Boosting Students’ Academic Performance: Trait-State Model of Academic Achievements in Higher Education
Yüklə 207,59 Kb.
|
- Bu səhifədəki naviqasiya:
- Data analysis
- Results Data description
- Test of the model
- Sensitivity analyses
- Test of mediation
- Key findings
- Potential applications
- Limitations and directions for future research
| Participants and procedures
An opportunity sample of 373 students from a London university, 93 (24.9%) males and 280 (75.1%) females with age range 18 to 54 years (M = 25.6, SD = 7.6), participated in the study; 203 (54.4%) students were from the Faculty of Science and Computing, and 170 (45.6%) students were from the Business School, the Faculty of Social Science and Humanities, and the Faculty of Law and International Relations; 328 (87.9%) were undergraduate students, and 45 (12.1%) were postgraduate students; 190 (50.9%) students were White, 68 (18.2%) students were Black, and the remaining 115 (30.9%) students were from other or mixed ethnic backgrounds. The background of this sample was similar to that of previous studies on students from the same university (e.g., Moneta & Spada, 2009: Rogaten & Moneta, 2013, 2015). A university ethics board approved the study. Participants were approached individually in common areas of the University. The study was described to participants as an investigation of the role of psychological factors in predicting study habits and academic performance. Following the granting of informed consent, participants were sent an email with the hyperlink to an online survey. Participants were debriefed online upon completion of the survey.
Consistent with the university’s assessment scheme, all students in the sample had their examinations at the end of a semester, and all coursework submissions took place in the second half of a semester. Students’ grades (expressed in percentage points, with 40 % representing the minimum passing grade) were retrieved from the university registry for the current and previous semester. The individual examination grades and individual coursework grades were separately identified for each participant and a student’s semester average was calculated separately for each of the two types of grades across all the courses taken in that semester 2. Process variables Approaches and Study Skills Inventory for Students (ASSIST Short Version; Entwistle, 2008). The ASSIST Short version consists of 18 items measuring individual tendencies to adopt deep (e.g., “When I am reading an article or book, I try to find out for myself exactly what the author means”), strategic (e.g., “I organise my study time carefully to make the best use of it”), and surface (e.g., “I concentrate on learning just those bits of information I have to know to pass”) approaches to studying, each using six dedicated items. The responses were recorded on a 4-point scale ranging from 1 (Disagree) to 4 (Agree). The scale has good internal consistency ranging from .80 to .87 and concurrent validity through positive correlations of deep and strategic approaches to studying, and a negative correlation of surface approach to studying, with self-reported academic performance (Tait, Entwistle, & MCcune, 1998). Positive and Negative Affect Schedule – Short Form (1-PANAS-SF; Thompson, 2007) consists of 10 adjectives measuring positive affect (e.g., “attentive”) and negative affect (e.g., “nervous”). The responses were recorded on a 5-point scale ranging from 1 (None) to 5 (Very much). The scale showed good internal consistency .80 (positive affect) and .74 (negative affect), good eight week test-retest reliability of .84 for both subscales, good concurrent validity through positive correlations of positive affect, and negative correlations of negative affect, with measures of happiness and subjective well-being (Thompson, 2007). Use of Creative Cognition Scale (UCCS; Rogaten & Moneta, 2015a). The UCCS consists of five items measuring frequency of use of creative cognition in study-related activities (e.g., “I find effective solutions by combining multiple ideas”). The responses were recorded on a 5-point scale ranging from 1 (Never) to 5 (Always). The scale has good internal consistency of .82 and concurrent validity through positive correlations with adaptive metacognitive traits, trait intrinsic motivation and positive affect, and good discriminant validity through lack of correlation with key maladaptive metacognitive traits (Rogaten & Moneta, 2015a). Evaluation Anxiety Scale (EVAN; Thompson & Dinnel, 2001). The EVAN consists of 15 items measuring anxiety in evaluative situations (e.g., “I get anxious when I am given a homework assignment that challenges my ability to do well”). The responses were recorded on a 7-point scale ranging from 1 (Not at all true of me) to 7 (Very true of me). The scale has good internal consistency of .85 and good concurrent validity through positive correlations with fear of negative evaluation, test anxiety and fear of failure (Thompson & Dinnel, 2001). Structure variables Work Preference Inventory (WPI; Amabile et al., 1994). The WPI consists of 30 items, trait intrinsic (e.g., “I enjoy tackling problems that are completely new to me”) and extrinsic motivation (e.g., “I am concerned about how other people are going to react to my ideas”), each using 15 dedicated items. The responses were recorded on a 4-point scale ranging from 1 (Never or almost never true of you) to 4 (Always or almost always true of you). The scores for trait intrinsic and extrinsic motivation were calculated by averaging the scores of their constituent items. The scale has satisfactory internal consistency of .70 for extrinsic motivation and .75 for intrinsic motivation, and has good concurrent validity through positive correlations with measures of personal development, autonomy, ability utilization and achievement (Loo, 2001). Positive Metacognitions and Meta-Emotions Questionnaire (PMCEQ; Beer & Moneta, 2010). The PMCEQ consists of 18 items measuring three adaptive metacognitive traits, each using six dedicated items. For the purpose of this study only factors 2 and 3 were used in the analysis: (2) confidence in interpreting own emotions as cues, restraining from immediate reaction, and mind setting for problem solving (e.g., “I tend to rationally evaluate unpredictable situations rather than getting anxious”), and (3) confidence in setting flexible and feasible hierarchies of goals (e.g., “I find it fairly easy to identify important needs and goals for me”). The responses were recorded on a 4-point scale ranging from 1 (Do not agree) to 4 (Agree very much). The subscale scores were calculated by averaging the scores of their constituent items. The subscales have good internal consistency in the .80 to .88 range, and good convergent validity through positive correlations of PMCEQ-2 and PMCEQ-3 with trait intrinsic motivation (Beer & Moneta, 2010). Meta-Cognitions Questionnaire 30 (MCQ-30, Wells & Cartwright-Hatton, 2004). The MCQ-30 consists of 30 items measuring five maladaptive metacognitive traits, each using six dedicated items. For the purpose of this study only the first four factors were used in the analysis: (1) positive beliefs about worry (e.g., “I need to worry in order to remain organised”), (2) negative beliefs about thoughts concerning uncontrollability and danger (e.g., “I cannot ignore my worrying thoughts”), (3) cognitive confidence (lack of, e.g., “I do not trust my memory”), and (4) beliefs about the need to control thoughts (e.g., “I should be in control of my thoughts all the time”). The responses were recorded on a 4-point scale ranging from 1 (Do not agree) to 4 (Agree very much). The subscale scores were calculated by averaging the scores for their constituent items. The subscale scores have good internal consistency in the range .72 to .93, and good convergent validity through positive correlations with obsessive-compulsive symptoms, worry, and trait anxiety (Wells & Cartwright-Hatton, 2004). Data analysis The model was tested using structural equation modeling in LISREL 8.8 (Jöreskog & Sörbom, 1996). All variables in the model were defined as latent variables with congeneric indicators in order to control for measurement error. Prior semester academic performance and current semester academic performance were defined as two latent variables using the corresponding semester average examination grade and semester average coursework grade as indicators. Positive affect, negative affect and creative cognition were defined as latent variables using their constituent items as indicators. Adaptive metacognition was defined as a latent variable using the PMCEQ-2 and PMCEQ-3 subscales as indicators. Maladaptive metacognition was defined as a latent variable using the MCQ-30-1 through MCQ-30-4 subscales as indicators. Indicators for all other variables in the model (trait intrinsic motivation, trait extrinsic motivation, evaluation anxiety, deep, strategic and surface approaches to studying) were created using parceling, which were formed using the “item-to-construct” method (Little, Cunningham, Shahar, & Widaman, 2002). The model includes 12 latent variables as predictors or mediators of academic performance. Although the sample size is generally sufficient to estimate the hypothesized model as a whole on the data of the present study, the test of indirect effects has low statistical power and is likely to be strongly biased by violations of the Normality assumption. The problem of low statistical power was addressed by a prudent interpretation of nonsignificant findings. The problem of bias in estimation was addressed using a bootstrap estimation procedure, based on 10,000 samples drawn from the covariance matrix of the model, which provides robust estimates. In particular, the statistical significance of the indirect effects was evaluated based on percentile 90% confidence intervals calculated on the bootstrapped samples. The chi-square test (Jöreskog & Sörbom, 1996) was used to assess the strict goodness of fit of the model. The model was then assessed for close fit using Hu and Bentler’s (1999) criteria with the cut-off point of .95 for the Comparative Fit Index (CFI) and the Non-Normed Fit Index (NNFI), and of .05 for the Root Mean Square Error of Approximation (RMSEA). Results Data description Table 1 shows the descriptive statistics, correlations, and internal consistency coefficients of the study variables. The study variables had from satisfactory to good internal consistency, with the exception of trait extrinsic motivation that just failed to reach the satisfactory level of .7. All hypothesized correlations were of the expected sign and significant, with the exception of the nonsignificant correlation between deep approach to studying and academic performance. -------------------------------------- Insert Table 1 about here --------------------------------------
The chi-square test for the hypothesized model was significant (chi-square = 1624.86, df = 827, p < .001), indicating that model does not fit strictly. However, the other indices reveal that the model fits closely (CFI = .96; NNFI = .95; RMSEA= .051, 90% CI = .047 – .055). The model explained 90% of variance in academic performance, 46% in positive affect, 48% in negative affect, 53% in deep, 45% in strategic, and 54% in surface approaches to studying, 53% in use of creative cognition in studying, and 34% in evaluation anxiety. Figure 2 shows the model with estimated standardized path coefficients. All the paths from approaches to studying to academic performance were non-significant, which does not support hypothesis 1. The path from positive affect to academic performance was positive and significant, whereas the path from negative affect to academic performance was nonsignificant; as such, hypothesis 2a is supported whereas hypothesis 2b is not. The paths from the use of creative cognition to positive affect, deep and strategic approaches to studying were all positive and significant, which supports hypothesis 3. The paths from evaluation anxiety to negative affect and surface approach to studying were all positive and significant, which supports hypothesis 4. The path from trait intrinsic motivation to deep approach to studying was positive and significant, whereas the path from trait extrinsic motivation to surface approach to studying was negative and significant; as such, hypothesis 5a is supported, whereas hypothesis 5b is disconfirmed. The paths from trait intrinsic motivation to use of creative cognition and from trait extrinsic motivation to evaluation anxiety were both positive and significant, which supports hypothesis 6. The paths from adaptive metacognition to positive affect and strategic approach to studying and the paths from maladaptive metacognition to negative affect and surface approach to studying were all positive and significant, which supports hypothesis 7 with the exception of hypothesis 7c stating a path from adaptive metacognition to deep approach to studying. Finally, the paths from adaptive metacognition to the use of creative cognition and the path maladaptive metacognition to evaluation anxiety were both positive and significant, which supports hypothesis 8. In all, hypotheses 2a, 3, 4, 5a, 6, 7a, 7b, 7d, 7e, and 8 are supported; hypotheses 1, 2b and 7c are not supported; hypothesis 5b is disconfirmed. Of the 16 hypothesized intermediate paths, the findings fail to support only that from adaptive metacognition to deep approach to studying. Of the five hypothesized direct effects on academic performance, the findings support only that from positive affect to academic performance. As such, the key finding is that the only significant direct predictors of academic performance are past semester academic performance and positive affect in studying. -------------------------------------- Insert Figure 2 about here --------------------------------------
Two sensitivity analyses were conducted. In the first, we examined the assumption that the structure variables (motivational orientations and metacognitions) and the first-level mediators (creative cognition and evaluation anxiety) do not have direct effects on academic performance. This was achieved by adding a direct path from each of these variables to academic performance, and comparing the expanded model to the hypothesized model by the difference of chi-square test. The difference in fit between the two models was not significant (chi-square = 6.77, df = 6, p < .343). Moreover, the t tests of each of the added paths were nonsignificant (results not shown). These findings suggest that the direct effects of the trait variables and first-level mediators on academic performance are negligible. In the second analysis, we examined the assumption that positive affect in studying originates from the actual experience of studying, rather than from the reinforcement provided by previously earned grades. This was achieved by adding a path from past academic performance to positive affect, and comparing the expanded model to the hypothesized model. The difference in fit between the two models was not significant (chi-square = .22, df = 1, p < .639), and the standardized coefficient of the added path was weak (.01). This finding suggests that the reinforcement process such that good past academic performance fosters positive affect is negligible, and hence that positive affect reflects mainly the positive emotions of studying.
Table 2 shows the estimated indirect effects of the structure variables and first-level mediators on academic performance, and their bootstrapped percentile 90% confidence intervals. Four indirect effects were significant, and each was of the hypothesized direction. With reference to the structure variables, both positive-adaptive traits had positive indirect effects on academic performance, whereas both negative-maladaptive traits had no indirect effects on academic performance. In particular, trait intrinsic motivation had a positive effect through the use of creative cognition (first-level mediator) and positive affect (second-level mediator) (path chain 1). Moreover, adaptive metacognition had a positive effect through the use of creative cognition (first-level mediator) and positive affect (second-level mediator) (path chain 4), and a less indirect positive effect through positive affect (path chain 7). As such, trait intrinsic motivation and adaptive metacognition are indirect predictors of academic performance, whereas trait extrinsic motivation and maladaptive metacognition are neither direct nor indirect predictors of academic performance. Turning attention to the first-level mediators, creative cognition had a positive indirect effect on academic performance through positive affect (path chain 10), whereas evaluation anxiety had no indirect effects on academic performance. The finding on creative cognition suggests that the use of creative cognition fosters academic performance independently of a student's levels of trait intrinsic motivation and adaptive metacognition. The finding on evaluation anxiety closes the circle on the negative-maladaptive submodel, showing that no variable in that submodel predicts academic performance either directly or indirectly. -------------------------------------- Insert Table 2 about here --------------------------------------
This study tested a model of end-of-semester academic performance, in which motivational orientations and metacognitions were the predictor variables, use of creative cognition and evaluation anxiety were the first-level mediator variables, affect and approaches to studying were the second-level mediator variables, and prior semester academic performance was the control variable. The model allowed to examine simultaneously the effect of adaptive-positive and maladaptive-negative structures and processes on students’ academic performance, and to determine which of the two are overall more influential on students’ academic performance. Five key findings from this study shed light on the psychological variables that most influence academic performance, and hence are candidates for future intervention studies. Key findings Positive affect as predictor Positive affect in studying was the strongest and sole psychological and direct predictor of students’ academic performance. This finding is consistent with prior empirical studies (e.g., Dosseville, Laborde, & Scelles, 2012; Rogaten et al., 2013) and various theories of emotions. Drawing from the mood-as-input theory (Martin et al., 1993), students who experience positive affect in studying are likely to interpret their emotions as a sign that the activity is enjoyable and hence are likely to devote more time and effort to studying. Drawing from the broaden-and-build model (Fredrickson, 1998, 2001), students who experience positive affect should accrue more psychological, physical, and social resources and, in turn, learn more and perform better academically. Finally, based on the control-process theory (Carver & Scheier, 1990, 2001), positive affect in studying can be seen as a subjective judgment on one’s own learning progress, in such a way that students who experience more positive affect in studying perceive themselves as learning faster than anticipated. In all, this study supports the crucial importance of positive affect to learning and academic performance, and is consistent with both the conjecture that positive affect fosters learning and the conjecture that positive affect signals that progress in learning is faster than anticipated. Positive affect as mediator Positive affect in studying mediated the positive effects of trait intrinsic motivation, adaptive metacognition, and use of creative cognition on academic performance. In particular, trait intrinsic motivation had a positive indirect effect on academic performance through the use of creative cognition (first-level mediator) and positive affect (second-level mediator). This finding is in line with Amabile and co-workers (1986) and Hennessey and co-workers (1989), who consistently found that intrinsic motivation – measured as either a state or trait – facilitates creative output, and with Amabile and co-workers (1996), who found that trait intrinsic motivation correlates with grades. Nevertheless, this finding goes beyond previous studies by showing that: (a) trait intrinsic motivation also enhances positive affect by fostering the use of creative cognition, and (b) the positive relation between trait intrinsic motivation and grades is entirely due to the mediating effects of use of creative cognition and positive affect, in that order. Moreover, adaptive metacognition had both a direct and indirect effect, through the use of creative cognition (first-level mediator), on positive affect (second-level mediator) and, in turn, academic performance. These findings are generally consistent with Antonietti and co-workers (2000) and Swanson (1990, 1992), who identified metacognition as an important contributor to creative problem solving, and suggest, in addition, that: (a) adaptive metacognition also enhances positive affect directly and indirectly, by fostering the use of creative cognition, and (b) the positive relation between adaptive metacognition and grades is entirely due to the chained mediating effects of use of creative cognition and positive affect, in that order. In all, students who are intrinsically interested in studying (trait intrinsic motivation) and have adaptable hierarchies of learning goals and a mind set for problem solving (adaptive metacognition) appear to be more likely to use divergent and convergent thinking, metaphorical and analogical thinking, perspective taking, or visualization strategies (use of creative cognition) when tackling academic problems and studying in general. In turn, students who use more creative cognition in studying would tend to experience more positive affect in studying and, in turn, perform better academically. As such, it is possible that positive affect in studying channels and converts positive structures and processes into better academic performance.
Besides functioning as a first-level mediator, the use of creative cognition was the strongest stand-alone direct predictor of positive affect in studying and the strongest stand-alone indirect predictor (through the mediation of positive affect) of academic performance. These findings are consistent with Rogaten and Moneta (2015b), who found that use of creative cognition in a semester predicts positive affect in the following semester, and provide two extensions: (a) the effect of use of creative cognition on positive affect appears to pass on academic performance, and (b) the use of creative cognition appears to have an effect on academic performance irrespective of a student's levels of trait intrinsic motivation and adaptive metacognition. As such, the use of creative cognition is a crucial variable in the model, and an attractive intervention target even for students who stand low on trait intrinsic motivation and adaptive metacognition. Uninfluential negative-maladaptive submodel All the psychological variables that were hypothesized to undermine learning and hence academic performance did not predict academic performance. On the one hand, maladaptive metacognition and trait extrinsic motivation predicted evaluation anxiety, negative affect, and surface approach to studying, consistent with a wealth of empirical studies (e.g., Prat-Sala & Redford, 2010; Spada & Moneta, 2012, 2014). On the other hand, all the maladaptive-negative psychological variables included in the model failed to explain additional variance in academic performance to that explained by adaptive-positive psychological variables, consistent with a similar study that, however, considered a subset of the variables included in the present study (Rogaten et al., 2013). These findings suggest that adaptive-positive structures and processes are on the whole better predictors of academic performance than maladaptive-negative structures and processes. In all, academic performance seems to be influenced more by the presence of positivity than by the absence of negativity. Independence of submodels The correlations between the variables of the adaptive-positive and maladaptive-negative submodels were weak. This implies that, for example, if a student has low levels of trait extrinsic motivation, maladaptive metacognition, evaluation anxiety, negative affect, and surface approach to studying, no inference can be made on that student's levels of trait intrinsic motivation, adaptive metacognition, use of creative cognition, adaptive approaches to studying, and positive affect in studying. In all, the investigated adaptive-positive and maladaptive-negative structures and processes in learning seem quite independent of each other, and the adaptive-positive structures and processes are way more relevant to academic performance. Potential applications The weak relationship between adaptive-positive and maladaptive-negative structures and processes has an important implication for the design of educational interventions: intervening on trait extrinsic motivation, maladaptive metacognition, evaluation anxiety, negative affect, and surface approach to studying may improve students' experience but will not result in the increase of adaptive-positive structures and processes, and will not result in higher academic performance. As such, the most promising opportunity for improving students’ academic performance is to intervene on adaptive-positive psychological predictors of academic performance. Based on the found relationships between the use of creative cognition in studying, positive affect in studying, and academic performance, it seems that educational interventions aiming to foster students’ academic success should be primarily directed at enhancing positive affect in studying. This can be achieved directly – e.g., through infusing enthusiasm in students, challenging students intellectually, and providing encouraging supervisory support – or indirectly, by intervening on variables that foster positive affect in studying. However, intervening directly on positive affect can be problematic, as sensitivity to emotion-eliciting stimuli is largely determined by temperament (Clark & Watson, 1999), notably extraversion (Gomez, Cooper, McOrmond, & Tatlow, 2004). It therefore is more viable to intervene on variables that foster positive affect, among which the use of creative cognition emerged as the target variable of choice in the present study. Given that every student can use creative cognition when coping with study problems, and can be encouraged and trained to do so, intervening on students’ use of creative cognition in studying is the most promising strategy for interventions aimed at fostering positive affect in studying and, in turn, academic performance. Although academic performance undoubtedly is an important target variable for any educational intervention, the emerging target variable in Higher Education is students’ creative ability (e.g., Csikszentmihalyi, 2014; Dino, 2015; Moyer & Wallace, 1995). As such, both academic performance and its best predictor – positive affect – can also be viewed as instrumental to the overarching goal of fostering students’ creative ability, as they provide intrinsic and extrinsic reinforcement to the use of creative cognition. Nevertheless, the use of creative cognition can and should also be targeted directly in order to foster development over and above “natural” development. In what follows, we propose four principles that should guide any such intervention. First, students should be given creative tasks, that is, tasks for which creativity is both possible and desirable. Amabile (1982, 1996) proposed a distinction between “algorithmic” and “heuristic” tasks, which can help to identify creative tasks. A task is algorithmic if someone is given beforehand a complete set of steps for completing the task, and completing the task is only a question of carrying out the steps. Instead, if discovering the steps is part of the task itself, then the task is heuristic. In order to be creative a problem must be heuristic, that is, it should not have a clear and readily identifiable path to a solution. As such, the minimal condition is that students be given plenty of heuristic problems to practice with. Moreover, students should be confronted with hard, ill-conditioned heuristics problems, such as problems with no clear path to a solution, problems with multiple paths to a solution, problems with no solution at all, problems with unstated constraints, and problems to which no general rule applies (e.g., Sternberg, 2006). These are the kind of problems humanity is confronting on a daily basis, such as predicting financial crises, addressing global warming, or preventing war, and hence it should not be hard to explain to students why they are asked to tackle tough problems. Second, when given creative tasks, students should be asked to work on them from beginning to end, completing all the phases of the creative process identified, for example, in Amabile’s (1983, 1996) componential model of the creative process: task representation, preparation, response generation, response validation, and outcome evaluation. The practical wisdom of doing so is that ideas that are creative but not well formed and well presented are rarely recognized and rewarded, and are sometimes stolen by somebody who knows how to develop them into full-fledged and winning ideas. A few historical examples could easily convince students of the importance of developing and bringing to fruition their creative ideas. Third, students should be given clear feedback on the contextual appropriateness of their creative attempts. As Kaufman and Beghetto (2013) humorously put it, whereas it is important to teach students to be creative, it is equally important to teach them when not to be creative. For example, it is not uncommon that a paragraph in an essay or report uses multiple terms to refer to the same concept or variable, creating unnecessary confusion in the reader. It is only by receiving appropriateness feedback that students can develop the metacognition of creativity and the ability to read the contextual cues that constrain the deployment of creativity. Finally, building on the previous points, it is necessary to assess students’ creative ability and their development in the course of their studies using performance-oriented methods in addition to standardized tests of divergent and convergent thinking. In this connection, the key assumption underlying the consensual definition and assessment technique of creativity (Amabile, 1982, 1996) is that although certain thinking processes – which can be measured using standardized creativity tests – and personality characteristics – which can be measured using standardized personality questionnaires – might be associated with creativity, they are not, themselves, creativity. Ultimately, it is in the fruit of those thinking processes and personality dynamics, in the actual work produced by the individual, that creativity manifests itself. From this perspective, the most appropriate measure of students’ creative ability is the level of creativity exhibited in their work – be it examination, coursework, or presentation – as evaluated by independent experts in the field who are blind in respect to students’ identity. For this reason, creative ability and its development should also be measured using the consensual assessment technique on numerous, repeated samples of student work produced throughout the course of study. Limitations and directions for future research The findings of this study should be considered in the light of four key methodological limitations. First, this study is cross-sectional and hence cannot imply causation. Future research should test the hypothesized causal relationships using longitudinal study designs. Second, the sample size is relatively small given the complexity of the model, and hence the power of tests is limited. This implies that some of the relationships that were found nonsignificant in this study – such as those involving maladaptive-negative psychological variables as predictors or mediators of academic performance – may turn out to be significant in larger samples. As such, replications on larger samples are needed. Third, this study gathered data from a heterogeneous sample of students from various faculties, degree levels and ethnic backgrounds, which is an appropriate choice of sample for an initial testing of the model. However, future research should test the model on larger and more homogeneous student samples to see if the found relationships hold, in particular, for different degree subjects and different degree levels. This is because the relationships between the variables in the model may be influenced by seniority, as students tend to focus more on their academic performance toward the end of their degree, and age, as mature age students tend to take their studies more seriously from the start. Finally, the classification of psychological variables as either structures or processes was based on theory and previous use of the measurement scales. The cross-sectional study design prevented modeling stability over time, and both structures and processes were measured using self-reports. Therefore, future studies should test whether the hypothesized structure variables are in fact more stable than the hypothesized process variables. Conclusion Despite its limitations, the present study advances our understanding of the relationship between psychological variables and academic performance in university students. The model explained nearly all the variance in academic performance, with prior academic performance and positive affect in studying being the only direct predictors. The use of creative cognition in studying mediated the positive effects of trait intrinsic motivation and adaptive metacognition on positive affect in studying and, in turn, academic performance. The variables thought to have an undermining effect on learning failed to predict academic performance. In sum, this study suggests that any intervention designed to improve students’ academic performance should concentrate on developing their adaptive-positive study behaviors, in particular the use of creative cognition in tackling study problems. Yüklə 207,59 Kb. Dostları ilə paylaş:
|