Effect of digital literacy on social entrepreneurial intentions and nascent behaviours among students and practitioners in mass communication

Ethical business practices, sustainability, and social responsibility have become central concerns for governments, corporations, and the public, making corporate social responsibility and ethical standards essential within mass communication sectors such as public relations and journalism. Social entrepreneurship has emerged as a promising avenue to create social value via market-based activities rather than relying on donations, addressing persistent social problems (e.g., poverty, unemployment, climate change). Prior research on social entrepreneurial intentions (SEI) has mainly assessed antecedents per the theory of planned behaviour, often assuming intentions translate into behaviours, leaving nascent social entrepreneurial behaviours underexplored. However, intentions may not always lead to action due to constraints such as doubt, aversion, and contextual barriers. This study therefore examines determinants of both SEI and social entrepreneurial behaviours, incorporating digital literacy and perceived social awareness of peers as predictors. The goals are: (1) revisit determinants of SEI, (2) add digital literacy and peers’ social awareness as predictors of SEI, (3) test the effect of SEI on social entrepreneurial behaviours, and (4) evaluate robustness across two samples: university students and mass communication practitioners in Taiwan.

Rationale for focusing on mass communication sectors: Mass communication spans strategic communication (public relations, advertising) and news media (journalism). Students in news media tend to be motivated by altruism, while strategic communication students focus more on revenue, aligning with social entrepreneurship’s dual social–economic value creation. Given industry shifts and layoffs, alternative career paths such as social entrepreneurial journalism are pertinent to sustain reporting on underserved issues. Digital era changes: Social media and digital technologies lower entry barriers for media startups, transform roles (e.g., freelancers replacing full-time journalists), and require new editorial and business models. Social entrepreneurial journalism advocates novel models to cover underreported social issues sustainably. Theoretical foundation: Grounded in the theory of planned behaviour (attitudes, perceived behavioural control, subjective norms) and social cognitive career theory (self-efficacy, outcome expectations). Prior models operationalise empathy, perceived social support, self-efficacy, and moral obligation; extend with experience with social problems as antecedents to these cognitions. Social cognitive career theory highlights dynamic self-efficacy shaped by experiences, plus contextual factors (education, role models). Intention–behaviour gap: Many SEI studies assume intentions lead to action; yet translation depends on skills, resources, opportunities, time, and effort—particularly demanding in social entrepreneurship (requiring financial, managerial, and problem-solving capabilities). Hence testing SEI’s effect on behaviours is essential. Digital literacy: Defined as the ability to use digital artefacts, platforms, and infrastructure for innovation and business operations, with social, technical, and cognitive dimensions. It supports stakeholder interaction, information processing, opportunity recognition, efficiency, and innovation—key for social entrepreneurs in digitally mediated media work. Hypotheses: H1 social support → (a) SEI, (b) behaviours; H2 self-efficacy → (a) SEI, (b) behaviours; H3 outcome expectations → SEI; H4 experience → SEI via social support, self-efficacy, outcome expectations (indirect); H5 peers’ social awareness → SEI; H6 digital literacy → (a) SEI, (b) behaviours; H7 SEI → behaviours.

Design and analysis: Cross-sectional web-based survey with structural equation modelling (SEM) and confirmatory factor analysis (CFA) using Amos 27.0 and maximum likelihood estimation. Separate structural models were estimated for students and practitioners to test robustness. Sampling and procedure: Targeted university students and practitioners in mass communication in Taiwan. The online survey was distributed via forums, Facebook, and email from March 18–27, 2022. Attention checks were included. Taiwan’s Internet penetration (85–91% in 2022) supports online reach (except older adults). Sample: N=814 valid responses (students n=373; practitioners n=441). Students: 46.1% men, 53.9% women; median age 22; regional distribution across Taiwan; 11.8% postgraduates. Practitioners: 49.9% men, 50.1% women; median age 36; education mostly undergraduate (72.8%) or postgraduate (17.2%). Measures: Experience with social problems, perceived social support, and social entrepreneurial self-efficacy (3 items each; Hockerts, 2017). Outcome expectations (3 items; Ip et al., 2021). Perceived social awareness of peers (4 items; Dinev & Hart, 2005; Lichtenstein et al., 2004). Digital literacy as a second-order construct (9 items total; 3 each for social, technical, cognitive dimensions; Neumeyer et al., 2021). SEI (4 items; Ip et al., 2017). Social entrepreneurial behaviours (5 items; Ip et al., 2022). Control variables: age, sex, education level, household income. All on 6-point Likert scales (1=strongly disagree to 6=strongly agree). Measurement model and validity: CFA showed acceptable fit for the pooled sample: χ²/df=2.87, RMSEA=0.05, SRMR=0.05, CFI=0.94, TLI=0.93. Standardised loadings >0.50 and composite reliabilities ≥0.74. AVE mostly ≥0.50; self-efficacy AVE=0.49 but acceptable given CR>0.70. Discriminant validity met via HTMT (0.28–0.84 <0.85/0.90). Structural models fit: Practitioners χ²/df=2.27, CFI=0.92, TLI=0.90, RMSEA=0.05, SRMR=0.06; Students χ²/df=2.02, CFI=0.91, TLI=0.90, RMSEA=0.05, SRMR=0.06.

Measurement validity: CFA acceptable (χ²/df=2.87; RMSEA=0.05; SRMR=0.05; CFI=0.94; TLI=0.93). CRs ≥0.74; AVEs mostly ≥0.50 (self-efficacy AVE=0.49, acceptable with CR>0.70). Discriminant validity achieved (HTMT 0.28–0.84). Structural model fits acceptable for both samples. Practitioners (n=441): • Perceived social support → SEI: β=0.36, SE=0.12, p<0.001 (H1a supported); → Behaviours: β=0.10, SE=0.09, p>0.05 (H1b rejected). • Social entrepreneurial self-efficacy → SEI: β=0.01, SE=0.19, p>0.05 (H2a rejected); → Behaviours: β=-0.03, SE=0.12, p>0.05 (H2b rejected). • Outcome expectations → SEI: β=-0.07, SE=0.10, p>0.05 (H3 rejected). • Experience → social support: β=0.76, SE=0.10, p<0.001; → self-efficacy: β=0.89, SE=0.09, p<0.001; → outcome expectations: β=0.58, SE=0.08, p<0.001. Indirect effect on SEI via mediators: β=0.24, SE=0.09, p<0.05, 95% CI [0.04, 0.41] (H4 supported). • Peers’ social awareness → SEI: β=0.22, SE=0.11, p<0.01 (H5 supported). • Digital literacy → SEI: β=0.31, SE=0.11, p<0.001 (H6a supported); → Behaviours: β=0.38, SE=0.09, p<0.001 (H6b supported). • SEI → Behaviours: β=0.53, SE=0.05, p<0.001 (H7 supported). • Controls: age, sex, education, income not significant. Students (n=373): • Perceived social support → SEI: β=0.25, SE=0.16, p<0.001 (H1a supported); → Behaviours: β=0.05, SE=0.12, p>0.05 (H1b rejected). • Social entrepreneurial self-efficacy → SEI: β=0.27, SE=0.25, p<0.01 (H2a supported); → Behaviours: β=-0.13, SE=0.19, p>0.05 (H2b rejected). • Outcome expectations → SEI: β=-0.08, SE=0.12, p>0.05 (H3 rejected). • Experience → social support: β=0.73, SE=0.08, p<0.001; → self-efficacy: β=0.88, SE=0.08, p<0.001; → outcome expectations: β=0.49, SE=0.06, p<0.001. Indirect effect on SEI via mediators: β=0.38, SE=0.10, p<0.001, 95% CI [0.20, 0.58] (H4 supported). • Peers’ social awareness → SEI: β=0.16, SE=0.09, p<0.05 (H5 supported). • Digital literacy → SEI: β=0.16, SE=0.10, p<0.05 (H6a supported); → Behaviours: β=0.36, SE=0.08, p<0.001 (H6b supported). • SEI → Behaviours: β=0.68, SE=0.06, p<0.001 (H7 supported). • Controls: men exhibited stronger SEI (β≈-0.13, coded such that men higher, p<0.01); higher education level associated with weaker SEI (β≈-0.11, p<0.01); household income positively affected behaviours (β=0.08, p<0.05). Overall: Across both samples, digital literacy and peers’ social awareness positively predict SEI; digital literacy and SEI positively predict behaviours; experience exerts a positive indirect effect on SEI; outcome expectations are not significant; social support aids SEI but not behaviours.

Findings address the research questions by demonstrating that beyond classical socio-cognitive drivers, digital literacy is a pivotal capability shaping both intentions and nascent social entrepreneurial actions among mass communication students and practitioners. Social support from networks strengthens intentions but does not directly translate into action, highlighting that support is necessary yet insufficient without operational capabilities. Self-efficacy predicts SEI among students but not practitioners, suggesting practitioners may regard problem-solving as a baseline competency, requiring more than confidence to initiate ventures. Outcome expectations did not predict SEI, possibly due to stronger orientation toward extrinsic rewards and pragmatic concerns in strategic communication roles. Experience with social issues strengthens SEI indirectly by enhancing self-efficacy, outcome expectations, and perceived support. The robust SEI → behaviour link confirms TPB’s mechanism in the social entrepreneurship domain and underscores the importance of interventions that both cultivate intentions and remove barriers to action. Relevance: Results inform media schools and industry by emphasising digital capabilities (social, technical, cognitive) as levers for fostering social entrepreneurial career paths and reinforcing the role of socially aware peer environments in shaping SEI.

Analysing Taiwanese mass communication students and practitioners, the study shows that perceived social support, peers’ social awareness, and digital literacy strengthen social entrepreneurial intentions, while digital literacy and intentions drive nascent social entrepreneurial behaviours. Digital literacy emerges as a critical driver of both cognition and action, indicating the value of educating for social, technical, and cognitive digital competencies for stakeholder engagement and decision-making. The comparatively weaker roles of self-efficacy (among practitioners) and outcome expectations suggest pragmatic capability-building may be more impactful than shaping favourable perceptions alone. Contributions: (1) Identifies digital literacy as central to fostering SEI and behaviours; (2) focuses on mass communication as a context facing digital disruption and offers curricular implications (e.g., entrepreneurial journalism, digital applications, social innovation/CSR communication); (3) empirically establishes the SEI → behaviour pathway in social entrepreneurship. Future research should test generalisability across countries and industries and examine social intrapreneurship within organisations.

Generalizability is limited by the focus on Taiwanese mass communication students and practitioners; results may not extend to other industries or countries. The study primarily considers social entrepreneurship as an alternative career path, not social intrapreneurship within organisations, which merits further investigation. Cross-sectional, self-report data limit causal inference; although SEM with validated measures was used, longitudinal or experimental designs could better capture intention–behaviour dynamics. Online sampling may underrepresent those with low Internet use, though national penetration is high.