Introduction
Preventing dropouts and promoting not only student retention but also academic success is an essential objective shared by international organisations and governments, researchers and universities (Michavila et al., 2012; Tinto, 2012). In this sense, the first year at the university seems to represent the most critical period for student retention. Although this moment of the transition to a different stage of life may involve positive changes and emotions, it may also increase the levels of uncertainty, anxiety and academic stress (Casuso, 2011; White, Slemp, & Murray, 2017), due to the necessary adaptation for a new and unfamiliar environment that presents different demands and responsibilities and requires putting many personal resources into play (Tinto, 2012). Research on student dropout and failure in higher education (HE) consistently reveal that most cases occur during the first two academic years (Crue Universidades Españolas, 2018; European Commission, 2015). An extensive body of research on this problem has been conducted in the fields of education and psychology.
Increasing attention is being placed on students' wellbeing as the main objective of HE institutions (Harvard, 2016). Positive psychology, oriented toward optimal human functioning, personal development, and welfare, contributes to this issue by focusing on the main role of human strengths and resources (Seligman & Csikszentmihalyi, 2001). According to its postulates, positive education entails incorporating students' well-being as a priority objective in education (Salanova & Llorens, 2016; Seligman & Adler, 2018). Therefore, it will be necessary to investigate which variables and conditions promote and enhance, not only students' performance but also their well-being and which pose a threat. It is also essential to pay particular attention to the interaction of personal strengths and resources with academic demands, and their impact in significant areas, such as academic success and failure (Williams, Horrell, Edmiston, & Brady, 2018), and psychological welfare (Harvard, 2016).
The Demands-Resources (D-R) Model applied to the Educational Field
The demands-resources model (Bakker & Demerouti, 2013; Schaufeli & De Witte, 2017), mostly applied in job settings (Schaufeli & Taris, 2014), could be a useful framework for addressing the research on personal resources that may help students to cope with their academic demands successfully increasing their wellbeing. The model assumes that environmental characteristics can be divided into two categories: resources (capacities of each to face tasks in a specific context) and demands (aspects of a specific context that require significant physical, cognitive and/or emotional effort) (Schaufeli & Taris, 2014). Accumulating evidence shows that an increase in (job and academic) resources correlates with an increase in wellbeing, whereas their loss -or the risk of their lossmay cause their decrease (Halbesleben, 2010; Mäkikangas, 2016; Schaufeli & De Witte, 2017). Resources also cushion the negative effect, per se, of demands; consequently, we could assume that the greater the resources, the higher the likelihood of addressing the demands, to where they may even be considered as challenges to overcome. A proper balance of resources to face demands will represent better levels of work, academic and psychological outcomes.
In academic environments, personal resources (e.g., intrinsic motivation, efficacy perception, etc.) would then relate to how students face demands (e.g., workload) and represent a source of well-being and success themselves (Carmona-Halty, Salanova, Llorens, & Schaufeli, 2018). Even though the number of studies focused on students'resources and personal strengths that lead them to face their academic demands successfully is increasing (e.g., Salanova, Martínez, Bresó, Llorens, & Grau, 2005; Schaufeli & Taris, 2014), according to White et al. (2017) the D-R model has been scarcely applied to university students (e.g., Barr, Sessa, Summer, & Bragger, 2015; Mokgele & Rothman, 2014). Nevertheless, D-R postulates could be well-suited to this field (Schaufeli & Taris, 2014). Likewise, work contexts, performance in academic settings also requires the use of individuals' resources to overcome those demands with which students must cope, such as tests/evaluations, deadlines, relationships with colleagues and supervisors, and so on (Mokgele & Rothman, 2014).
In line with these postulates, it is crucial to determine which personal strengths and resources should be mainly considered in academic settings to enhance students' wellbeing and success. Conversely, more research is needed to establish which characteristics could increase students' perception of demands that boost discomfort and the risk of dropping out.
The present study is centred in two main psychological characteristics that have proven to present strong evidence of their positive vs negative impact on individuals' performance and well-being, both in work and academic settings: engagement and burnout, respectively (Leiter & Maslach, 2017; Taris, Ybema, & Van Beek, 2017). Accordingly to the D-R model, we aimed to identify those personal resources, demands and learning coping processes that lead to developing academic engagement and burnout in first-year university students.
Psychological Resources and their Relationships with Academic Engagement
Engagement has been widely addressed and conceptualised in the literature (e.g., Halbesleben, 2010; Leiter & Maslach, 2017; Martínez & Salanova, 2003). A common and shared conceptualization among positive psychologists defines engagement as a persistent positive psychological state over time composed of vigour, corresponding to high levels of mental energy and persistence in the face of obstacles; dedication or enthusiasm towards the task or set of relevant tasks performed; and absorption, or ability to concentrate deeply, with a feeling that time "flies by" (Salanova, Bresó & Schaufeli, 2005). In education, results repeatedly confirm the positive impact of academic engagement on student retention (Garbanzo, 2007), wellbeing (Carmona et al., 2018) and academic performance (Michavila et al., 2012), identified as the most important characteristics for preventing student failure and dropout (Tinto, 2012).
Academic engagement has been previously related to different motivational, cognitive and emotional psychological resources. Considered, primarily, as a positive motivational construct (Schaufeli, Salanova, González- Romá, & Bakker, 2002), engaged students are highly intrinsically motivated (Ryan & Deci, 2000). That would explain their high persistence in the face of difficulties (Salanova, Martínez et al., 2005; Tinto, 2012), their greater academic satisfaction and, even, their better academic performance (Garbanzo, 2007).
Engagement has also been linked to other personal resources that have proven relevant in the domains of organisations and academia, such as personal initiative (Frese & Fray, 2001), a behavioural pattern characterised by the self-initiated, proactive and persistent performance that is objectives-driven and action-oriented. Significant positive relationships with students' performance and psychological well-being have also been found in students with high levels of personal initiative (Lisbona, Palací, Salanova, & Frese, 2018).
Another important resource related to students' engagement levels is the perception of efficacy (Tinto, 2012), defined as the degree of preparation that the person self-evaluates as possessing in a specific field (Bandura, 2001). Thus, students who perceive themselves to be more efficacious attribute their failures to a lack of effort, rather than to insufficient capacity, are more likely to initiate study behaviours due to a higher expectation of success, and are prone to set more self-improvement goals requiring more persistence and determination, raising their academic motivation and subsequently, their engagement (Pérez et al., 2018; Salanova, Martínez et al., 2005). Consequently, the self-perception of efficacy could be a positive and necessary precursor for engagement. From a longitudinal perspective, Salanova, Bresó, and Schaufeli (2005) propose a positive spiral through which the perception of efficacy and engagement wield a positive, reciprocal and continuous influence.
Engagement and resilience have also been related (Martínez, Peñalver, & Meneghel, 2016). Given the inherent difficulties and challenges that characterise university life, resilience, defined as the ability to recover and cope with adversity with a more positive perception of problems and difficulties (You, 2016), can be valued as a significant personal strength. Furthermore, resilient students present higher levels of intrinsic motivation (Martínez et al., 2016), well-being (Grant & Kinman, 2012), academic performance (Martin & Marsh, 2006) and retention (European Commission, 2015).
Also, engaged students seem to make use of more proactive coping strategies (e.g., Grant & Kinman, 2012; Sánchez-Elvira-Paniagua, Fernández, & Amor, 2006). In this sense, good management of demands and challenges has been related to these active coping strategies, such as persistence, by mediating those demands that impact on variables such as academic performance (O'Connor, Nguyen, & Anglim, 2016; Salmela-Aro & Read, 2017). To the contrary, academic engagement has a negative association with academic procrastination (González-Brignardello & Sánchez-Elvira-Paniagua, 2013). Academic procrastination is defined as the tendency to intentionally delay the start, execution and completion of academic tasks and duties “despite expecting to be worse off for the delay” (Steel & Klingsieck, 2016, p. 37).
Demands in Academic Settings and their Relationship with Academic Burnout
Academic burnout is defined as a persistent negative mental state regarding studies, mainly characterised by emotional exhaustion accompanied by discomfort, cynicism or detachment from studies and lack of academic self-efficacy in students without psychological disorders (Llorens, García-Renedo, & Salanova, 2005; Maslach & Leiter, 1997).
The experience of academic burnout has been consistently linked to the greater perception of academic demands (Mokgele & Rothman, 2014) and academic stress (Casuso, 2011), and usually interpreted as a consequence (García-Ros, Pérez-González, Pérez-Blasco, & Natividad, 2012). Situations such as university transition, academic overload, lack of time, social relations and performance on academic tests may be considered as academic demands, which are likely to generate stress if students perceive that they lack proper resources and/or strategies to face them (Casuso, 2011; García-Ros et al., 2012; White et al., 2017). According to the transactional model of Lazarus and Folkman (1987) and the D-R model (Bakker & Demerouti, 2013), people will perceive stress and, eventually, burnout, when they consider that the demands exceed their resources or capacities. Burnout not only entails distress but also impaired academic performance (Schaufeli, Martínez, Marqués-Pinto, Salanova, & Bakker, 2002) and a greater tendency toward dropping out, as well (Martínez & Marques, 2005).
Furthermore, academic burnout has been associated with fewer resources (Schaufeli & Taris, 2014) and harmful learning strategies such as procrastination (González- Brignardello & Sánchez-Elvira-Paniagua, 2013), which has been inversely related to self-regulatory processes, self-efficacy, subjective well-being and academic performance (Kim & Seo, 2015; Klaassen, Krawchuk, & Rajani, 2008; Steel, 2007; Steel & Klingsieck, 2016).
Remembering what has been exposed before, we could postulate that academic engagement would be facilitated by a set of positive motivational, emotional and cognitive, personal resources that contribute toward perceiving fewer demands in the environment and a greater capacity for facing them properly. The outcome would be a positive impact on students' well-being and performance that would consolidate the resources and, eventually, result in greater engagement. On the contrary, academic burnout could be induced by a higher perception of demands and insufficient resources for coping with them, and poorer coping strategies for dealing with the demands properly, deleteriously affecting students'performance and wellbeing.
The previous relationships discovered between engagement and burnout with relevant personal resources and perceived demands guided the selection of the variables investigated in the present study.
Academic Engagement and Burnout: Opposite poles?
Engagement and burnout have been postulated as two states that show different and, in many cases, opposite relationships with different psychological variables inherent to academic performance. As seen before, previous evidence highlights that engagement relates more to personal resources and burnout to the perception of existent demands and stress (Taris et al., 2017).
However, some discrepancies currently exist regarding the conceptualisation and measurement of engagement and burnout (e.g., Schaufeli & De Witte, 2017; Taris et al., 2017). Although some researchers argue that they represent opposite extremes of the same construct (e.g., Maricotui, Sulea, & Lancu, 2017), others point out that research is not unequivocally conclusive in this regard (e.g., Schaufeli & De Witte, 2017).
Whereas there is evidence of a clear antagonism of the dimensions that comprise the so-called "core" of engagement and burnout: vigour vs exhaustion and dedication vs cynicism (Schaufeli & Salanova, 2007), nevertheless, the perception of lack of efficacy as a component of burnout poses another controversy among researchers. It has been postulated, first, that the measurement of burnout using a subscale with reversed-positive items of efficacy may not reliably assess this characteristic (Schaufeli & Salanova, 2007); and, second, that it could be a necessary precursor, more so rather than a real component, of burnout (Salanova et al., 2004). Also, the perception of efficacy could precede engagement levels, acting as a resource and a possible mechanism for self-motivation, in those who perceive themselves as capable of facing a task often put more of their resources into action, are more persistent and set further-reaching goals (Salanova et al., 2004). In that case, it would be interesting to explore the contribution of the perception of efficacy in developing engagement and burnout alike.
Objectives
Based on the previous evidence and within the framework of the demands-resources model, the main objective of the present study was to explore, under a multivariate approach, the contribution of a set of relevant personal resources, academic demands, learning and coping strategies, affect and satisfaction in the prediction of students' engagement and burnout in their first university year, before their first exams. Another objective was to look for the likely contribution of efficacy as an independent psychological resource.
Personal resources were expected to predict engagement, whereas burnout, and core burnout specifically, was expected to be mostly predicted by the perception of more academic demands.
Method
Sample
One hundred and seventy-three incoming first-year students from 20 Spanish face-to-face (traditional learning) universities (68.8 % female), mostly enrolled in Social and Legal Sciences, volunteered to participate in this study. The mean age was 22.99 years old (ranging from 18 to 58 years, SD = 7.62; see Table 1). This average age was high due to some extreme values; nevertheless, 71.7 % of our sample was between 17 and 21 years old.
Instruments
A Spanish online battery of tests was designed to be voluntarily completed. The battery included socio-demographic, academic and psychological variables, as described below. Internal reliability found in previous studies was adequate (García-Ros et al., 2012; Notario- Pacheco et al., 2011; Salanova, Bresó, & Schaufeli, 2005; Salanova, Schaufeli, Llorens, Peiró, & Grau, 2000; Sandín et al., 1999).
Socio-demographic variables. Ad-hoc items were used to collect information about students' age, gender, university, the branch of knowledge, current academic year and total number of ECTS (European Credit Transfer System) in which they were enrolled.
Academic engagement. Engagement was measured with the student version of the Utrecht Work Engagement Scale, UWES (Salanova et al. 2000). It is comprised of 15 items and a Likert-type scale ranging from 0 (never) to 6 (always). It consists of three dimensions: vigour (e.g. “when I study, I feel like I am bursting with energy”), dedication (e.g., “I am proud of my studies”) and absorption (e.g., “I am immersed in my studies”). The global reliability coefficient in our study was .92 and .83, .84, .84 for absorption, dedication & vigour subscales, respectively.
Academic burnout. Burnout was measured with the Maslach Burnout Inventory-Student Survey, MBI-SS (Schaufeli, Salanova et al., 2002) in its Spanish adaptation (Salanova, Martínez et al., 2005). It is comprised of 15 items grouped into three dimensions: exhaustion (e.g., “I feel used up at the end of a day at university”), cynicism (e.g., “I doubt the significance of my studies”) and efficacy (e.g., “During class I feel confident that I am effective in getting things done”) (reverse positive items). The MBI-SS uses a Likert-type scale ranging from 0 (never) to 6 (always). The global reliability coefficient was .86 and 84, .86, .84 for exhaustion, cynicism and efficacy subscales, respectively. Core burnout, comprising the dimensions of exhaustion and cynicism, showed a .83 reliability coefficient.
Psychological resources
Reasons for degree selection. Students were asked about the reasons why they chose their current studies, through 14 items, and a Likert-type scale ranging from 1 (not important at all) to 4 (very important). Items were grouped into two subscales: intrinsic (“for personal satisfaction”) and extrinsic ("it is a way to increase the chances of getting a promotion") motives. The reliability coefficients were .86 for intrinsic motives and .71 for extrinsic ones.
Academic motivation. Intrinsic and extrinsic motivation subscales (Batería de Escalas de Aprendizaje Autorregulado, Sánchez-Elvira-Paniagua et al., 2006). It consists of a 13-item Likert-type scale ranging from 1 (not at all) to 4 (almost always). "I put a lot of effort into my studies" or "the fear of not fulfilling my studies is what motivates me the most" represent examples of two items. The reliability coefficients were .73 for identified intrinsic motivation, .55 for pure intrinsic motivation, and .74 for extrinsic motivation.
Personal initiative. Measured with the Self-reported Initiative Scale (Frese, Fay, Hilburger, Leng, & Tag, 1997) in its Spanish adaptation (Las-Hayas, Lisbona, & Palací, 2018). It is comprised of seven items (e.g., "I solve problems without waiting to see what's next") and a Likert-type scale ranging from 1 (strongly disagree) to 4 (strongly agree). The scale reliability coefficient was .81.
Resilience. Connor-Davidson Resilience Scale (Connor & Davidson, 2003), in its Spanish adaptation (Notario-Pacheco et al., 2011). It is comprised of 10 items and a Likert-type scale ranging from 0 (not at all) to 4 (almost always). One example of an item is "I believe that I am a strong person when I face the challenges and vital difficulties". The scale reliability coefficient was .89.
Learning and Coping Strategies
Academic procrastination. The Academic Procrastination questionnaire (González-Brignardello & Sánchez- Elvira-Paniagua, 2013) is comprised of 18 items (e.g., “I work in a non-systematic way (disorderly)” and a Likerttype scale ranging from 1 (absolutely not) to 5 (totally). The scale reliability coefficient was .90.
Coping strategies. Measured with the Batería de Escalas de Aprendizaje Autorregulado (Sánchez-Elvira-Paniagua et al., 2006). Due to the objectives of this study, three of their subscales were selected, comprised of 19 items: persistence (e.g., “I strive for studying subjects, whether I like them or not”), avoidance strategies in the face of difficulties (e.g., “I usually throw in the towel facing significant difficulties when studying”) and academic anxiety (e.g., “I feel nervous while I'm studying”). The reliability coefficients for each subscale were .83, .88 and .60, respectively.
Academic Demands
They were measured with the Cuestionario de Estrés Académico en la Universidad (University Academic Stress Questionnaire; García-Ros et al., 2012). According to the objectives of this study, only two subscales were selected, comprised of 13 items and a Likert-type scale ranging from 1 (no stress) to 6 (much stress): academic obligations (e.g., “doing exams”) and academic records and future perspectives (e.g., “to obtain high grades in the subjects”). The reliability coefficients were .81 for academic obligations and .79 for academic records and future perspectives.
Satisfaction and Affect
Affect. The Positive and Negative Affect Schedule (Watson & Clark, 1988) in its Spanish adaptation (Sandín et al., 1999). It is comprised of 20 items grouped into two subscales, positive and negative affect. Students were asked about how intensely they experienced last week each of 20 feelings (e.g., enthusiastic, ashamed) with a Likert-type scale ranging from 1 (very slightly) to 4 (extremely). The reliability coefficients were .88 for positive affect and .91 for negative affect. Likewise, the affect balance measure (positive affect - negative affect) was calculated.
Degree of satisfaction with the studies. It was measured with a single ad-hoc item (“indicate the satisfaction level with your studies”) and a Likert-type scale ranging from 1 (none) to 5 (very much).
Procedure
The faculty of the participating universities distributed the link to the battery of online questionnaires. The test was administered prior to the first semester evaluation period, between November and December. The Qualtrics© software used allowed for responding to the instruments using a computer, tablet or phone, and included explicit, informed consent.
Statistical Analyses
Pearson correlation coefficients were conducted to analyse the relationships of academic engagement and burnout with the rest of the variables of the study. Subsequently, to determine what variables predicted the levels of engagement and burnout, two multiple regression analyses, method forward, were carried out. Those variables with a correlation with the criterion ones equal to or higher than .35 were introduced in the model as predictor variables. All statistical analyses were performed with the SPSS 24 software.
Results
Correlational Analyses
Correlational analyses between engagement, burnout and core burnout scales and its subscales were carried out.
Engagement and burnout were significantly and negatively correlated (-.53), though that correlation did not allow us to conclude that the constructs were opposite. Also, the correlation between engagement and core burnout were substantially smaller (-.33). That was because efficacy correlated higher with engagement (.63) than with core burnout (-.36). The efficacy subscale (measured in its original, positively-worded items) presented significant correlations with both global measures and its subscales, except exhaustion (-.13); additionally, positive and significant correlations with vigour, dedication and absorption were found, and negative correlations with cynicism. The strongest correlation was with vigour (.65) (Table 2). Given that efficacy correlated with engagement subscales more than with the burnout ones (except for absorption) and given the theoretical background previously explained, the efficacy subscale (here considered as the belief in ones'abilities and skills to cope and deal satisfactorily with study demands) was hereafter considered an independent personal resource and a predictor variable instead of a burnout subscale, for this study.
M (SD) | (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | (9) | |
---|---|---|---|---|---|---|---|---|---|---|
Engagement (1) | 4.69 (1.00) | — | -.53** | -.33** | .88** | .82** | .91** | -.10 | -.55** | .63** |
Burnout (2) | 3.10 (.90) | — | .91** | -.35** | -.49** | -.55** | .70** | .77** | -.71** | |
Core burnout (3) | 3.03(1.05) | — | -.20** | -.32** | -.34** | .86** | .72** | -.36** | ||
Absorption (4) | 4.31 (1.18) | — | .56** | .72** | -.03 | -.40** | .45** | |||
Dedication (5) | 5.58 (1.04) | — | .63** | -.04 | -.62** | .54** | ||||
Vigour (6) | 4.19 (1.23) | — | -.17* | -.45** | .65** | |||||
Exhaustion (7) | 3.80 (1.31) | — | .28** | -.13 | ||||||
Cynicism (8) | 2.16 (1.18) | — | -.50** | |||||||
Efficacy a (9) | 3.22 (1.12) | — |
Note. **p < .01 (bilateral),
*p < .05 (bilateral).
Table 3 shows the correlations obtained between the global scales of engagement, burnout and core burnout and the set of variables studied. Engagement showed moderate to high positive and significant correlations with all the personal resources, except extrinsic motivations, with the correlations with pure intrinsic motivation (.69) and efficacy (.63) being the highest. These correlations were also significant, but inverse, with burnout, although their value was significantly reduced when core burnout was considered, with low effect sizes.
M (SD) | Engagement | Burnout | Core burnout | |
---|---|---|---|---|
Reasons for degree selection | ||||
Intrinsic motivations | 3.28 (.63) | .49** | -.37** | -.26** |
Extrinsic motivations | 2.26 (.65) | .12 | .02 | .08 |
Personal resources | ||||
Personal initiative | 3.37 (.71) | .57** | -.41** | -.16* |
Pure intrinsic motivation | 2.7 (.64) | .69** | -.48** | -.30** |
Identified intrinsic motivation | 3.71 (.37) | .59** | -.42** | -.28** |
Extrinsic motivation | 2.79 (.72) | .24** | .09 | .22** |
Resilience | 3.67 (.82) | .45** | -.37** | -.25** |
Efficacy |
3.22 (1.12) | .63** | -.36** | |
Learning and coping Strategies | ||||
Academic procrastination | 3.05 (.85) | -.41** |
.52** |
.40** |
Persistence | 3.12 (.54) | .64** | -.38** | -.23** |
Anxiety | 1.50 (.49) | -.08 | .49** | .52** |
Avoidance | 2.45 (.76) | -.39** | .45** | .34** |
Academic demands | ||||
Academic obligations | 3.52 (.78) | -.02 | .50** | .54** |
Future perspectives | 3.02 (.96) | -.02 | .39** | .51** |
Enrolled ECTS | 56.06 (17.25) | -.04 | .03 | .11 |
Satisfaction and Affect | ||||
Degree Satisfaction | 3.95 (.82) | .54** | -.50** | -.42** |
Positive affect | 2.54 (.62) | .53** | -.49** | -.38** |
Negative affect | 2.09 (.76) | -.06 | .45** | .51** |
Affect balance | .45 (1.14) | .32** | -.57** | -.54** |
Note. **p < .01 (bilateral),
*p < .05 (bilateral).
Regarding the ways students coped with their studies, our results revealed that the higher the engagement, the greater the persistence and lower the procrastination, avoidance and anxiety. On the contrary, the higher the burnout, the higher levels of procrastination, avoidance and anxiety when facing academic tasks and the less persistence. These correlations decreased to some extent when core burnout was considered, except for academic anxiety, which experienced a slight increase.
Engagement did not show significant correlations with the perception of academic demands, unlike burnout, which showed positive and significant correlations, and core burnout, which showed even higher correlations.
Finally, engagement was positively correlated with positive affect but not with negative affect, in a different way to burnout, which correlated significantly both with negative affect and inversely with positive affect, and core burnout, which presented an even increased correlation with negative affect. Due to this pattern, burnout and core burnout alike presented significant, inverse correlations with a positive balance of affect.
Given the correlations obtained, we can conclude that in general engagement and burnout demonstrated a pattern of inverse relations with personal resources and coping strategies, academic demands, affect and satisfaction levels. Besides, core burnout presented stronger relationships with academic demands and negative affect than with burnout. Finally, it should be noted that engagement and both burnout measures did not correlate with the number of enrolled ECTS.
Predictive Analyses of Engagement, Burnout, and Core Burnout
Three multiple regression analyses, method forward, were carried out using engagement, burnout and core burnout levels as criterion variables (considering the positive-worded efficacy scale as an independent variable). Those variables with a correlation with the criterion ones equal to or higher than .35 were introduced as predictor variables in each model. All variable scores were standardised (Table 4).
Criteria | Predictors | Adjusted R2 |
F | df | ß | T | |
---|---|---|---|---|---|---|---|
Engagement | .68 | 52.15*** | 7 | ||||
Pure intrinsic motivation | .22 | 3.35*** | |||||
Degree satisfaction | .17 | 3.41*** | |||||
Efficacy | .16 | 2.73** | |||||
Persistence | .19 | 3.29*** | |||||
Intrinsic motivations for degree selection | .12 | 2.46* | |||||
Identified intrinsic motivation | .16 | 2.92** | |||||
Positive affect | .12 | 2.3* | |||||
Burnout | .63 | 49.82*** | 6 | ||||
Engagement | -.22 | -3,27** | |||||
Academic obligations | .35 | 6,08*** | |||||
Academic procrastination | .23 | 4.46*** | |||||
Identified intrinsic motivation | -.21 | -3.59*** | |||||
Negative affect | .16 | 2,75** | |||||
Degree satisfaction | -.15 | -2,58* | |||||
Core burnout | .48 | 32.55*** | 5 | ||||
Academic obligations | .26 | 3.50*** | |||||
Degree satisfaction | -.22 | -3.70** | |||||
Academic procrastination | .21 | 3.58*** | |||||
Future perspectives | .18 | 2.56* | |||||
Negative affect | .17 | 2.35* |
Note. *p < .05,
**p < .01,
***p < .001.
Seven predictors explained 68 % of the variance of engagement. These predictors were, fundamentally, intrinsic motivational factors, concerning both the reasons for choosing the degree and the levels of academic motivation. A perception of efficacy, the use of persistence as an active coping strategy, positive affect and satisfaction with studies also contributed to engagement prediction. Therefore, a pattern associated with personal resources and active coping, with beneficial consequences for students, was found in the prediction of engagement levels.
Six predictors explained 63 % of the variance of burnout. On the one hand, the experience of burnout was positively predicted by the perception of academic obligations, academic procrastination and negative affect; and on the other hand, negatively by engagement, identified intrinsic motivation and degree satisfaction. Accordingly, a pattern associated with academic stress and poor learning strategies, and also disengagement, dissatisfaction and demotivation, predicted burnout levels significantly. When core burnout was considered, the percentage of variance explained decreased to 48%. Main predictors were also the experience of academic demands, negative affect, procrastination and dissatisfaction with studies. Therefore, core burnout was predicted by the perception of demands but not by the lack of personal resources, according to previous studies and hypotheses.
Discussion
The present study aimed to explore, in depth, the role of some relevant psychological characteristics in predicting the levels of engagement and burnout experienced by first-year students of face-to-face teaching universities before their first semester exams, using a multivariate demands-resources approach. For this purpose, relevant variables previously related to engagement and burnout, and supported by empirical evidence of their impact on well-being and academic performance, were considered (psychological resources, academic coping strategies, academic demands, affective states and academic satisfaction).
Our data demonstrated by using a multivariate approach that the engagement levels of first-year university students were strongly predicted by a set of relevant personal resources, mainly driven by intrinsic motivations and efficacy. The predictive value of having chosen a degree for intrinsic reasons is worth highlighting. In previous studies, this motivation was related to academic performance and dropout prevention (e.g., Vizoso & Arias- Gundín, 2016). Also, persistence (as a proactive way of coping with studies), positive affect and academic satisfaction were significant predictors of the levels of engagement experienced by students before their first exams.
To the contrary, academic burnout was mainly predicted by the perception of academic demands and procrastination, and inversely by engagement (a motivational construct per se; Schaufeli, Martínez et al., 2002), and intrinsic motivation; finally, dissatisfaction with studies and negative emotions were significant predictors, as well, of the burnout experienced by students. Low levels of engagement (explained by a set of relevant personal resources as previously mentioned), would highlight the significant link between academic burnout and lack of resources, in line with other studies (e.g., Schaufeli & Taris, 2014).
These analyses are also congruent with the demandsresources model, in which resources are related to greater well-being and, conversely, demands show a direct link to dissatisfaction (e.g., Bakker & Demerouti, 2013; Mäkikangas, 2016).
Additionally, academic core burnout revealed similarities and differences with global burnout, although burnout was predicted by a greater percentage of the variance. Both constructs were mainly predicted by the perception of academic demands and procrastination, and also by dissatisfaction and negative affect. Nevertheless, engagement and identified intrinsic motivation did not predict core burnout levels when efficacy was considered an independent personal predictor, congruently with its strong relationship with these personal variables. These results are similar to those reported in previous studies (Bedewy & Gabriel, 2015; García-Ros et al., 2012; Schaufeli & Salanova, 2007; Salanova et al., 2004). Nonetheless, only 48 % of core burnout variance could be explained in our study. Therefore further studies would have to gain a better understanding of this construct and its predictors.
In summary, our multiple regression analyses confirmed the hypotheses that different variables predict engagement and burnout, mainly the perception of personal resources (e.g., Barr et al., 2015) or academic demands (e.g., Mokgele & Rothman, 2014), respectively.
Likewise, the study contributes to the current debate (Taris et al., 2017) as to whether engagement and burnout are two poles of the same construct. Moreover, to whether efficacy beliefs, as measured by the burnout scale, should be better considered a triggering factor in developing engagement and burnout, instead of a real component of the latter (e.g., Maricotui et al., 2017; Schaufeli & De Witte, 2017).
First, our results indicate, in coherence with previous studies (e.g., Goering Shimazu, Zhou, Wadac, & Sakai, 2017; Leiter & Maslach, 2017), that although engagement and burnout were inversely related, they were not strictly antagonistic. Likewise, low scores in one dimension do not necessarily imply high scores in the other (Schaufeli & De Witte, 2017).
Second, our results revealed that although the positively worded efficacy scale was significantly and inversely related to the so-called core burnout, it was even more strongly and positively related to engagement (global scale and subscales). These results are congruent with Schaufeli & Salanova (2007), who caution when assuming that a reversed efficacy scale is a reliable assessment tool of the perception of inefficiency that characterises burnout. As these authors claim, efficacy and inefficacy beliefs should not be considered opposite constructs, and suggest that the former could be an element of engagement and the latter of burnout; in this case, different assessment instruments would be required. In parallel, according to the discussion raised by several authors as to whether efficacy actually represents a component of burnout or is instead a necessary precursor (e.g., Taris et al., 2017), several authors have postulated, on the one hand, that a previous crisis of efficacy could lead to burnout instead of being a real component of the same (e.g., Salanova, Bresó, & Schaufeli, 2005). On the other hand, another proposal is an "extended" engagement comprising the original dimensions and the efficacy subscale -stated positively- (Schaufeli & Bakker, 2001; Schaufeli & Salanova, 2007). In our study, efficacy beliefs, considered an independent resource, predicted engagement but not core burnout. Our results deserve further attention concerning the role and measurement of efficacy in the experience of engagement and burnout.
A clear limitation of our design is its transversal nature since it does not allow for establishing any causal relationships between the analysed variables. Nevertheless, our data encourage exploring the hypothesis that the factors involved in student's welfare and academic success will coexist in a continuous transaction that researchers such as Bresó, Schaufeli, and Salanova (2011) have identified as positive and negative spirals. Thus, personal resources, such as intrinsic motivation, personal initiative or efficacy, would lead to implementing better coping strategies and the experience of positive emotions (Bandura 1997; Fredrickson, 2004), which, the latter, would reinforce the individual psychological resources (Pérez et al., 2018). On the contrary, low levels of personal resources would contribute to procrastination and maladaptive coping with academic demands (Salmela-Aro & Read, 2017; O'Connor et al., 2016), followed by the experience of higher levels of stress (Bedewy & Gabriel, 2015) and negative emotions (Aguayo-Muela & Aguilar-Luzón, 2017), that would negatively affect personal resources.
Another limitation is that the age of our student sample was slightly higher than the average for first-year students, due to a few extreme values, although in our distribution 71.7 % of the sample was between 17 and 21 years old. Additionally, a larger number of sample subjects would allow us a more precise analysis of the intervening variables.
The exclusive use of self-report measures is another limitation of the study, given the potential influence of the acquiescence bias on results. Besides, because participation in our study was voluntary, more likely that students were more motivated were eager to participate, whereas those who were less motivated were not. Nevertheless, our sample yielded a wide range of scores for the analysed variables.
Finally, given the relevance of the variables involved in engagement and burnout, and their role in the wellbeing and satisfaction of first-year students, some future studies and actions we propose include:
To expand in longitudinal studies and to test explanatory analysis models would allow for analysing the spirals models or virtuous and vicious circles to a greater extent, from a broader perspective.
To develop institutional actions to enhance students' engagement by promoting intrinsic motivation and efficacy beliefs, guiding them in their development of positive learning strategies, such as persistence, while simultaneously reducing the risk of students' burnout by orientating them to neither procrastinate nor avoid difficulties, and coping better with stress. The ultimate aim would be to increase students' psychological well-being and academic performance, reduce dropouts and improve retention (Seligman & Adler, 2018), especially during their first academic year.
To focus primarily on the first academic year at the university as a critical period (Michavila et al., 2012; Tinto, 2006, 2012) during which initial expectations are confronted (Merhi, 2011), according to data on student dropout rates (e.g., Crue Universidades Españolas, 2018; European Commission, 2015).
To provide students proper orientation in an early stage (e.g., in secondary education) that encourages them to choose a degree from among those that motivate them more intrinsically (Garbanzo, 2007; Vizoso & Arias- Gundín, 2016).