Learning under Stress and the Influence of Future Time Perspective
Masterscriptie Onderwijswetenschappen Universiteit van Amsterdam
S.F. Koene (10901507) Begeleiding: Dr. L. Andre Tweede beoordelaar: Prof. Dr. T. T. D. Peetsma Amsterdam, 24 juli 2020
2 Abstract
The present study aimed to investigate the moderating effect of future time perspective (FTP) on the relationship between mathematics-related stress manifested as anxiety, and
self-regulated learning (SRL) behavior in mathematics among students in Dutch secondary education. A total of N = 349 students reported their levels of class anxiety, learning anxiety, test anxiety, investment in learning, and future time perspective through a digital
questionnaire. Results showed a significant negative association between the different types of anxiety and investment in learning, as expected. A significant negative association between FTP and anxiety was found, as expected. The moderation analyses of FTP on these
relationships were also found to be significant, as expected. However, contrary to the
proposed hypothesis, FTP was found to be enhancing the negative association between stress and SRL behavior rather than regulating it in a positive direction. Results and possible explanations are thoroughly discussed.
Keywords: Future time perspective, investment in learning, stress, class anxiety, learning anxiety, test anxiety.
Introduction
Students have to maneuver themselves between a lot of potential stressful factors like homework, exams, and conflicting deadlines. Too much stress can influence their learning behavior and possibly even lower academic performance. It is important to help students have a successful academic career and therefore to understand the influence of stress on their learning behavior and to identify possible moderating variables. Moderating variables may mitigate or increase the effect of one variable on the other. Lowering the possible negative effects of stress on learning behavior can therefore be achieved by the influence of
moderating factors.
Stress is known to have a negative effect on learning behavior and thus influence academic achievement in students (Vogel & Schwabe, 2016). Individuals under low amounts of stress exhibit flexible and cognitive learning styles like self-regulated learning behavior (SRL). SRL behavior refers to the constructive and active process of learning and is known to increase motivation and performance (Pintrich, 2000). Identifying the negative effects of stress on SRL behavior is therefore an important step to increase academic performance.
3 One of the possible variables that could moderate the relationship between stress and SRL behavior is future time perspective (FTP). Individuals with a high FTP score have a positive outlook on the future with clear goals (Peetsma, 2000). Having a high FTP score is a motivating factor for many different aspects in life and can therefore increase self-regulation. FTP can also make students more resilient to contextual influences on their self-regulated learning behavior such as stress and therefore possibly moderate the effect that stress has on SRL behavior.
The relationship between stress, SRL behavior and FTP is not clearly determined yet, and the current study aims to contribute to clear this up. Knowing the effect of FTP on the relationship between stress and SRL behavior might help students perform better. If FTP can help students cope better with stress and improve their academic achievement this could result in motivated and successful students as well as better support systems for students suffering from stress. Therefore, this study aims to determine the moderating effect of FTP on the association between perceived mathematics-related stress and SRL behavior of students.
Theoretical framework
Perceived stress
Stress is a part of life that every individual encounters and can be seen as the reaction of one’s body and mind to adverse situations (Shahmohammadi, 2011). Lazarus and Folkman (1984) emphasized the relationship between a person and their environment when defining stress. They stated that stress is caused by stressors, external factors in the environment that can trigger stress in an individual. Perceived stress therefore depends on the character of the individual as well as the nature of the environment and the appraisal of this relationship by the individual. The individual evaluates whether their relationship with the environment is
manageable or noxious to their well-being. The amount of stress and the way individuals cope with it can therefore differ greatly from person to person (Lazarus, 1993).
When studying stress as a construct it is important to note that different types of stress have different impacts on an individual’s psychological state. Selye (1975) divided stress in two categories: positive stress (eustress) and negative stress (distress). Positive stress can motivate an individual, whereas negative stress causes negative physical and psychological reactions such as physical pain, anxiety, anger or worry (Shahmohammadi, 2011). LePine, LePine and Jackson (2004) determined that positive stress can improve learning performance, whereas negative stress reduces learning performance. The current study will focus on
4 negative stress because gaining insight on how to reduce the influence of distress on learning behavior can help students perform better.
Stress can trigger both physical and psychological reactions in the human body. These results of stress are called stress reactions (Lazarus, 1993). Stress reactions can be measured in various ways and can provide an indication on the stress that an individual experiences. Emotions are part of psychological stress reactions and are therefore an interesting field of study. Lazarus (1993) identified 15 emotions and the current study will focus on one of them; anxiety, which is a common stress reaction to negative stress. Previous research has shown that anxiety was the emotion reported most often by students, especially in relation to exams, but also while doing homework or being in class (Pekrun, Goetz, Titz & Perry, 2002).
Pekrun (2006) puts anxiety under a category of emotions that are tied directly to achievement activities and outcomes called achievement emotions. Achievement emotions can be subcategorized in two categories: activity emotions and outcome emotions. Anxiety can be seen as both an activity emotion and an outcome emotion, depending on the situation. For example, when a student feels anxious during mathematics class or while doing
homework because they find mathematics difficult, this can be seen as an activity emotion. They are not anxious about the outcome of anything, but about the task itself. However, being anxious about failing a test is an outcome emotion. This applies for anxiety experienced before and after a test as well as during. Although one could argue that anxiety experienced during a test can also be an activity emotion as the anxiety might pointed towards the activity more than the possible failure.
Pekrun (2006) also addressed the difference between trait- and state achievement emotions. State emotions are linked to a given situation and a given moment in time, while trait emotions are a characteristic of an individual. Trait emotions can, however, still be situation-specific, such as trait test anxiety. This study will focus on trait emotions in three specific situations: during mathematics class, while making mathematics homework and before/during mathematics tests. These three situations are separated to distinguish activity emotions from outcome emotions. Also, from previous studies it is known that general anxiety correlates less with students’ achievement than test anxiety (Hembree, 1988). So, the relationship between test anxiety and SRL behavior is expected to be stronger than the relationship between class anxiety or learning anxiety with SRL behavior.
As mentioned before, an emphasis will be put on mathematics-related stress in this study. Mathematics anxiety is a significant factor in student’s avoidance of math-related career tracks (Ashcraft, 2002). Stress related to mathematics is a common problem among
5 students and therefore mathematics is an interesting school subject to investigate when
studying stress (Ahmed, Minnaert, Kuyper & van der Werf, 2012). Besides, because the majority of students in Dutch secondary education take mathematics classes, a large and diverse sample group can be gathered. By focusing only on mathematics-related stress, students do not need to report on all the stressful events in their lives, making it a less intrusive study for them. Mathematics-related stress was determined by measuring mathematics-related learning anxiety, classroom anxiety and test anxiety.
Self-regulated learning behavior
Self-regulated learning (SRL) is an indicator of academic achievement that has been defined in multiple ways in the field of educational sciences (Mega, Ronconi, & De Beni, 2014). The definition used in the current study is that of Pintrich (2000) and describes SRL as an “active, constructive process whereby learners set goals for their learning and attempt to monitor, regulate and control their cognition, motivation, and behavior, guided and
constrained by their goals and contextual features in the environment”. In this definition four areas of self-regulation are included; cognition, motivation, behavior, and context. The first three areas are internal to the learner and show overlap in many cases. For example, the behavioral area overlaps with the cognitive component, because cognition, forethought and planning are needed to monitor and regulate behavior effectively (Pintrich, 2000). The fourth area (context) is external and therefore less in the learner’s control. However, the context and environment of a learner can influence their degree of self-regulation greatly.
This study will focus on the behavioral component of SRL, referred to as SRL
behavior. Pintrich (2000) included in this area all behavioral strategies that individuals use to regulate their learning and motivation. For example, learning behavior, decision making, seeking help and level of persistence, making it a rather broad term. To measure SRL behavior, the current study will focus on one aspect of SRL behavior; school investment (or investment in learning). Peetsma and van der Veen (2011) argued that investment in learning is in fact a component of self-regulated learning and therefore measuring it is a means to measure SRL behavior. They found investment in learning to be positively related with academic achievement and FTP.
As mentioned before, SRL behavior can be influenced by the context or environment of the learner. One of the contextual features that could constrain SRL behavior is perceived stress. Learning behavior in general is known to be negatively affected by negative stress (Vogel & Schwabe, 2016). Where individuals would normally exhibit flexible and cognitive
6 learning styles, individuals under stress behave more rigid and display habit-like learning behavior. Drawing on these conclusions, the first hypothesis of the current study is as follows; perceived stress manifested as mathematics-related class anxiety, learning anxiety, and test anxiety relates negatively to SRL behavior in mathematics manifested as investment in learning.
Future Time Perspective
Almeida (2005) described several factors that can increase or decrease a person’s vulnerability to stress. One of the psychosocial factors that he mentioned are life goals. Life goals can be a motivational factor that helps an individual cope with stress and “pull through” to reach their goal. However, not just the goals an individual has for the future are weighing in, also their motivation and the extent to which they think about the future, and the action that an individual puts in present for researching those goals in the future. The sum of all these factors is defined as future time perspective (FTP). Many definitions and
conceptualizations of FTP have been used in the literature over the years (Andre, 2018; Shipp, Edwards, & Lambert, 2009). This study will assume the definition used by Peetsma (2000) and Husman and Shell (2008) stating that FTP is “an attitude towards goals in the distant future that motivates learners through the value they attach to distant future goals”. This attitude contains three components; feelings, cognition, and behavioral intentions/behaviors that individuals have towards the future. Feelings towards the future are emotions that can be positive and negative and therefore can influence an individual’s attitude (Andre, van Vianen, Peetsma, & Oort, 2018). Cognition includes any expectations, beliefs, or ideas an individual has about the future. Finally, behavioral intentions/behaviors include the set goals, plans and actions that an individual has for the future. These three components together sum up the construct of FTP.
In the literature, FTP has been regarded as a motivator in different life domains. Peetsma (2000) distinguished four objects or life domains; school and professional career, social relations, personal development, and leisure. The current study is concerned with FTP that influences the effect of math-related stress on learning behavior and therefore will focus on the FTP of school and professional career.
Specifically applied to academic related stress, the motivation to obtain a diploma or to be promoted to the next schoolyear can impact the influence of stress on academic
achievement greatly. The life goals mentioned by Almeida (2005) are assumed to have a moderating influence on the relationship between stress and well-being by increasing a
7 person’s resilience to stress. Drawing on this assumption, a similar, possibly even stronger effect could be expected from FTP. This also matches the theory of stress appraisal and coping that Folkman and Lazarus (1984) formulated. Having a positive outlook on the future can also mean having a positive attitude towards stressful situations. By appraising stressful situations more positively because you “take it lightly”, it is easier to regulate distressing emotions. Folkman, Lazarus, Pimley and Novacek (1987) defined this coping technique as emotion-focused coping. FTP can also be a motivator to change a stressful situation at hand to reach future goals, this coping technique is called problem-focused coping (Folkman et al., 1987). A previous meta-analysis study has indeed found a negative relationship between FTP and anxiety of moderate to large effect size (Kooij, Kanfer, Betts, & Rudolph, 2018). Drawing on these conclusions, the second hypothesis of the current study is as follows; FTP on school and professional career is negatively associated with perceives stress, manifested as
mathematics-related class anxiety, learning anxiety and test anxiety.
Besides influencing stress, FTP also has an influence on SRL behavior. De Bilde, Vansteenkiste and Lens (2011) found that FTP was found to be positively associated with identified regulation. Identified regulation is the most internalized form of self-regulation (Ryan & Deci, 2000). Therefore, the positive effect that FTP has on identified regulation is assumed to also increase SRL behavior. By both lowering stress and increasing SRL behavior, FTP is expected to be a moderator to the negative association between stress and SRL
behavior (see Figure 1). These expectations lead to the third hypothesis of the current study; FTP is a significant moderator to the negative association between perceived mathematics-related stress and SRL behavior. By testing the three hypotheses proposed above, the current study ultimately aims to test the accuracy and significance of the model shown below (Figure 1).
Figure 1: Overview of hypothesized relationship between FTP, SRL behavior and perceived
stress of the current study.
Method
8 Participants of this study were Dutch students involved in secondary education from one big school. The average age of the students was 15.4 years (SD = 2.66, range 12.5 – 19.2). This school offers mavo1, havo2 and vwo3 level of education. These levels are indicative of how vocational or theoretical the provided education is.
Participation was preceded by an informed-consent procedure that required active consent from both students and parents (for students who were below 16 years old). An online questionnaire was administered that students could complete during their mentor class by clicking on the link or scanning the QR code to access the questionnaire. The total sample size was N = 351 students (128, males, 190 females, and 31 students who did not report their gender).
Measures
Perceived stress. Perceived stress was assessed by measuring students’ negative
response to stress, for which anxiety is a measure. Thus, the Anxiety scale from the
Achievement Emotions Questionnaire – mathematics (AEQ-M) by Pekrun, Goetz, & Frenzel (2005) was used. The AEQ-M is a questionnaire in which respondents self-report about their achievement emotions related to mathematics. In the current study only questions about anxiety as an achievement emotion were used. The anxiety scale included three subscales that measured different types of students’ anxiety in mathematics: class anxiety, learning anxiety and test anxiety. The validity of the AEQ-M questionnaire was confirmed in a cross-cultural study (Frenzel, Thrash, Pekrun, & Goetz, 2007).
Class anxiety subscale consisted of four questions with a five-point Likert scale ranging from 1 to 5. This questionnaire contained questions like: “I am so afraid of
mathematics that I would rather not go to class”. The subscale had a good reliability in the current study (Crohnbach’s α = .86). The subscale for learning anxiety consisted of four questions with a five-point Likert scale ranging from 1 to 5 and included questions like: “When I am studying for mathematics, I feel tense and nervous”. This subscale had a good reliability in the current study as well (Crohnbach’s α = .85). And finally, the subscale for test anxiety consisted of seven questions with a five-point Likert scale ranging from 1 to 5 and consisted of questions like: “During a mathematics test I am worried I will receive a bad
1 Mavo: middelbaar algemeen voortgezet onderwijs [Lower general secondary education] 2 Havo: hoger algemeen voortgezet onderwijs [Higher general secondary education]
9 grade”. This subscale had an excellent reliability in the current study (Crohbach’s α = .93). Over all three subscales there were no items that needed recoding.
SRL behavior. SRL behavior was determined by measuring students’ investment in
learning in class through a questionnaire based on the investment in learning in class subscale of the School Investment Questionnaire developed by Roede (1989) that was adjusted for mathematics. The subscale consisted of five questions with a five-point Likert scale ranging from 1 to 5 through which respondents self-reported about their investment in learning during mathematics classes. The subscale was operationalized by three aspects: the onset of students' behavior; the degree of intensity in learning, and perseverance in learning (Andre et.al., 2019). The validity of this questionnaire has been confirmed in previous studies (Roede, 1989; Peetsma, 2000; Peetsma, Hascher, & van der Veen, 2005; Andre et. al., 2019). The questionnaire included items like: “During mathematics class I work hard”. This scale had a good reliability in the current study (Crohnbach’s α = .86). There were no items that needed recoding.
FTP on school and professional career. FTP on school and professional career was
measured using the Future scale of the Time Perspective Questionnaire (TPQ) developed by Peetsma (1992). The measurement of FTP on school and professional career was conducted simultaneously with the measurement of SRL behavior and perceived stress. The Future scale of the TPQ is a questionnaire in which respondents self-report about their long-term future related to school. In contrast to other FTP scales, this scale includes a combination of measures of cognition, affect and behavioral intent towards the future while specifying context of future thinking as being related to school and professional career. The validity of this scale has been confirmed in previous studies (Peetsma, 1993; Stouthard & Peetsma, 1999). The questionnaire included seven questions with a five-point Likert scale determined the overall FTP on school and professional career score of the individuals. The questionnaire included items like: “I enjoy thinking about my future studies or work”. This questionnaire had a good reliability in the current study (Crohnbach’s α = .71). Three negatively phrased items were recoded.
Data analysis plan
The analyses of data consisted of four steps. Firstly, as part of the preliminary analyses, missing data was checked for and the main assumptions relevant for the analyses
10 were tested. Secondly, bivariate Pearson correlation analyses were conducted to test the relationship between the three types of mathematics-related anxiety with investment in learning and FTP (Agresti & Franklin, 2013). Thirdly, to determine the explained variance of perceived daily stress to SRL behavior, three single regression analyses with 95% probability interval were conducted with class anxiety, learning anxiety and test anxiety as the predictors and investment in learning as the dependent variable (Agresti & Franklin, 2013). Finally, to explore the moderating influence of FTP on the relationship between daily perceived stress and SRL behavior, three moderator analyses were conducted on the regressions between the three types of anxiety and investment in learning with FTP as the moderating value. The analyses were conducted using the PROCESS macro for SPSS, version 3.5 (Hayes, 2020). The difference between males and females on each variable was tested using a same-sample t-test.
Results
Data screening and preliminary analyses
Descriptive statistics of all the study variables are shown in table 1, including correlations between all study variables. Little’s MCAR test was significant (p = .23), indicating that there was no pattern in the missingness of our data. It was therefore not necessary to do any data replacement. SPSS’ listwise deletion of missing data was used in further analyses. For the investment in learning scale, class anxiety, learning anxiety and test anxiety, less than 5% of data was missing. While future time perspective 7.1% of data were missing. Overall, the total amount of missing values was less than 5% and therefore
considered non-problematic.
Of the N = 351 students’ responses, several outliers were identified by generating boxplots of which two were removed due to having filled in the same response on each question. This brought the final sample size to N = 349 students that were included in the main analyses. Other outliers were not removed because no patterns were detected in the responses and reversely coded items were answered adequately.
The kurtosis and skewness tests showed that all variables were normally distributed as all values fell between 1.96 and -1.96. All variables were very homoscedastic and VIF’s well below 5 showed there were no concerns about multicollinearity among variables.
A significant difference between males and females was found concerning class anxiety (Mfemales= 2.02, SD = 0.97; MMales = 1.59, SD = 0.65), t(316) = -4.410, p = .000. The
11
SD = 0.69), t(316) = -5.728 , p = .000 and for test anxiety (Mfemales= 2.85, SD = 1.12; MMales =
2.00, SD = 0.85), t(316) = -7.244 , p = .000. These results show that female students had significantly more mathematics related anxiety on class-, learning- and test anxiety than male students. No significant differences were found between males and females concerning FTP or investment in learning.
Regression analyses were done with and without controlling for the gender of the students, yielding similar results. Therefore, the results for correlation, regression and moderation are further presented without control for the students’ gender.
Table 1. Descriptive statistics and correlations of all study variables.
Variables 2. 3. 4. 5. M SD 1. Class Anxiety .81* .71* -.30* -.15** 1.86 .90 2. Learning Anxiety - .79* -.28* -.15** 1.92 .95 3. Test Anxiety - -.22* -.05*** 2.51 1.11 4. Investment in Learning - .37* 3.07 .90 5. FTP School - 3.78 .65
Note. * p < .001, **p < .01, ***Not significant. All scales had a range of 1-5.
Data analyses
Correlations between the three types of anxiety and investment in learning are reported in Table 1. As shown, each type of anxiety was negatively related to investment in learning. That is, test anxiety, class anxiety and learning anxiety all showed a significant negative correlation of medium effect size with investment in learning, as hypothesized.
Correlations between the three types of anxiety and FTP are reported in Table 1. As expected, the relationship between FTP and all types of anxiety are negative. The effect size of the correlation between learning anxiety and FTP and between class anxiety and FTP can both be considered moderate (Keith, 2006). The effect size of the correlation between test anxiety and FTP can be considered small, but meaningful. The correlation between FTP and test anxiety was, however, not significant.
To test the assumption of the model shown in figure 1 that perceived stress is a predictor for SRL behavior, three single regression analyses were carried out with class
12 anxiety, learning anxiety and test anxiety as predictors and investment in learning as the dependent variable. As shown in Table 2, the explained variances (R2) were small but
significant for all three types of anxiety. As stated by Keith (2006), effect sizes of larger than .05 can be considered meaningful. This indicates that indeed class anxiety, learning anxiety and test anxiety are predictors for investment in learning.
Table 2. Class anxiety, learning anxiety and test anxiety as predictors of investment in learning. Variable Investment in learning B R2 F df 95% CI* Class anxiety -.30 .087 31.77 334 [-.40, -.19] Learning anxiety -.28 .079 28.68 333 [-.36, -.17] Test Anxiety -.22 .049 17.23 331 [-.27, -.09]
Note. *p < .01 for all three regressions.
Moderation analyses were carried out for each type of mathematics related anxiety. For the moderation of FTP on the relationship between class anxiety and investment in learning the effect was significant and largest (B = .20, SE = .08, p < .05, 95% CI [.346, -.050]). The moderation of FTP on the relationship between learning anxiety and investment in learning was significant as well (B = -.16, SE = .074, p < .05, 95% CI [-.307, -.017]). Both these effect sizes can be considered medium-sized (Keith, 2006). The moderation of FTP on the relationship between test anxiety and investment in learning was marginally significant (B = -.11, SE = .064, p = .086 , 95% CI [-.236, .016]), which corresponds with the lack of
13 Figure 2. The interactions of FTP and different types of anxiety on investment in learning a.
Class anxiety, b. Learning anxiety and c. Test anxiety.
When plotting the moderation effects, comparable results were found for class anxiety, learning anxiety and test anxiety. As shown in Figure 2a, the relationship between investment in learning and class anxiety was very large for individuals with a high FTP (B = -.40, SE = 0.07, p < .001, 95% CI [-.55, -.25]), large for individuals with moderate FTP (B = -.27, SE = 0.05, p < .001, 95% CI [-.37, -.17]) and moderate for individuals with a low FTP (B = -.14, SE = 0.06, p < .05, 95% CI [-.27, -.01]). In Figure 2b is shown that the relationship between investment in learning and learning anxiety is large for individuals with a high FTP (B = -.35, SE = 0.07, p < .001, 95% CI [.49, .20]), moderate for individuals with moderate FTP (B = -.24, SE = 0.048, p < .001, 95% CI [-.34, -.15]) and also moderate for individuals with a low FTP (B = -.14, SE = 0.063, p < .05, 95% CI [-.26, -.01]). Finally, in Figure 2c it is shown that the relationship between investment in learning and test anxiety is large for individuals with a
14 high FTP (B = -.26, SE = 0.061, p < .001, 95% CI [-.38, -.14]), moderate for individuals with moderate FTP (B = -.19, SE = 0.041, p < .001, 95% CI [-.27, -.11]) and moderate for
individuals with a low FTP (B = -.11, SE = 0.056, p < .05, 95% CI [-.22, -.005]) as well. From these results it becomes apparent that the third hypothesis of the current study needs to be rejected. Even though FTP is a statistically significant moderator for the
relationship between anxiety and investment in learning, the effect is opposite of what was expected. Students with high FTP score show a larger negative correlation between anxiety and investment in learning than students with a low FTP score. This indicates that a high FTP score increases the effect of anxiety on a student’s investment in learning, rather than
moderates it. A few explanations for these results will be proposed in the discussion.
Conclusion and Discussion
The aim of this study was to empirically investigate the statistical significance of the model shown in Figure 1. To this end, three hypotheses were proposed; (1) mathematics-related anxiety is negatively associated with investment in learning, (2) Future time perspective (FTP) on school and professional career is negatively associated with
mathematics-related anxiety and (3) FTP is a moderator for the negative association between perceived stress and self-regulated learning (SRL) behavior.
Investment in learning is a measure for SRL behavior (Peetsma & Van der Veen, 2011). Any result found for investment in learning in this study was therefore generalized to a result for SRL behavior. Class anxiety, learning anxiety and test anxiety are variables that can be considered part of an individual’s perceived stress. As Lazarus (1993) and Pekrun, Goetz, Titz and Perry (2002) argued, anxiety is an emotion heavily correlated with stress and a common stress reaction. Anxiety can therefore be considered a measure for perceived stress. Therefore, any result found for mathematics-related anxiety in this study was generalized for perceived mathematics-related stress.
Stress and SRL behavior
It was hypothesized that perceived mathematics-related stress was negatively
associated with SRL behavior. The current study showed that significant negative associations were found between all three types of anxiety and investment in learning. This effect was large for class anxiety and learning anxiety and moderate for test anxiety (Keith, 2006). The explained variances of the three types of anxiety as predictors of investment in learning were
15 small but meaningful. This finding supports the conclusion Vogel and Schwabe (2016) made in their article; learning behavior is negatively affected by negative stress.
As mentioned above, the explained variances of SRL behavior by perceived stress are small but meaningful and significant. The fact that the explained variances are so small can be explained by evaluating the definition Pintrich (2000) proposed for SRL behavior. In Pintrich’ definition, SRL behavior is guided and constrained by and individual’s goals and context. Perceived stress is only a small part of what can be considered the “context” of an individual. Many other variables might play a role such as motivation, eustress, buoyancy, self-concept, and meta-emotions and will be discussed in detail below. Some of these variables might be directly correlated with SRL behavior and others might be considered a possible moderator on the relationship between perceived stress and SRL behavior.
First, positive stress or eustress is known to have a positive effect on learning (Selye, 1975; Shahmohammadi, 2011). Anxiety is an emotion that only occurs when experiencing negative stress or distress. But by measuring other stress reaction more associated with eustress, the influence of eustress on SRL behavior can be investigated. Another variable that can have a positive influence on SRL behavior might be motivation. From earlier studies it is known that motivation and SRL behavior both mediate the effect of negative emotions on academic achievement (Mega, Ronconi, & De Beni, 2014). Moreover, Mega, Ronconi and De Beni (2014) found that SRL behavior and motivation significantly correlate with one another and are therefore directly linked. Motivation could therefore increase SRL behavior and make the influence of stress on SRL behavior less extreme.
Other variables such as buoyancy and meta-emotions can be considered possible moderators to the relationship between stress and SRL behavior. Academic buoyancy refers to a students’ ability to successfully deal with academic setbacks (Martin, & Marsch, 2008). Students with high academic buoyancy are less prone to be negatively influenced by stressful situations and might experience less anxiety because of that. It could also be that the anxiety they experience as a lesser effect on SRL behavior because they deal with these negative emotions better. Either way, academic buoyancy is a social construct that is likely to influence the interactions between stress and SRL behavior and a possible moderator.
Meta-emotions are defined as the feelings an individual has about their feelings. In the case of the current study, meta-emotions about anxiety might contribute to the influence the anxiety has on the SRL behavior of the individual. As Goetz, Pekrun, Titz and Perry (2002) stated in their study, the role of meta-emotions has been confirmed in coping with anxiety. Especially a feeling of anger about one’s anxiety seems to be a motivator to find constructive
16 coping techniques that lower anxiety. This might indicate that the measurement of meta-emotions could be of added value to further studies about academic anxiety and stress.
Finally, math self-concept is also known to be negatively correlated with math anxiety (Ahmad, Minnaert, Kuyper, & van der Werf, 2012) and hence is also likely to lower the impact of stress on SRL behavior. Math self-concept is defined as the self-appraisal of an individual regarding their mathematics skills. If a student has trust in his or her own
mathematics skills, he or she is less likely to experience mathematics-related anxiety. Math self-concept is therefore an interesting control variable or study variable to include in further studies.
Future time perspective and stress
It was hypothesized that future time perspective (FTP) on school and professional career was negatively associated with mathematics-related anxiety and thus with perceived
mathematics-related stress. The current study found that FTP on school and professional career was negatively correlated with learning anxiety and class anxiety, the effect sizes were moderate and significant. The correlation of FTP on school and professional career with test anxiety was also negative, but of a small effect size and not significant. This could be
explained when considering the nature of test anxiety in contrast to learning anxiety and class anxiety. Test anxiety is an outcome emotion while learning anxiety and class anxiety are activity emotions (Pekrun, 2006). Test anxiety is linked to a short-term outcome like getting a bad grade or anxiety experienced during a test. This type of anxiety is so temporary that it apparently is not affected by an individual’s thought and expectations about a more distant future. However, since the correlations of FTP with class anxiety and learning anxiety are significant and the direction of the non-significant correlation with test anxiety is still negative, the second hypothesis cannot be simply rejected. FTP does seem to be negatively associated with mathematics-related anxiety.
Future time perspective as a moderator
The third hypothesis of the current study was that FTP on school and professional career moderates the negative association between perceived mathematics-related stress and SRL behavior. The moderation analyses were found to be significant, but the effect was opposite of what was expected. When comparing the regression between investment in learning and the different types of anxiety, the results showed that students with a high FTP were more strongly influenced by anxiety in their investment in learning. In other words, having a
17 positive outlook on the future makes the influence of perceived mathematics-related stress on SRL behavior in mathematics larger. An intervention to enhance a students’ FTP would lower perceived stress and enhance SRL behavior. However, the influence of the perceived stress on the SRL behavior would probably increase, making the choice for such an intervention
doubtful. Instead of helping students cope with the stress they experience, it might have an opposite effect.
From the literature brought forward in the theoretical framework of this study, there is no apparent and direct explanation to these results. Since the moderation analysis was
statistically significant, the model cannot simply be discarded, but possible additions can be considered. A possible explanation to the unexpected negative effect of FTP on the
relationship between stress and SRL behavior could be that there is another variable in the model that goes in between FTP and the relationship between stress and SRL behavior, a variable X (Figure 3).
Figure 3. Proposed model for additional variable in between FTP and the relationship
between perceived stress and SRL behavior
Variable X could be a variable that has a negative moderating effect on the relationship between perceived stress and SRL behavior but is positively affected by FTP or vice versa. From scientific literature a few possible variables can be extracted. First of all, De Bilde, Vansteenkiste and Lens (2011) found that FTP is positively correlated with introjected regulation, a form of motivation based on feelings of guilt and shame. Introjected regulation can have a negative effect on SRL behavior as it is associated with maladaptive coping strategies and a fear of failure (Ryan, & Connell, 1989). Drawing on these conclusions, it is possible that introjected regulation, or rather a lack of introjected regulation, can take the place of variable X in the model proposed in Figure 3.
Secondly, another possible variable to take the place of variable X could be performance pressure. If a student thinks a lot about their future this could possibly distract them from learning or performing well on tests. This phenomenon is known as choking under pressure
18 (DeCaro, Thomas, Albert, & Beilock, 2011). Choking under pressure refers to pressure that induces distraction from the task at hand and therefore negatively influences learning behavior. Of course, pressure experienced due to expectations of the future are not directly linked to the tasks occurring while in mathematics class, doing homework or completing a task. Therefore, this is a distracting form of pressure. Drawing on these conclusions, it could be possible that performance pressure is positively influenced by FTP while negatively influencing the association between perceived stress and SRL behavior. Lowering performance pressure due to expectations of the future could therefore be an interesting intervention objective when trying to enhance SRL behavior or lower the impact of stress.
Finally, another variable that could be considered is the degree of internal control students have over their school performance. From previous research it is known that students with a high degree of internal control show a lower correlation between FTP on school and
professional career and their investment in learning, compared to students with an average or low degree of internal control (Peetsma, 2000). Drawing on these conclusions, internal locus of control can be considered a moderator to the relationship between FTP and SRL behavior and therefore could also show interactions between FTP and the relationship between perceived stress and SRL behavior.
Limitations and recommendations
The current study has some limitations that should be acknowledged. First, the data was collected at a single school making it hard to rule out any confounding variables specific to the context of these students (cohort effect). Therefore, it is recommended for future studies to include more schools from different areas or even different countries. However, the data was collected over different educational levels, grades and from both genders, making it somewhat generalizable.
Secondly, the model proposed in Figure 1 implies a causal relationship of stress negatively influencing SRL behavior. However, based on the current data it is not possible to strongly assume a causal relationship yet. To do so, it is recommended to use a longitudinal design for future research.
In studies like the current one, the issue of self-report is usually mentioned as a limitation. To have a more objective view of certain characteristics of individuals,
observations should be included in the study next to self-report methods. However, anxiety is an emotion and therefore impossible to accurately report about through observations. The same would go for FTP as this is a complex construct related to an individual’s feelings and
19 ideas about the future. The only variable used in the current study that could be considered for observation is investment in learning as this has behavioral components.
Another relevant recommendation would be to make use of state-emotion data to complement trait-emotion data as was also recommended by the developer of the AEQ (Pekrun, 2006). This could be achieved by using a daily diary type of data collection rather than a questionnaire on a set point in time. The advantage of measurements of state emotions over trait is that they might be a closer representation of an individual’s anxiety levels. When filling in a questionnaire about test anxiety, a respondent must think back about their last mathematics test which might be some time ago. The self-report of emotions of the past is probably less accurate than a measurement on the same day the emotion was experienced. Of course, trait emotions can still be situation specific as Pekrun (2006) stated. Therefore, the data collection for test anxiety of the current study is still a relevant approximation of anxiety experienced during mathematics tests.
As discussed earlier, the small values for the explained variance of anxiety for SRL behavior imply that the model is more complex than proposed in Figure 1. Possible relevant variables like motivation, eustress, meta-emotions, math self-concept and buoyancy have been mentioned in this discussion as possible influences on the proposed model. Therefore, it is recommendable to include some of these variables in further research. To explain the
interaction between FTP and the relationship between stress and SRL behavior, variables like locus of control, introjected regulation and performance pressure are also relevant variables to include in future research.
The current study has brought some relevant evidence to explain the relationships between FTP, stress and SRL behavior. FTP related positively to investment in learning. However, when students would experience high levels of anxiety and FTP they would invest less in their learning. To take further steps to a comprehensive model, more constructs need to be included in future research.
20 References
Agresti, A. & Franklin, C. (2013). Statistics: The art and science of learning from data. Third edition. Harlow, UK: Pearson Education.
Ahmed, W., Minnaert, A., Kuyper, H., & van der Werf, G. (2012). Reciprocal relationships between math self-concept and math anxiety. Learning and Individual Differences, 22, 385-389.
Almeida, D.M. (2005). Resilience and Vulnerability to Daily Stressors Assessed via Diary Method. Current Directions in Psychological Science, 14, 64-68.
Andre, L. (2018). Driven by the future: Future time perspective across life domains and
cultures. Ridderprint BV: ISBN 978-94-6299-941-1.
Andre, L., van Vianen, A. E. M., Peetsma, T. T. D., & Oort, F. J. (2018). Motivational power of future time perspective: Meta-analyses in education, work, and health. PLoS ONE,
13. doi:10.1371/journal.pone.0190492
Ashcraft, M.H. (2002). Math Anxiety: Personal, Educational, and Cognitive Consequences.
Current Directions in Psychological Science, 11, 181-185.
DeCaro, M., Thomas, R., Albert, N., & Beilock, S. (2011). Choking Under Pressure: Multiple Routes to Skill Failure. Journal of Experimental Psychology: General, 140, 390–406. Deci, E.L., & Ryan, R.M. (2000). The “what” and “why” of goal pursuits: human needs and
the self-determination of behaviour. Psychological Inquiry, 11, 227-268.
De Bilde, J., Vansteenkiste, M., & Lens, W. (2011). Understanding the association between future time perspective and regulated learning through the lens of
self-determination theory. Learning and Instruction, 21, 332-344.
Folkman, S, Lazarus, R.S., Pimley, S., & Novacek, J. (1987). Age differences in stress and coping processes. Psychology and Aging, 2, 171-184.
Frenzel, A. C., Thrash, T. M., Pekrun, R., & Goetz, T. (2007). Achievement Emotions in Germany and China: A Cross-Cultural Validation of the Academic Emotions
Questionnaire—Mathematics. Journal of Cross-Cultural Psychology, 38(3), 302–309. Hembree, R. (1988). Correlates, causes, effects, and treatment of test anxiety. Review of
Educational Research, 58, 47-77.
Jose, P. E. (2013). Doing statistical mediation & moderation. First edition. New York: Guilford Press.
Kooij, D., Kanfer, R., Betts, M., & Rudolph, C. (2018). Future Time Perspective: A
21 Lazarus, R. S. (1993). From psychological stress to the emotions: A history of changing
outlooks. Annual Review of Psychology, 44, 1-21.
Lazarus, R.S., & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer Publishing Company, Inc.
LePine, J.A., LePine, M.A., & Jackson, C.L. (2004). Challenge and hindrance stress:
Relationships with exhaustion, motivation to learn, and learning performance. Journal
of Applied Psychology, 89, 883-891.
Martin, A. J., & Marsh, H. W. (2008). Academic buoyancy: Towards an understanding of everyday academic resilience. Journal of School Psychology, 46, 53-83.
Mega, C., Ronconi, L., & De Beni, R. (2014). What makes a good student? How emotions, self-regulated learning, and motivation contribute to academic achievement. Journal
of Educational Psychology, 106, 121-131.
Peetsma, T. T. D. (1992). Toekomst als motor? Toekomstperspectieven van leerlingen in het
voortgezet onderwijs en hun inzet voor school [The future as an incentive? Secondary
education students’ perspectives concerning their future and their investment in school] (Doctoral dissertation). University of Amsterdam, Amsterdam, The Netherlands.
Peetsma, T.T.D., (1993) Future Time Perspective as an Attitude: the validation of a concept, paper presented at the 5th Conference of the European Association for Research on
Learning and Instruction (EARLI), Aix-en-Provence, France.
Peetsma, T. T. D. (2000). Future time perspective as a predictor of school investment.
Scandinavian Journal of Educational Research, 44, 177–192.
Peetsma, T.T.D., & van der Veen, I. (2011). Relations between the development of future time perspective in three life domains, investment in learning, and academic achievement. Learning and Instruction, 21, 481-494.
Peetsma, T.T.D, Hascher, T, van der Veen, I., & Roede, E. (2005). Relations between adolescents’ self-evaluations, time perspectives, motivation for school and their achievement in different countries and at different ages. European Journal of
Psychology of Education, 20, 209-225.
Pekrun, R. (2006). The control-value theory of achievement emotions: assumptions, corollaries, and implications for educational research and practice. Educational
22 Pekrun, R., Goetz, T., Titz, W., & Perry, R. (2002). Academic emotions in students’
self-regulated learning and achievement: a program of qualitative and quantitative research. Educational Psychologist, 37, 91-105.
Pintrich, P. R. (2000). The role of goal orientation in self-regulated learning. In M. Boekaerts, P. R. Pintrich, & M. Zeidner (Eds.), Handbook of Self-Regulation (pp. 451–502). San Diego, CA, US: Academic Press. doi:10.1016/b978-012109890-2/50043-3
Hayes, A. (2017). Introduction to mediation, moderation, and conditional process analysis: a
regression-based approach (Second edition.). The Guilford Press.
Husman, J., & Shell, D. F. (2008). Beliefs and perceptions about the future: a measurement of future time perspective. Learning and Individual Differences, 18, 166–175.
Roede, E. (1989). Explaining student investment, an investigation of high school students’ retrospective causal accounts of their investment in school. Amsterdam: Universiteit van Amsterdam.
Selye, H. (1975) Confusion and Controversy in the Stress Field. Journal of Human Stress, 1, 37-44.
Shahmohammadi, N. (2011). Students’ coping with stress at high school level particularly at 11th & 12th grade. Procedia – Social and Behavioral Sciences, 30, 395-401.
Shipp, A. J., Edwards, J. R., & Lambert, L. S. (2009). Conceptualization and measurement of temporal focus: The subjective experience of the past, present, and future.
Organizational Behavior & Human Decision Processes, 110, 1–22.
Stouthard, M.E.A, & Peetsma, T.T.D. (1999). Future-time perspective: analysis of a facet-designed questionnaire. European Journal of Psychological Assessment, 15, 99-105. Vogel, S., & Schwabe, L. (2016). Learning and memory under stress: implications for the
