Quality of instructor, fear of COVID-19, and students' anxiety as predictors of student satisfaction and academic effort in online classes

The study addresses how human factors—specifically perceived quality of instructor and students’ psychological states (fear of COVID-19 and anxiety)—relate to student satisfaction with online classes and students’ academic behavior (effort/performance) during the COVID-19 pandemic. In the context of intensified competition and the centrality of student satisfaction as a key service quality outcome in higher education, online learning’s rapid expansion during pandemic disruptions heightened the need to understand determinants of satisfaction and academic outcomes. Prior work emphasized technological and institutional factors, but there was limited insight into instructor quality and students’ mental health impacts. The research aims to examine whether instructor quality increases satisfaction, whether satisfaction enhances academic effort, and how fear of COVID-19 and anxiety interplay and potentially affect satisfaction and effort in an online learning context in Croatia.

The theoretical framework posits instructor quality as pivotal for student satisfaction in online classes (communication, feedback, pedagogical/technological skills). Student satisfaction is linked to motivation, engagement, and performance. The COVID-19 context exacerbated mental health issues (anxiety, depression) and introduced fear-related stressors that could impact academic behavior. Prior studies show mixed effects of anxiety on performance and satisfaction, and the Fear of COVID-19 Scale (FCV-19S) relates fear to anxiety/depression. Hypotheses: H1: Quality of instructor positively affects student satisfaction. H2: Student satisfaction positively affects students’ academic effort/performance. H3: Satisfaction mediates the relationship between instructor quality and academic effort/performance. H4: Fear of COVID-19 positively affects student anxiety. H5: Anxiety negatively affects academic effort/performance. H6: Anxiety mediates the relationship between fear of COVID-19 and academic effort/performance. H7: Anxiety negatively affects satisfaction with online classes.

Design and context: Cross-sectional descriptive study conducted in 2022 while COVID-19 restrictions persisted in Croatia. Participants: N=359 undergraduate students from two Croatian higher education institutions (one public, Croatian-language; one private, English-language). Gender: 40.4% male, 59.3% female. Age: 18–20 (25.9%), 21–23 (59.1%), ≥24 (15.0%). Year levels 1–5 represented. Recruitment targeted ~340 responses (10× observed variables). Data collection: Online questionnaire, administered in English or Croatian according to program language. Pretesting involved a representative student group to ensure clarity. Ethical approval granted by RIT Croatia IRB (No. 2022-04-28); informed consent obtained. Measures: - Quality of instructor (QUALI): 8 items on a 5-point Likert scale adapted from Gopal et al. (2021), including prompt feedback item; α=0.91. - Anxiety (STAI trait short form): Items from Spielberger (1983) with attention to short-form validation literature; final 6 items on a 4-point frequency scale; α=0.86 after item removals due to low loadings/reverse-scoring issues. - Student satisfaction (SA): 7 items on a 5-point Likert scale (Bangert, 2004; Wilson et al., 1997; Gopal et al., 2021); after item removal, 5 items; α=0.94. - Fear of COVID-19 (FCOV): Initially 5 items from FCV-19S (Ahorsu et al., 2022; Bitan et al., 2020; Zolotov et al., 2022); refined to 3 items capturing ‘emotional fear reactions to COVID-19’; α=0.85. - Academic effort and performance (ACEP): Initially 4 items adapted from Tolken (2011); due to factor structure, focused on Academic Effort (2 items; α=0.85). Data screening and assumptions: Incomplete responses (>15% missing) removed; six unengaged responses excluded; final N=359 had no missing data. Outliers checked via Cook’s distance (none >1). Skewness/kurtosis within acceptable thresholds. Multicollinearity not detected (VIF<4). Common method bias assessed by Harman’s single-factor test (largest factor 33.31% <50%). Analytic strategy: - EFA (ML extraction, Promax rotation; SPSS) guided item reduction: removed items with low extraction (<0.40) or cross-loadings. For STAI, reverse-scored items exhibited separate factor; retained 6 nonproblematic items. ACEP refined to Academic Effort (ACE1–ACE2). FCOV refined to ‘emotional fear reactions’. - EFA results (refined model): KMO=0.907; χ²=5448.95, df=276, p<0.001; five-factor solution; 63.14% variance explained; loadings 0.615–0.935. - CFA (AMOS): Good fit for refined measurement model: χ²=458.649; df=242; χ²/df=1.895; TLI=0.953; CFI=0.959; RMSEA=0.050 (90% CI [0.043, 0.057]); SRMR=0.047. Reliability/validity: Cronbach’s α and composite reliability (CR) for all constructs ≥0.85; AVE≥0.50; Fornell-Larcker discriminant validity satisfied. - SEM (AMOS): Structural model fit: χ²/df=1.923; TLI=0.952; CFI=0.958; RMSEA=0.051 (90% CI [0.044, 0.058]); SRMR=0.062. Hypotheses updated to reflect refined variables: H1 (QUALI→Satisfaction), H2 (Satisfaction→Academic Effort), H3 (Satisfaction mediates QUALI→Academic Effort), H4 (Emotional fear→Anxiety), H5 (Anxiety→Academic Effort), H6 (Anxiety mediates Emotional fear→Academic Effort), H7 (Anxiety→Satisfaction). Mediation tested via indirect effects and bootstrapped confidence intervals.

- Sample and measurement: N=359 students; validated five-construct measurement model with strong reliability (α/CR ≥0.85) and validity (AVE ≥0.50). - Model fit: SEM fit indices indicated good fit (χ²/df=1.923; TLI=0.952; CFI=0.958; RMSEA=0.051; SRMR=0.062). - Direct effects: • Quality of instructor → Student satisfaction: β=0.695, t=12.550, p<0.001 (H1 supported). • Student satisfaction → Academic effort: β=0.373, t=4.711, p<0.001 (H2 supported). • Emotional fear reactions to COVID-19 → Anxiety: β=0.368, t=5.744, p<0.001 (H4 supported). • Anxiety → Academic effort: β=-0.045, t=-0.750, p=0.454 (H5 not supported). • Anxiety → Student satisfaction: β=-0.060, t=-1.354, p=0.176 (H7 not supported). - Mediation: • Student satisfaction fully mediated Quality of instructor → Academic effort: indirect β=0.369, t=4.011, p<0.001; direct β=0.215, p>0.05 (H3 supported; full mediation). • Anxiety did not mediate Emotional fear reactions → Academic effort: indirect β=-0.017, t=-0.654, p>0.05 (H6 not supported). - Correlational validity (Table 4): Key correlations included positive links between quality of instructor and satisfaction (r≈0.70) and between satisfaction and academic effort (r≈0.49); emotional fear correlated with anxiety (r≈0.37). - Scale performance: All constructs demonstrated satisfactory reliability (e.g., QUALI α=0.911; Satisfaction α=0.942; Anxiety α=0.860; Emotional fear α=0.851; Academic effort α=0.853).

Findings confirm that instructor quality strongly elevates student satisfaction in online classes, and satisfaction, in turn, increases students’ academic effort. Instructor influence on effort operates fully through satisfaction, underscoring the centrality of instructional communication, prompt feedback, accessibility, enthusiasm, and respectful interactions for engaging students in coursework. Emotional fear reactions to COVID-19 increased student anxiety, yet neither fear-induced anxiety nor anxiety itself translated into reduced academic effort or lower satisfaction. Possible explanations include temporal adaptation to pandemic conditions, lower fear levels among university students relative to the general population, and increased motivation/optimism as students adjusted to online learning. The results add nuance to mixed prior evidence on the performance effects of anxiety and suggest that, in this context and period, academic engagement was more sensitive to instructional quality and resultant satisfaction than to anxiety or fear. These insights emphasize targeting instructional practices to sustain student engagement during disruptions and beyond.

This study contributes by demonstrating that: (1) perceived quality of instructor is a strong predictor of student satisfaction in online classes; (2) satisfaction positively drives academic effort; and (3) satisfaction fully mediates the effect of instructor quality on effort. Although emotional fear reactions to COVID-19 increased anxiety, neither fear nor anxiety significantly affected satisfaction or academic effort in this sample. Practical implications include prioritizing instructor development in communication, feedback, accessibility, and student-centered practices to sustain satisfaction and effort in online learning, particularly under disruption. Future research should pursue longitudinal designs, cross-cultural replications, and inclusion of additional stakeholders (e.g., instructors, administrators), and explore other personal and contextual moderators to clarify when and how anxiety and fear affect academic outcomes.

- Cross-sectional design precludes causal inference and tracking changes over time. - Sample limited to students from two universities in Croatia; generalizability is constrained despite inclusion of international students in programs. - Perspectives of other stakeholders (e.g., faculty, administrators) were not included. - Limited sociodemographic/health-related variables (e.g., physical and mental health history) that might influence coping and academic behavior were not measured. - Measurement adjustments (item removals, short forms) may affect comparability with full-scale implementations, though reliability and validity were adequate.