of the TPB: Behavioural Beliefs; Normative Beliefs and Perceived Behavioural Control. As shown in Table 7, internet self‐efficacy was positively associated with behavioural beliefs and perceived behavioural control in both groups, indicating that participants who perceived greater benefits from internet use and greater control over their ability to go online also reported higher confidence in using the internet. Normative beliefs showed weaker and less consistent associations with self‐efficacy. TABLE 7Means, standard deviations and correlations between internet self‐efficacy and the theory of planned behaviour subscales. M SD1234Group with intellectual disability1.Internet SE32.5819.4210.40*** 0.100.28* 2.Behavioural beliefs35.384.9010.170.39*** 3.Normative beliefs14.904.3110.21* 4.PBC20.4985.591Group without intellectual disability1.Internet SE62.4113.5210.67*** 0.32* 0.35** 2.Behavioural beliefs37.123.5010.200.50*** 3.Normative beliefs13.7843.4710.034.PBC23.611.921Abbreviations: Internet SE, internet self‐efficacy; PBC, perceived behavioural control.* p < 0.05.** p < 0.01.*** p < 0.001. Means, standard deviations and correlations between internet self‐efficacy and the theory of planned behaviour subscales. Abbreviations: Internet SE, internet self‐efficacy; PBC, perceived behavioural control. p < 0.05. p < 0.01. p < 0.001. Motivation Motivation is operationalised through TPB‐consistent beliefs (behavioural, normative and control beliefs) and is examined as a potential correlate of self‐efficacy rather than as a direct measure of behaviour. To examine the relationship between internet self‐efficacy and behavioural intention as measured by the TPB Questionnaire, multiple regression analyses were used to predict internet self‐efficacy from the three TPB subscales. Separate analyses were conducted for the group with intellectual disability and the group without intellectual disability. Both the multiple regression models significantly predicted internet self‐efficacy (Table 8). For both groups, the only variable which added significantly to the prediction was Behavioural Beliefs. TABLE 8Multiple regression results for the prediction of internet self‐efficacy from the theory of planned behaviour subscales.Internet self‐efficacy B 95% CI for B SE B ß F (df) R 2 ∆R 2 LLULGroup with intellectual disabilityModel4.444 (3, 63)0.18** 0.14** Behavioural beliefs1.333** 0.34422.3220.4950.336** Normative beliefs0.026−1.0331.0860.5300.006PBC0.514−0.3581.3870.4370.148Group without intellectual disabilityModel15.429 (3, 45)0.51*** 0.47*** Behavioural beliefs2.464*** 1.5033.4240.4170.639*** Normative beliefs0.734−0.1061.5750.4170.189PBC0.168−1.5461.8830.8510.024Abbreviations: ∆R 2, adjusted R 2; B, unstandardized regression coefficient; CI, confidence interval; df, degrees of freedom; F, F ratio; LL, lower limit; Model, ‘Enter’ method in SPSS statistics; PBC, perceived behavioural control.; R 2, coefficient of determination; SE B, standard error of the coefficient; ß, standardised coefficient; UL, upper limit.** p < 0.01.*** p < 0.001. Multiple regression results for the prediction of internet self‐efficacy from the theory of planned behaviour subscales. Abbreviations: ∆R 2, adjusted R 2; B, unstandardized regression coefficient; CI, confidence interval; df, degrees of freedom; F, F ratio; LL, lower limit; Model, ‘Enter’ method in SPSS statistics; PBC, perceived behavioural control.; R 2, coefficient of determination; SE B, standard error of the coefficient; ß, standardised coefficient; UL, upper limit. p < 0.01. p < 0.001. Discussion The present study confirms the persistence of a digital divide between people with and without intellectual disability, reinforcing a well‐established body of evidence that digital inequality is a multidimensional, systemic issue affecting many aspects of daily life (Chadwick 2022; Chadwick and Fullwood 2018; Gómez‐Puerta and Chiner 2022). This section focuses on the study's three central contributions: group differences in internet use and access, key demographic and contextual correlates of internet self‐efficacy and the additional contribution of TPB‐informed motivational beliefs. Significant group differences were found across multiple aspects of internet use, including the timing of first use, the source and mode of internet skill acquisition, satisfaction with current skills, access to internet‐enabled devices and mobile data, and the frequency and type of online activities. These findings extend previous research by providing cross‐group comparisons across a wider adult age range, and by highlighting how digital participation for adults with mild or moderate intellectual disability remains heavily influenced by social and environmental dependencies (Chadwick 2022; Delgado et al. 2024). In contrast to the group without intellectual disability, where most participants expressed satisfaction with their digital skills, approximately 20% of participants with intellectual disability indicated they wanted to learn more. This signals a significant and ongoing need for accessible, strengths‐based digital skill development programmes. Structural inequality also played a significant role in shaping internet participation. While all participants reported owning at least one internet‐enabled device, those without intellectual disability were far more likely to own multiple devices and to have flexible, independent access. The most striking difference was observed in smartphone access and connectivity. Although half of the participants with intellectual disability owned a smartphone, only a small fraction (7%) had mobile data access, compared to 80% of the comparison group. This substantially limits spontaneous and community‐based internet use, with likely flow‐on effects for autonomy, communication and service access. Recent studies have similarly identified gaps in device ownership, connectivity and digital agency (Borgström 2023; Chiner et al. 2021), with implications for autonomy, employment, communication and participation in civic life. These findings contribute to international efforts to understand and address digital exclusion as a barrier to valued lifestyles, and support calls for inclusive technology policy, community‐based digital support and co‐designed interventions that reflect the needs and preferences of people with intellectual disability. Differences in frequency of internet use appeared closely tied to disparities in access to mobile data and independent connectivity. The absence of session duration data is a limitation, and future research should incorporate objective indicators such as usage logs or digital trace data to better understand the intensity and contexts of online participation. The largest disparity in internet activity was found in access to government services online. While 96% of participants without intellectual disability used the internet for this purpose, only 23% of the participants with intellectual disability reported doing so. This reflects a clear second‐level digital divide in terms of functional use. It also reinforces earlier findings by Jenaro, Flores, Cruz, et al. (2018), who reported that people with intellectual disability are more likely to engage with the internet for social and entertainment purposes. Recent studies conducted in a range of international contexts have confirmed this pattern (Hebblewhite et al. 2022; Jacob and Pillay 2023). The lack of engagement with government and informational services may also reflect a third‐level digital divide, where the offline benefits gained from internet use differ between groups (van Deursen and Helsper 2015). Although this study did not directly measure offline outcomes, the usage patterns observed suggest reduced access to the broader advantages of digital inclusion, such as access to health services, civic participation and self‐determined decision‐making. These gaps have important implications for practice and policy and must be addressed to promote equitable and valued lifestyles for people with intellectual disabilities. Significant associations were identified between cognitive profile and internet self‐efficacy, reinforcing the importance of cognitive accessibility in digital inclusion. These findings are consistent with previous research showing that the cognitive demands of navigating online environments can limit participation among people with intellectual disability (Blanck 2014). A moderate negative correlation was also found between age and internet self‐efficacy, aligning with prior studies in both the general population and disability‐specific contexts (McCausland et al. 2023). Participants living in outer regional areas reported significantly lower levels of internet self‐efficacy compared to those living in major cities. This pattern of findings suggests that intersecting factors such as age, geographic location and disability may compound disadvantage and exacerbate digital exclusion. These results extend previous work on intersectionality and digital inequality by confirming that people with intellectual disability who are older or living in regional areas face additional barriers to online participation (Collings et al. 2017; Watchman 2019; Wilson et al. 2019). Addressing these overlapping forms of exclusion is essential to promoting equitable participation in digital society. The type of internet support was a predictor of internet self‐efficacy. Participants with intellectual disability who received support from professional IT staff reported higher internet self‐efficacy than those with no support person, suggesting that access to skilled, targeted support may be an important lever for digital inclusion. While differences between other support types were not statistically significant, this finding is consistent with concerns raised in previous studies about the capacity of support workers and family members to provide adequate digital guidance. Both Borgström (2023) and Gómez‐Puerta and Chiner (2022) emphasise that many support workers lack the confidence, training or resources required to effectively teach digital skills. These limitations can reduce opportunities for people with intellectual disability to build competence, develop autonomy and navigate risk online. Further research is needed to identify which forms of support are most effective for fostering digital skill development and building sustained confidence. This includes examining co‐learning models, capacity‐building for support networks, and peer mentoring approaches. This study also examined the utility of the TPB in predicting internet self‐efficacy. The TPB was applied as a framework to organise and interpret motivational correlates of internet self‐efficacy, rather than as a comprehensive explanatory model of digital participation among adults with intellectual disability. Although the TPB explained a relatively small portion of the variance in self‐efficacy for participants with intellectual disability (14%), the findings were statistically significant and suggest that motivational beliefs play a meaningful role in shaping digital engagement. Among the three TPB components, behavioural beliefs emerged as the strongest and only significant predictor for both groups. This highlights the potential value of designing internet training or inclusion programmes that target individuals' beliefs about the benefits and positive outcomes of going online (Ajzen 1991). For example, belief‐focused digital inclusion interventions could incorporate structured digital skills workshops that explicitly highlight practical benefits of internet use, such as accessing health information, managing appointments or maintaining social connections online. These programmes might combine hands‐on digital training with opportunities for participants to reflect on positive online experiences, helping to strengthen beliefs about the usefulness and value of internet engagement. In addition, the finding that access to professional IT support was associated with higher self‐efficacy suggests that disability services could consider integrating specialist digital mentors or technology support roles within service models, enabling individuals to receive tailored assistance while developing independent digital skills. Similar belief‐focused approaches have shown utility in other applied behaviour change contexts (e.g., Kothe et al. 2012; Sabouri et al. 2020; Stead 2004). Despite its promise, the TPB accounted for a much smaller proportion of variance in internet self‐efficacy among participants with intellectual disability compared to those without (14% vs. 47%). This disparity suggests that while behavioural motivation is important, it is unlikely to be sufficient on its own. These findings reinforce the need to treat behavioural theory as complementary to, rather than a substitute for, structural and contextual accounts of digital inequity. For individuals with intellectual disability, internet engagement depends on a broader set of interacting conditions, including cognitive ability, environmental access, technical support and inclusive design. Interventions must therefore be multifaceted, combining motivational strategies with structural improvements in accessibility, digital literacy training and social support. Only through this comprehensive approach can digital inequality be meaningfully addressed, enabling greater autonomy, inclusion and participation in contemporary society for people with intellectual disability. Methodological Considerations and Limitations While this study provides valuable insights into digital participation among adults with intellectual disability, several limitations should be acknowledged. Participants with intellectual disability were recruited via disability service providers in Queensland, and no information was collected about individuals who declined participation; a small number who initially expressed interest did not complete the assessment, and reasons were not systematically recorded. This recruitment pathway may have favoured individuals already engaged with formal services and therefore more likely to have access to digital supports or exposure to internet use, which may influence the interpretation of findings regarding digital participation and self‐efficacy. The study relied on structured, interviewer‐administered self‐report measures with fixed response options, which supported accessibility but may still be influenced by response style and comprehension. Finally, although participants were drawn from diverse metropolitan and regional locations, findings may not fully generalise to adults with intellectual disability in other Australian states and territories where service systems and regional disadvantage differ. Conclusions In summary, this study highlights three key implications for improving digital inclusion among adults with intellectual disability: the importance of equitable access to digital devices and connectivity, the role of skilled support in developing digital confidence, and the potential value of interventions that strengthen positive beliefs about internet use. These findings reinforce the importance of tailored, person‐centred supports to promote digital participation and autonomy. Importantly, the study highlights that internet self‐efficacy is significantly higher when individuals receive support from skilled IT professionals, underscoring the value of targeted digital inclusion initiatives. By demonstrating how both environmental and psychological factors contribute to digital inequality, this research offers practical insights for professionals across disciplines, including educators, support workers, policy makers and technologists, working to enhance digital access and quality of life. Further investigation into the role of caregivers and support strategies is critical to ensuring people with intellectual disability are not only connected, but meaningfully included in the digital world. Funding This research was supported by a PhD scholarship awarded by the Australian National Disability Insurance Agency and administered by the University of the Sunshine Coast (1068349). Ethics Statement This study was approved by the Human Research Ethics Committee at the University of the Sunshine Coast (Approval Numbers: S171089 and S171141) and was conducted in accordance with the ethical standards of the Committee and the principles outlined in the Declaration of Helsinki. All participants, including those with intellectual disability and, where applicable, their legal guardians, gave informed consent prior to participation in the study. Conflicts of Interest The authors declare no conflicts of interest. Acknowledgements The authors sincerely thank the adults with intellectual disability and their caregivers who generously gave their time to participate in the research. Their insights and contributions were invaluable to the study and are deeply appreciated. Open access publishing facilitated by University of the Sunshine Coast, as part of the Wiley ‐ University of the Sunshine Coast agreement via the Council of Australasian University Librarians. Data Availability Statement The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions. References Associated Data Data Availability Statement The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.