Study Demands and Resources in Distance Education—Their Associations with Engagement, Emotional Exhaustion, and Academic Success

Abstract

Distance learning offers enhanced flexibility and reduced access restrictions, making it increasingly popular among non-traditional students and those juggling academic studies with professional and family obligations. This study explored the associations between study demands and resources (decision latitude and social support from lecturers and peers) and different study outcomes by applying the Job Demands-Resources Model in a distance learning context. Based on the model’s assumptions, we hypothesized that academic demands negatively predict study success in distance learning, while decision latitude and social support from lecturers and peers positively affect it. These associations were expected to be mediated by emotional exhaustion and different dimensions of engagement. The cross-sectional online study involved 286 psychology students from a German distance university. The multivariate path model revealed an association of demands and decision latitude with perceptions of competence and study satisfaction. While demands were significantly correlated with the grade point average, decision latitude was not. Consistent with the model’s assumptions, these effects were partially mediated by exhaustion and engagement. We did not find significant incremental associations of social support with the outcomes. The findings concerning measures to support students in distance education were discussed.

1. Introduction

The digital revolution in higher education, driven by rapid technical innovation and global challenges and promoted by the coronavirus pandemic, has accelerated in recent years, fundamentally reshaping how knowledge is delivered and acquired. The use of distance learning or online learning material is an ongoing trend, with the advantages of synchronous courses, e.g., by video conferences, and asynchronous lessons, e.g., by recorded or written material. For example, one-third of all EU internet users were engaged in online learning in 2024 (Eurostat, 2025). In 2021, 28% of all enrolled undergraduate students in the United States were enrolled in distance learning courses exclusively (National Center for Education Statistics, 2023)

As distance learning usually offers more openness, accessibility, and flexibility in contrast to traditional face-to-face education, it is highly attractive for so-called non-traditional students. These students often start their studies at a higher age, have a professional occupation during enrollment, a discontinuous biography, or are studying part-time (e.g., Wolter, 2012; Chung et al., 2014). While all distance students must deal with higher demands for self-organization and self-regulation (e.g., Bothma & Monteith, 2004), non-traditional students in distance education often struggle with additional obligations from job and family (e.g., Kohler Giancola et al., 2009), resulting in a higher risk for dropout (e.g., Choy, 2002; Stoessel et al., 2015). Hence, distance education programs offer an opportunity to non-traditional students to overcome inequalities driven by sociodemographic characteristics, but these students are still at higher risk for failure.

Several studies have explored individual predictors for success in distance education settings, such as the use of learning strategies (Neroni et al., 2019) or personal goals (Stoessel et al., 2015). However, still a gap in the literature is the limited research on how subjectively perceived study conditions affect student motivation and success. For example, Chen and Jang (2010) found correlations between contextual support and online learners’ motivation, and Ghasempour et al. (2023) reported positive correlations between learning environment factors (like instructor support, student interaction, and student autonomy) and different aspects of academic success, such as general academic skills. The lack of research is particularly noteworthy because understanding these perceived conditions could allow institutions to tailor more supportive learning environments that benefit a broader range of students. Therefore, the present study will shed light on the academic outcomes of study demands and different study resources in higher distance education.

1.1. The Job Demands-Resources Model and the Study Demands-Resources Framework

In the occupational context, Demerouti et al. (2001) established the Job Demands-Resources model (JD-R) to explain the effects of demands like time pressure or high task complexity and resources of the work environment like decision latitude or feedback on different organizational and individual outcomes. The model predicts that high job demands and a lack of job resources result in high strain or burnout (health impairment process) that, in turn, predict negative outcomes like absenteeism or job change. On the other hand, job resources can initiate engagement (motivational process) that in turn predicts positive outcomes like high performance, commitment, or job satisfaction (Figure 1). Hence, the model considers both the pathogenic effects of job characteristics and the salutogenic effects of job resources on motivation. It also allows for testing a broad range of different demands and resources regarding their effect on job outcomes. Many studies confirmed the model’s basic assumptions, and some of them also found evidence for the cross paths with resources reducing burnout and demands having a negative impact on engagement (see Bakker et al., 2023).

Since work and higher education share many characteristics like time-bound, measurable goals and a comparable workload and outcomes such as achievement, satisfaction, commitment, and well-being (Cilliers et al., 2018; Cotton et al., 2002; Ouweneel et al., 2011), the JD-R has also been applied to the academic context for traditional students in face-to-face educational settings, leading to the proposition of the Study Demands-Resources-Framework (Lesener et al., 2020). Sieverding et al. (2013) showed that both high demands and poor control were predictive of low study satisfaction. It was also demonstrated that high study demands, and a lack of resources can promote the occurrence of burnout (Mokgele & Rothmann, 2014; Robins et al., 2015). Gusy et al. (2016) and Lesener et al. (2020) found evidence for the health impairment path as well as for the motivational path and their effect on students’ well-being. Since the evidence for the effects of different demands and resources on engagement and burnout as well as on different student outcomes was almost exclusively found for face-to-face education (for an overview see Bakker & Mostert, 2024), the present study extents this literature by focusing on (non-traditional) students and their success in distance education.

1.2. Study Demands and Resources in Distance Education

Study demands in face-to-face settings are usually assessed by psychological demands (e.g., Schmidt et al., 2019), difficulty of the learning material, time pressure, or incompatibility of studies and private life (e.g., Gusy et al., 2016; Lesener et al., 2020), which also seem to be relevant demands in distance education. Some resources, however, might play a more pronounced role in distance settings than in face-to-face study (e.g., Akpen et al., 2024), such as decision latitude and social support.

In the occupational context, decision latitude is considered a fundamental resource that is often described as equivalent to or as an important part of autonomy (Gagné & Deci, 2005). Decision latitude consists of two theoretically distinct subdimensions (i.e., Hackman & Oldham, 1975), i.e., decision authority that refers to the extent of making own decisions and skill discretion describing the extent to which the task requires different skills and creativity. Schmidt et al. (2019) found that decision latitude was associated with lower stress and higher study satisfaction. Due to the many additional obligations that non-traditional distance students need to balance with their studies, decision latitude that allows flexibility concerning time and place of studying and the selection of learning strategies might be a particularly important resource in this student population.

Social support comprises material and psychological resources to help the recipient cope with stressful events and enhance engagement (Cohen, 2004). In the educational context, social support has many sources ranging from the personal community (like support from peers) to the course community (like support from lecturers) (Borup et al., 2020). There is evidence that instructors’ and peers’ support directly affect students’ engagement, achievement, and well-being (Cilliers et al., 2018; Nielsen et al., 2016). Because in distance education, student-student and student-lecturer interactions are not as self-evident and immediate as in face-to-face study, it is highly informative for designing online learning settings if the provision of social support achieves the same effects in distance education. Ghasempour et al. (2023) were able to find evidence that, even in the online context, social support from instructors plays a role in academic success. However, the participating students were only active in an online setting for two semesters during the pandemic and had already completed two semesters of face-to-face learning beforehand. This prior experience may have allowed them to establish social contacts during the initial face-to-face phase, which could then be successfully maintained online. This possibility does not exist in pure distance learning environments and could be a relevant difference.

Concerning student engagement, most of the literature on the study demand-resources framework and parts of it, relies on the Utrecht Work Engagement-Scale for students (UWES-S, Schaufeli et al., 2002) that was originally developed for the occupational context and conceptualizes engagement as a conglomerate of vigor, dedication, and absorption. In the school context, however, the most widely used model of engagement by Fredricks et al. (2004) distinguishes emotional, cognitive, and behavioral engagement. Whereas emotional engagement includes aspects like enjoyment and positive emotions, cognitive engagement regards planning, investing, self-perception, or self-regulation. Behavioral engagement includes participation and time-on-task, but also effort and persistence. In line with this conceptualization, the present study distinguishes these different aspects of academic engagement.

Finally, academic success comprises both institutional or objective criteria, like grades or duration of study, and subjective criteria like study satisfaction or the experience of competence (e.g., Rindermann & Oubaid, 1999). Hence, we considered both grade point average (GPA) as well as study satisfaction and perceived professional competence as important outcomes in higher distance education.

Based on these assumptions and previous empirical findings, we derived the following hypotheses:

H1. 

Study intensity negatively predicts academic success, mediated by higher emotional exhaustion and lower engagement.

H2. 

Decision latitude positively predicts academic success, mediated by lower emotional exhaustion and higher engagement.

H3. 

Social support from lecturers positively predicts academic success, mediated by lower emotional exhaustion and higher engagement.

H4. 

Social support from peers positively predicts academic success, mediated by lower emotional exhaustion and higher engagement.

2. Materials and Methods

2.1. Sample

286 psychology students of Germany’s only public distance university, which is also Germany’s largest university concerning student numbers (for more information see Stürmer et al., 2018), participated in a voluntary online questionnaire in the winter term of 2020. 206 students were female, 217 worked in addition to their studies, and about 40% worked more than 20 h/w. Students’ ages ranged from under 20 to over 60 years, about 50% were older than 30 years (for more details, see

Table 1. Descriptive Statistics for the sample.

).

2.2. Materials

Table 2. Details for the scales used.

Scales	M	SD	α	Range	Items	Example Items	Source
Study intensity	2.7	0.6	0.86	1–4	6	My distance learning requires a lot of effort from me.	Richter et al. (2000)
Decision latitude	2.8	0.5	0.63	1–4	6	I can organize and schedule my distance learning on my own.	Richter et al. (2000)
Support from lecturers	4.0	1.2	0.93	1–6	3	The lecturers are cooperative and open-minded.	Schaeper and Weiß (2016)
Support from peers	4.2	1.0	0.80	1–6	3	In general, students support each other.	Schaeper and Weiß (2016)
Emotional exhaustion	2.4	0.6	0.83	1–4	8	There are days when I feel dull even before I start with my study tasks.	Reis et al. (2015)
Emotional engagement	3.9	1.0	0.87	1–7	10	I am completely absorbed in what I am doing.	Rheinberg et al. (2003)
Cognitive engagement	3.5	0.7	0.72	1–5	6	I ask myself questions about the material to check whether I have understood everything.	Klingsieck (2018)
Behavioral Engagement	4.1	0.7	0.67	1–5	3	I don’t give up, even if the subject matter is difficult or complex.	Klingsieck (2018)
Competence	3.4	0.8	0.85	1–5	6	I can reproduce important terms/concepts of my distance learning so far.	Braun et al. (2008)
Study satisfaction	7.0	1.90	-	1–10	1	Overall, I am satisfied with my studies.	Self-constructed

presents all scales and their measurement properties as well as descriptive statistics.

To assess study demands, we adapted the work intensity scale of the questionnaire about the experience of work intensity and decision latitude (FIT) by Richter et al. (2000) to the study context. The six items were measured on a Likert scale ranging from “does not apply” (1) to “applies” (4). A higher mean score indicates a higher level of perceived study intensity.

To assess decision latitude, we adapted the decision latitude scale with the subdimensions skill discretion and decision authority of the FIT questionnaire by Richter et al. (2000) to the study context. Whereas skill discretion refers to the use and enhancement of different skills, decision authority comprises the extent of making own decisions. The seven items were measured on a Likert scale ranging from “does not apply” (1) to “applies” (4). A higher mean score indicates a higher level of perceived decision latitude.

To assess social support from lecturers and social support from peers, we used two subscales of the National Education Panel Study (NEPS, Schaeper & Weiß, 2016) about rapport with lecturers and peers, respectively. The three items on each scale were measured on a Likert scale ranging from “does not apply at all” (1) to “fully applies” (6). A higher mean score indicates a higher level of perceived support.

To capture the students’ emotional exhaustion, we used the respective subscale of the Oldenburg Burnout Inventory in the student version (OLBI-S; Reis et al., 2015). The eight items were measured on a Likert scale ranging from “strongly disagree” (1) to “totally agree” (4). A higher mean score indicates a higher level of emotional exhaustion.

For emotional engagement, we used the “straight automated process” and “absorption” subscales of the Flow Short-Scale by Rheinberg et al. (2003). The third subscale, “concern,” was omitted due to its similarity to “exhaustion”. The ten items were measured on a Likert scale ranging from “does not apply” (1) to “applies” (7). A higher mean score indicates a higher level of emotional engagement.

For cognitive engagement, we used the two subscales “control” and “regulation” of the questionnaire for measuring learning strategies of students (LIST) by Wild and Schiefele (1994) in the adapted version of Klingsieck (2018). These subscales represent important parts of cognitive and metacognitive engagement, especially in the distance learning context, where requirements for self-regulation and self-organization are particularly high (e.g., Bothma & Monteith, 2004). The six items were measured on a Likert scale ranging from “very rare” (1) to “very often” (5). A higher mean score indicates a higher level of cognitive engagement.

For behavioral engagement, we also used the subscale “effort” from the LIST questionnaire in the adapted version of Klingsieck (2018). The three items were measured on a Likert scale ranging from “very rare” (1) to “very often” (5). A higher mean score indicates a higher level of behavioral engagement.

The grade point average (GPA) was surveyed in categories ranging from “3.6–4” (6, worst) to “1.0–1.5” (1, best). For the distribution of different grades, see

Table 3. Frequencies of different GPAs.

Variable	Value	N	%
GPA	6 = 1.0–1.5 (best)	12	4.2%
	5 = 1.6–2.0	49	17.3%
	4 = 2.1–2.5	80	28.3%
	3 = 2.6–3.0	69	24.4%
	2 = 3.1–3.5	58	20.5%
	1 = 3.6–4.0 (worst)	15	5.3%

Note. N = 285.

.

Study satisfaction was measured with a single item, “Overall, I am satisfied with my studies” on a scale ranging from 1 to 10, where a higher mean indicates a higher level of satisfaction (as suggested, e.g., by Damrath, 2006). In the German context, the “Fragebogen zur Studienzufriedenheit” (Schiefele & Jacob-Ebbinghaus, 2006; Westermann et al., 1996) dominates, yet its scale development was not firmly grounded in psychological theory. As a result, its subscales—satisfaction with study content, study conditions, and coping with study-related stress—lack strong theoretical justification. According to Kegel et al. (2025), a single-item measure of overall study satisfaction is not only efficient but also similarly correlated with other study-relevant variables as the corresponding 3-item or 5-item scale.

For the perceived professional competence, we used the subscale about professional competence of the Berlin Evaluation Instrument for self-evaluated student competences (BEvaKomp, Braun et al., 2008). The six items were measured on a Likert scale ranging from “does not apply” (1) to “completely applies” (5). A higher mean score indicates a higher level of perceived competence.

All scales showed high to acceptable levels of homogeneity, with the lowest scores for behavioral engagement (α = 0.67) and decision latitude (α = 0.63), which includes items for both skill discretion and decision authority.

2.3. Data Analysis

Analyses were done with Mplus 7.4 (Muthén & Muthén, 2015) applying the maximum likelihood method. We fitted a saturated manifest path model that included direct paths from study intensity, decision latitude, and social support from lecturers and peers on GPA, perceived competence, and study satisfaction, as well as indirect effects via emotional exhaustion and the three dimensions of engagement. Because the proof of the significance of indirect effects uses a product of two or more regression coefficients, this product often misses normal distribution. Therefore, we used bias-corrected bootstrapping with 10,000 bootstraps and confidence intervals as recommended (e.g., MacKinnon, 2008).

3. Results

Means and standard deviations of the scales are provided in Table 2. The sample mean for study intensity is approximately at the scale mean. The sample means for the three resources, decision latitude, social support from lecturers and peers, are slightly above the scale mean. However, for the two forms of support, the variance around the mean is relatively high. While the sample means for the mediators, emotional exhaustion and emotional engagement, are also approximately at the scale mean, those for cognitive engagement and particularly behavioral engagement are substantially above it. The sample mean for the subjective study success criterion competence is slightly above the scale mean, whereas the value for study satisfaction is substantially above it. Correlations of the study variables are reported in

Table 4. Intercorrelations.

.

The correlation analyses revealed many significant associations between the predictors, mediators, and outcomes. Higher study intensity was significantly correlated with more emotional exhaustion and less emotional engagement. Furthermore, higher intensity was negatively correlated with all three academic outcomes. The resource decision latitude was significantly positively correlated with all three engagement variables, but not with exhaustion. Additionally, it was positively associated with the two subjective student outcomes, study satisfaction and competence. More support from lecturers was significantly positively correlated with higher cognitive and behavioral engagement. Regarding the outcomes, it was only associated with higher satisfaction. More support from peers was correlated with more cognitive and behavioral engagement, but not with emotional engagement or emotional exhaustion. Furthermore, it was positively associated with the subjective academic outcomes. Emotional exhaustion and all engagement variables were significantly correlated with the subjective study outcomes as well as with the GPA. Finally, the low to moderate correlations among the different resources, the different measures of engagement, and the different measures of academic success support the idea of investigating their unique effects rather than composite scores.

The multivariate model considering all variables simultaneously is presented in Figure 2. The total effects are given in

Table 5. Fully standardized total effects.

, and the indirect effects with confidence intervals are presented in

Table 6. Fully standardized indirect effects of the mediating variables.

. The predictors explained a substantial amount of variance in academic success with R² of 33.2% and 28.4% for the subjective measures of study satisfaction and professional competence, respectively, and R² = 10.3% for objective achievement as measured by GPA. Also, study intensity, resources, and covariates explained substantial amounts of variance in the mediator’s emotional exhaustion (24.8%), and the three engagement measures emotional (9.0%), cognitive (9.3%), and behavioral (8.0%). The correlations between the outcomes were small, and those between the independent variables and between the mediators were moderate at around .30. Only the correlation between emotional exhaustion and emotional engagement was larger, with .51.

Looking at the a-paths of the multivariate mediation model, study intensity was significantly associated with all mediator variables, with the strongest association with emotional exhaustion. Of all three resources, only decision latitude was significantly predictive of the mediator variables. Social support from peers and lecturers was not incrementally predictive of the mediator variables in the multivariate model.

Concerning the associations between the mediator variables and the outcomes (b-paths), emotional exhaustion significantly predicted lower study satisfaction, emotional engagement positively predicted all study outcomes, and cognitive and behavioral engagement only predicted the subjective experience of professional competence.

Finally, we tested the twelve indirect effects that included significant a- and b-paths—hence, six indirect effects for study intensity and decision latitude each.

Study intensity had a significant negative total and indirect effect on study satisfaction via higher emotional exhaustion and lower emotional engagement, as well as a remaining significant direct effect. This means that the effect of study intensity on study satisfaction was partially mediated by emotional exhaustion and emotional engagement. Study intensity also had a significant negative total and indirect effect on competence via decreasing all three dimensions of engagement, as well as a remaining significant direct effect. In summary, the negative effect of study intensity on competence was partially mediated by engagement. In addition, study intensity had a significant negative total and indirect effect on GPA via lower emotional engagement but no direct one, so the effect of study intensity on GPA was fully mediated by emotional engagement. Hence, hypothesis 1 was mostly supported.

Decision latitude had a significant positive total and indirect effect on study satisfaction via lower emotional exhaustion and higher emotional engagement, as well as a remaining significant direct effect. Therefore, the positive effect of decision latitude on study satisfaction was partially mediated by emotional exhaustion and emotional engagement. Furthermore, decision latitude had a significant positive total and indirect effect on professional competence via enhancing all three dimensions of engagement, as well as a remaining significant direct effect. The effect of decision latitude on professional competence was partially mediated by engagement. In addition, decision latitude had no significant total effect on GPA, but a significant positive indirect effect via higher emotional engagement. There were neither more indirect effects nor a significant direct one, so the effect of decision latitude on GPA was fully mediated by emotional engagement. Hence, hypothesis 2 was mostly supported.

There was neither a significant total nor an indirect effect of social support from peers and lecturers on the outcomes. Hence, hypotheses 3 and 4 were not supported.

4. Discussion

Based on the assumptions of the JD-R model or SD-R model, respectively, our study aimed to examine how perceived study intensity and different study resources predict study success in distance learning and if the associations are mediated by emotional exhaustion and engagement. We analyzed the bivariate correlations between the variables and their unique incremental, direct, and indirect effects using multivariate path analyses. The hypotheses were mostly supported for study intensity as an indicator of study demands and decision latitude as a study resource. Social support by peers and lecturers, however, was not incrementally predictive of exhaustion, engagement, and academic achievement. The multivariate model explained a substantial amount of variance in both the subjective indicators of academic success (33% and 28% of study satisfaction and professional competence, respectively) and, to a smaller degree, the objective criterion GPA (10.3%).

4.1. Emotional Exhaustion, Engagement, and Academic Success

We start the discussion with the associations between emotional exhaustion and engagement, and academic success. We measured engagement, in contrast to previous studies, differentiated into three dimensions, which helped us gain more nuanced insights. While most of the earlier studies confirmed the correlation between engagement and different academic outcomes (see Bakker & Mostert, 2024), Mokgele and Rothmann (2014) and Gusy et al. (2016) did not find these effects. This might be due to different aspects of engagement that these studies looked at. In our study, all aspects of engagement showed substantial correlations with all three indicators of academic success. In the multivariate model, however, this remained true for emotional engagement, whilst cognitive and behavioral engagement were incrementally predictive of perceived competence only. Since emotional engagement was measured by the experience of flow, it is closely related to intrinsic motivation (e.g., Deci & Ryan, 1985), subjective feelings of a smooth learning process, optimal fit between demands and abilities, and effortlessness (e.g., Rheinberg et al., 2003). Previous studies have shown that this has a positive effect on learning performance (e.g., Engeser et al., 2005), which is supported in our study by its effects on both subjective and objective criteria of academic success. The results regarding cognitive and behavioral engagement show parallels to those of Boerner et al.’s (2005) study in a blended learning environment. Similarly to our findings, they found that cognitive strategies (e.g., control and regulation) and behavioral strategies like effort were not predictive of GPA, but behavioral engagement (corresponding to the subscale “effort” of the LIST questionnaire) was associated with competence. In contrast to their results, in our study, cognitive engagement positively predicted competence, while Boerner et al. (2005) did not find this effect.

Unlike emotional engagement, emotional exhaustion as the second mediator showed substantial bivariate correlations with all three indicators of academic success, but in the multivariate model, however, exhaustion was incrementally predictive just for study satisfaction. Hence, despite many exhausted students still being able to achieve, they nevertheless tend to be unsatisfied with their studies, which is a risk factor for drop-out in general and in distance education, in particular (Choi & Park, 2018).

4.2. The Relevance of Study Demands in Distance Education

Hypothesis 1 predicted that study intensity is negatively associated with academic success, mediated by higher emotional exhaustion and lower engagement. This could be mostly confirmed.

Both in the bivariate and the multivariate analyses, study intensity was significantly associated with all mediators and all objective and subjective indicators of academic success. Study intensity negatively predicted GPA (total effect β = −.14; p = .038), professional competence (total effect β = −.25; p < .001) and study satisfaction (total effect β = −.28; p < .001) mostly mediated by higher emotional exhaustion and lower emotional, cognitive, and behavioral engagement. Presumably, high study intensity is a particular strain in distance education, given students’ extensive extracurricular obligations. The perception of high demands has both emotional consequences (more exhaustion and less emotional engagement) for students, and it negatively affects the effectiveness of their learning process (less cognitive and behavioral engagement). In the long run, the negative emotions, maladaptive study behavior, and low academic success might set off a downward spiral of perceiving the demands as increasingly overburdening, which in turn further decreases performance and well-being, finally resulting in drop-out (e.g., Demerouti et al., 2009). Future studies should address these longitudinal reciprocal effects using repeated assessments over a longer study period.

4.3. The Relevance of Study Resources in Distance Education

Hypothesis 2 predicted that decision latitude is positively associated with academic success, mediated by emotional exhaustion and lower engagement. This could be partly confirmed, too.

Decision latitude had significant associations with all dimensions of engagement, competence, and study satisfaction. In the multivariate model, it was also associated with emotional exhaustion. Decision latitude did not significantly predict GPA (total effect β = −.01; p = .857). The significant positive associations with competence (total effect β = .31; p < .001) and study satisfaction (total effect β = .33; p < .001) were mostly mediated by lower emotional exhaustion and higher emotional, cognitive, and behavioral engagement.

It seems that the autonomy provided by decision latitude leads to greater satisfaction when it elicits more flow and absorption during learning, but not when it is met with increased effort, regulation, or control. The association between decision latitude and experienced competence was partially mediated by all dimensions of engagement. This could indicate a sense of empowerment that results from using autonomy to engage in learning. GPA was only indirectly predicted via emotional engagement. It appears that decision latitude can enhance achievement when it translates into emotional engagement, but in higher amounts, it may no longer be advantageous in all cases. This is consistent with the idea of a Vitamin-Model (Warr, 2009) that suggests that the positive effects of some job characteristics, including decision latitude, might turn into negative effects once a certain level of the characteristic is exceeded.

Hypotheses 3 and 4 predicted that social support from lecturers and peers is positively associated with academic success, mediated by emotional exhaustion and lower engagement. These hypotheses had to be rejected.

Whereas there were some bivariate associations between social support from lecturers and peers with emotional exhaustion or engagement, these were no longer statistically significant in the multivariate model. These findings are in contrast to some earlier research about social support as a resource in face-to-face studies (e.g., Gusy et al., 2016). Cilliers et al. (2018) are in line with our results for support from lecturers: The correlation between support and engagement disappeared after controlling for personality. Boerner et al. (2005) also found no significant correlations between rapport with peers and grade or learning success in the learning field in a blended learning environment. Ghasempour et al. (2023) found evidence for associations between instructor support and academic success in online learning. However, the participants had previously studied face-to-face for two semesters, and they had not studied exclusively in distance. Furthermore, their questionnaire on social support also included, in contrast to ours, the aspect of feedback from instructors. This may have been the decisive variable in this case. Although the students in our study obviously felt well supported (mean of 4.0 for support from lecturers and 4.2 for support from peers and a scale mean of 3.5) it seems that social support has just a very limited incremental influence on engagement, exhaustion and the tested outcomes after controlling for study intensity and decision latitude. The lack of relevance might be due to the operationalization with only three items each and without considering different aspects of social support. Maybe a more differentiated view on the nature of functional support (e.g., Cohen & Wills, 1985), separated into emotional, instrumental, and informational, can shed more light on the effect of social support in distance learning. Furthermore, in this context, maybe it would be more meaningful to investigate the influence of feedback provided by instructors (and peers) instead of social support. Schneider and Preckel (2017), par example, highlighted in their review the crucial role of social interaction for the success of higher education.

In our study, we focused on resources within the direct environment of the university. However, future research should also consider social support from the private and working environment (par example, from friends, colleagues, and family), which might play a crucial role for distance learners due to their numerous other commitments beyond their academic demands.

Overall, the application of the SD-R model proved to be fruitful in the context of distance learning for examining the relationships between subjectively perceived study conditions and various outcomes of academic success. We were able to explain a significant proportion of the variance in academic success, particularly in the subjective criteria of study satisfaction and the experience of professional competence. While high study demands proved to be particularly detrimental to the subjective criteria, decision latitude helps in better managing one’s studies and enhances satisfaction as well as the experience of competence.

4.4. Limitations

The generalizability of the results is restricted due to the sampling of psychology students from one distance university. Since participation in the survey was voluntary, there may have been a selection bias favoring less exhausted and particularly engaged students who did not perceive the demands as remarkably high.

We assessed study intensity as a set of various demands that arise solely from the studies themselves. Beyond that, it can be assumed that among the group of non-traditional students, private, professional, and study-related demands might be more intertwined compared to traditional students. Even though we used an emotional exhaustion questionnaire specifically for the study setting, academic demands and their effects may be perceived as particularly high when burdens from other areas already exist and lead to additional interactions. This should be considered in future surveys to gain a more nuanced insight into how the interplay of demands affects outcomes in distance education.

The significance of GPA may be lower for students in distance learning, as the motivations and goals associated with their studies can vary significantly between different student groups. For example, older students or those who have already completed an apprenticeship and are working part-time or full-time might be less focused on achieving good grades and more interested in learning, growth, and skill development. Here, learning goal orientation might take center stage. Neroni et al. (2018) examined the associations between goal orientation and academic achievement in adult distance education. They did not find a significant association between the mastery approach and academic performance. Future research could focus on interaction effects based on different student characteristics.

This study focuses on the relationship between subjectively perceived study conditions and study success. In the future, however, individual factors and resources should also be considered. For instance, Bakker et al. (2015) found in their diary study that on days when students had greater access to personal resources, they also reported higher levels of study resources, and vice versa. Tho (2023) demonstrated that students with more personal resources exhibited greater initiative in enhancing their social and structural study resources. Also in the professional context, evidence suggests that individuals with greater personal resources also have increased access to environmental resources (e.g., Xanthopoulou et al., 2009)

We used self-reported data for our study. Since our focus was on examining the effects of subjectively perceived study conditions as predominantly occurs in the research about the JD-R model in the work context (e.g., review of Demerouti & Nachreiner, 2019), the data collection method was appropriate. However, self-reports -especially of emotions- are subject to fluctuations and may be influenced by a recall bias (e.g., Althubaiti, 2016). Therefore, it could be beneficial to conduct a diary study with multiple measurement points to analyze relationships more precisely.

The use of cross-sectional data does not allow the causal inferences proposed by the SD-R model. To validate our data, we therefore need longitudinal studies, which at least facilitate the study of correlations between earlier assessments and behavior and later success by using random intercept cross-lagged panel models that enable the decomposition of measurements into between- and within-person components (e.g., Speyer et al., 2023).

4.5. Practical Implications

The detrimental effects of high demands and the positive effects of decision latitude should be highlighted in universities. On the one hand, universities offering distance education can use the assumptions of the model as a diagnostic tool and develop questionnaires to assess perceived autonomy and demands. This can help identify where demands are considered particularly high and decision latitude is especially low, allowing for subsequent adjustments.

The demands can be both essential (e.g., engaging with the learning material, preparing for exams) and unnecessary, such as when organizational efforts are disproportionately high or multiple exams are combined. Lecturers should be trained to minimize unnecessary demands as much as possible. This also applies to administrative processes originating from the university’s administration.

On the other hand, instructors should be trained to increase the autonomy available to students and offer opportunities to expand their skills, both subject-specific skills and those that enable students to effectively use their autonomy for meaningful learning activities and to handle challenging learning material and other demands.

5. Conclusions

In our study, we were able to show that the job demands-resources model—or alternatively, the study demands-resources model—is also highly suitable for analyzing the effects of demands and resources in the context of distance education. We observed negative correlations between study intensity and academic success, which were partially or fully mediated by emotional exhaustion and various dimensions of engagement. In contrast, the resource decision latitude had a positive effect on subjective academic success, though it did not significantly predict the GPA. These effects were partially mediated by emotional exhaustion and engagement. Moreover, we found no significant association between the resource social support from peers or lecturers and emotional exhaustion, engagement, or academic success.

To the best of our knowledge, our study is the first to indicate that it may be beneficial to differentiate the mediator “engagement” into distinct dimensions, as emotional, (meta)cognitive, and behavioral engagement each showed different associations with both the independent and dependent variables. Furthermore, it appears to be useful distinguishing necessary from avoidable demands to minimize the negative effects on academic success. In the future, these results should be replicated through longitudinal research that also takes individual resources into account.