Exploring fatigue as a moderator for the test anxiety-test performance relationship

(1)

FATIGUE AS A MODERATOR

Exploring fatigue as a moderator for the test anxiety-test performance relationship Theres B. Ekblad & Eric Nylén Johansson

Örebro Universitet

Abstract

Fatigue and test anxiety are variables that have been identified as prevalent in students of all ages. Test anxiety has been shown to affect test performance. And yet, no studies have considered a possible interaction between these three variables. The aim of this study was to address this relation, specifically, whether fatigue moderates the test anxiety-test performance relationship. This concept was further investigated using a motivational model of fatigue. Self-report questionnaires were answered by students at Swedish universities, using validated scales on sleep, pain, fatigue, and test anxiety. Correlation tests showed that there was an association between test anxiety, test performance, and fatigue, but not motivational fatigue. A moderation analysis showed that fatigue does not moderate the test anxiety-test performance relationship. Thus, we suggest that further research be done with a larger sample, and a controlled environment, instead of being measured through self-reports.

Keywords: fatigue, anxiety, exam, students, motivation

(2)

2

Utforskning av utmattning som en moderator för tentamentsångest och prestation på tentamen Theres B. Ekblad & Eric Nylén Johansson

Örebro Universitet

Sammanfattning

Utmattning och tentamensångest är förekommande hos studenter i alla åldrar. Studier visar även att ångest inför en tentamen påverkar prestationen under tentamen. Ändå har inga studier gjorts där man har utforskat en möjlig interaktion mellan tentamensångest, utmattning och tentamensresultat. Målet med denna studie var att undersöka relationen mellan de tre variablerna, mer specifikt om utmattning modererar relationen mellan tentamensångest och hur man presterar på en tentamen. Konceptet utforskades vidare genom att använda en motivationell modell av utmattning. Enkäter som innehöll validerade skalor på sömn, smärta, utmattning och tentamensångest besvarades av studenter på några svenska universitet. Ett korrelationstest visade samband mellan utmattning, tentamensångest och prestation på tentamen, men motivations-utmattning korrelerade inte. Moderationsanalysen visade att utmattning inte modererar sambandet mellan tentamensångest och prestationen under tentamen. Därför rekommenderar vi vidare forskning med ett större urval och kontrollerad miljö, istället för självrapporter. Nyckelord: utmattning, ångest, tentamen, studenter, motivation

Handledare: Martien Schrooten Psykologi III

(3)

3

Exploring fatigue as a moderator for the test anxiety-test performance relationship Over the years, research has consistently suggested that test anxiety is common amongst students (Ashcraft, 2007; Bitsika, Sharpley, & Bell, 2009; Gürses, Kaya, Doğar, Günes, & Yolcu, 2010). Studies have shown that test anxiety can have negative implications on decision making, working memory and cognitive ability (Ashcraft, 2002; Ashcraft & Krause, 2007; Dutke & Stöber, 2001; Eysenck, Derakshan, Santos & Calvo, 2007; Sommer & Arendasy, 2015). Furthermore, test anxiety has been tied to general anxiety (i.e., anxiety and worry over everyday life events), chronic anxiety (i.e., a term used to describe long-term anxiety), and transitory anxiety (i.e., anxiety or worry over situational factors) (Ackerman & Heggestad, 1997; Hembree, 1988). Thus, test anxiety is thought to negatively impact test performance, and may also be linked to other types of anxiety.

In addition, students often express physical, psychological, and emotional tiredness, due to heavy workloads. This is more commonly known as fatigue, which has also been found to be common amongst students (Beckne, 1995). In a recent article Van Damme, Becker, and Van der Linden proposed that there may be a motivational dimension to fatigue, explaining it as “[---] an aversive motivational state urging disengagement from effortful behaviour of which the costs are currently estimated as exceeding the benefits” (2017, p. 1). If the cost of a certain type of

behaviour is greater than the benefit of the result, interruption of such a behaviour is extremely likely (e.g., if a student is constantly presented with heavy workloads, and still frequently receives poor grades, they are more likely to drop out). Other characteristics in motivation are continual effort, persistence and vigour. This applies no matter the situation, if it’s the primitive need of finding food or a more modern need of learning a new language (Kruglanski & Higgins, 2017). Thus, they propose a model of fatigue, primarily founded by motivational components.

(4)

4

Furthermore, studies have previously found pain and sleep patterns to be strongly associated with fatigue (Menting, Tack, & Knoop, 2017; Luntamo, Sourander, Santalahti, Aromaa, & Helenius, 2012; Kawada, 2017). However, due to their close similarities, it may be important to make certain distinctions. Shen, Barbera and Shapiro noted the similarities between sleepiness and fatigue, writing:

Sleepiness and fatigue often coexist as a consequence of sleep deprivation, and are often grouped together by such patients under the complaint of being ‘tired’. On closer examination, however, it can be seen that sleepiness and fatigue are two distinct, albeit interrelated symptoms (2006, p. 72).

Likewise, researchers have noted an association between fatigue and chronic pain when they are severe (Menting, Tack & Knoop, 2017; Van Der Slot, et al., 2012). Thus, sleepiness, pain and fatigue, while distinctly different phenomenon’s, have been shown to be closely interconnected.

Interestingly, the relationship between test anxiety and test performance has been found to be affected by mental distortion (Putwain, Connors, & Symes, 2010), and by self-regulatory functions (i.e., the ability to control emotions and behaviours; Morosanova & Fomina, 2017). However, very little is known about the things that might influence the test anxiety-test performance relationship. It is important that the relation is explored to better understand and treat these problems.

What’s more, studies have found a link between fatigue, state anxiety (See transitory anxiety; Ardebil, Fekri, Goleyakh, & Moradzadeh, 2010; Bitsika, et al., 2009), and academic performance (Law, 2007; Nagane, 2004; Palmer, et al., 2014). In line with the previously mentioned findings of Ackerman and Heggestad (1997), and Hembree (1988), it may not be entirely far-fetched to assume that fatigue may also be associated with test anxiety. With all of this in consideration, it is rather surprising that we’ve been unable to locate any previous research on the subject.

(5)

5

The aim of this self-report survey was to examine whether fatigue affects the relationship between test anxiety and test performance. Our focus was on the motivational component of fatigue, as proposed by Van Damme, et al. (2017). Therefore, we hypothesize that high levels of both test anxiety and fatigue will have a predictive association with the lowest score on test performance, and low levels of both test anxiety and fatigue will have a predictive association to the highest score on test performance. In addition, we hypothesize that low levels of test anxiety, but high levels of fatigue will have a predictive association to lower scores on test performance than if levels of test anxiety were high, and levels of fatigue were low.

Method Sample

Data was collected from 186 students at two Swedish universities (Örebro University and Skövde University). 137 (73.70%) students reported being female, 47 (25.30%) students reported being male, meanwhile two (1.10%) identified as ‘other’. They were between the ages 19-42 (M: 23.67, SD: 4.10). Inclusion of respondents was based on their comprehension of the Swedish language (measured through an item on Swedish fluency), and if they were students at a

university. The exclusions were made based on whether they could remember the grade on their last exam or not. After accounting for the inclusion and exclusion criteria’s, 163 persons

remained, whereof 119 (73.00%) were female, 43 (26.40%) male, and one (0.60%) who identified as ‘other’. The number of semesters studied for was also addressed (See Table 1). Links and QR-codes that directed the user to the questionnaire page were posted on the Örebro University campus, as well as on official social media groups for Örebro students and Skövde students.

(6)

6

Table 1

Semesters studied by students after exclusions

Semesters Female Male Other Total (N)

n % N % n % n % First semester 32 19.60 10 6.10 0 0 42 25.80 1-3 semesters 37 22.70 16 9.80 1 0.60 54 33.10 4-5 semesters 33 20.20 15 9.20 0 0 48 29.40 6 semesters or more 17 10.40 2 1.20 0 0 19 11.70 Total (N) 119 72.90 47 26.40 1 0.60 163 100

Note. N = Total population; n = Group population

Measures

Demographic. Demographic questions were asked on gender, birth date, semesters

studied without a break, and total semesters studied. The control questions for speaking Swedish fluently, and current student status were also presented here (See Appendix 1).

Test anxiety. Students’ test anxiety was measured using a Swedish version of the Revised

Test Anxiety scale (RTA; Kim, 2014), originally developed by Benson, Moulin-Julian,

Schwarzer, Seipp & El-Zahhar (1992). The RTA is a self-report questionnaire, designed to assess test anxiety in students. Although the Swedish version of the RTA hasn’t previously been tested for validity and/or reliability, it may be worth noting studies have reported good construct

(7)

7

validity and reliability for the original scale (Benson, et al., 1992; Benson & EI-Zahhar, 1994; Hodapp & Benson, 1997; McIlroy, Bunting & Adamson, 2000). In our sample the translated test had a Cronbach's alpha of .91. The instrument consists of 20 items, all of which are statements (e.g., During exams I feel very tense), assessed through a Likert-scale consisting of four ordered responses (1 = almost never, 4 = almost always; See Appendix 1). The RTA is composed of four dimensions; worry, tension, test-irrelevant thinking and bodily symptoms. The possible range of the total scores was 20 to 80, with a cut-off score of 45. The cut-off score was established, using the data provided by Kim (2014), which reported that the norm of student anxiety was at 38-45. We settled for the normative scores being cut-off points, due to previous research reporting test anxiety as common among students (Ashcraft, 2007; Bitsika, et al., 2009; Gürses, Kaya, Doğar, Günes, & Yolcu, 2010). Evidently, a higher score represents a higher level of test anxiety (McIllroy, et al., 2000).

Performance motivation and test performance. Students’ performance motivation and

test performance were assessed through five statements, and two self-report questions. The statements were used to measure the motivational component of fatigue (e.g., I spent a lot of energy on my studies before the exam), and was assessed on a five-point Likert scale (1 = Yes, that is correct, 5 = No, that is incorrect; See Appendix 1). They were developed according to literature on the subject (Boksem, Meijman, & Lorist, 2006; Van Damme, et al., 2017). Specifically, Boksem, et al. (2006) discuss the effort-reward model provided by Tops, Lorist, Wijers & Meijman (2004), wherein it is argued that motivation is sustainable only so long as there is a balance in effort-reward. In short, they argue that an individual will maintain their efforts, so long as the benefit is considered worthwhile. As such, questions aiming to assess to what degree students value their grades were developed. Thus, two dimensions were measured,

(8)

8

namely cost/effort (two items: “I spent a lot of energy on my studies before the exam” and “I spend a lot of energy on trying to understand and answer the questions during the exam”), and reward/motivation (three items: “I felt motivated to study for the exam”, “It was important for me to get a good grade on the exam” and “It was important for me to pass the exam”). Two questions measured the grade that the students had received on their last exam (i.e., “What grade did you get on your last exam?” and “How many correct responses did you get on your last exam?”).

Fatigue. Data on students’ fatigue was collected using the Swedish version of

Multidimensional Fatigue Inventory scale (MFI-20; Fürst & Åhsberg, 2001), originally

developed by Smets, Garssen, Bonke & De Haes (1995). The MFI-20 consists of 20 items, all of which are statements (e.g., Dread doing things), assessed through a five-point Likert-scale (1=Yes, that is correct, 5=No, that is incorrect; See Appendix 1). Studies have reported good construct validity and reliability in the MFI-20 (Chung, Yu, Yung, Yeung, Ng, & Ho, 2014; Ericsson & Mannerkorpi, 2007; Schneider, 1998). In our test the Cronbach's alpha is .92. The MFI-20 measures five dimensions of fatigue; general fatigue (four items: e.g., I feel tired) ⍺ .71, physical fatigue (four items: e.g., I’m physically out of shape) ⍺ 86, reduced activity (four items: e.g., I don’t get much done) ⍺ .83, reduced motivation (four items: e.g., I dread doing things) ⍺ .70, and mental fatigue (four items: e.g., It takes a lot of effort to concentrate) ⍺ .76. The possible range of the total scores was 20 to 100, with a cut-off score of 60 (Tian & Hong, 2013). 10 items were reversed, to accurately measure fatigue (e.g., Physically I feel able to do only a little).

Sleepiness and sleep quality. To better understand fatigue, data was also collected on

sleep. This was measured using two subscales of the Karolinska Sleep Questionnaire (KSQ; Kecklund, & Åkerstedt, 1992): the sleep quality/insomnia index (four statements; e.g.,

(9)

9

difficulties waking up) and the sleepiness index (five statements; e.g., sleepy during work), in addition to an independent question (i.e., Lack of sleep (at least two hours less than main sleep period)) were used. The statements were measured through a six-point Likert-scale (1 = never, 6 = always; See Appendix 1). Good construct validity and reliability has repeatedly been reported for the KSQ and its subscales (Kaida, et al. 2006; Nordin, Åkerstedt, & Nordin, 2013; Åkerstedt, Ingre, Broman, & Kecklund, 2008). The Cronbach’s alpha for both indexes together was α .84. The possible range of the total scores was 0 to 50, with a cut-off score of 30 (Kecklund, & Åkerstedt, 1992).

Chronic pain. Students’ experiences with chronic pain was measured with seven

statements taken from the Short Form-Örebro Musculoskeletal Pain Screening Questionnaire (SF-ÖMPSQ; Linton & Hallden, 1998). The seven items used were ‘How would you rate the pain that you have had during the past three months?’ (with the period being adjusted from past week to past three months to better reflect the purpose of the study), ‘I can do light work for an hour’, ‘How tense or anxious have you felt in the past week’, ‘How much have you been bothered by feeling depressed in the past week?’, ‘In your view, how large is the risk that your current pain may become persistent?’, ‘An increase in pain is an indication that I should stop what I’m doing until the pain decreases’, and ‘I should not do my normal work with my present pain’. These questions were specifically selected under the assumption that they could accurately assess chronic pain that may impact grades. The seven statements that remained (e.g., I can do light work for an hour), were deemed to be enough to measure chronic pain. This on the premise that the scale was developed using questions already controlled for good test-retest reliability, as well as validity (Linton, 1999). The scale was assessed using a 10-point scale (1 = no pain, 10 =

(10)

10

worst pain imaginable; See Appendix 1)1. Good construct validity and reliability has previously been reported (Forsbrand, et al., 2015; Nonclercq & Berquin, 2012; Ruokolainena, Haapea, Linton, Korniloff, Häkkinen, Paananen & Karppinen, 2016). In this study the Cronbach's alpha was .61 for the seven questions that we used. One item was reversed (i.e., I can do light work for an hour). The possible range of total scores was 0 to 63, with a cut-off scores of 32. This is done in a similar fashion to the full scale, with a cut-off score at half of the range of total scores (Linton & Hallden, 1998).

Procedure

Students were recruited to complete the survey anonymously and online via a secure survey system known as Artologik. The survey was advertised using printed information with QR-codes and links to the questionnaire, posted at various locations on Örebro University; campus Örebro and campus USÖ. Information and a link to the questionnaire was also posted on two different pages on Facebook, one group dedicated to the students on Örebro University (“Dom kallar oss studenter”) and one group dedicated to students at Skövde University (“Studenter vid Högskolan i Skövde”) to obtain as many respondents as possible. Groups for other universities were also contacted, however no response was given under the duration that the survey was left open for.

The survey started with a page, containing information on what kind of questions to expect and the ethical considerations taken, specifically how their anonymity was protected, and that they could decline at any time, simply by closing the survey page (See Appendix 1). They were also informed that by sending in their response they agreed to participate. Information on

1_{Due to an error, answers were assessed on a 10-point scale, unlike the original that was assessed through an} 11-point scale. Of course, this would only complicate things if a direct comparison were to be made between the original scale and the adjusted 7-item scale.

(11)

11

the actual purpose of the study was provided at the end of the survey. After information had been provided, participants were presented with questions on demographics. Test anxiety was the first variable to be presented, followed by performance motivation and test performance. This

decision was actively made for two reasons, primarily for ethical reasons, but also on the account that emotionally anxious responses may influence the results on the RTA. The same reasoning was applied to the questions on cost/benefit, being presented before the test performance. Fatigue, sleepiness, and chronic pain were subsequently measured. Fatigue was presented prior to the items on sleepiness and chronic pain, since they served as control variables. The

questionnaire was left open between 15/11/2017 to 01/12/2017. Statistical analyses

All statistical analyses were performed using SPSS 24.0.0.0. Mean, standard deviation and range of total scores was presented for all relevant variables. To control for any association between sleepiness, pain and fatigue, a Pearson correlation was performed. An additional Pearson correlation test was performed on test anxiety, performance motivation, test

performance, fatigue, and motivation index. Variables not significantly correlating with test anxiety and test performance were excluded from further analysis. A single moderation analysis was performed with test anxiety as a predictor, fatigue as a moderator, and test performance as the dependent variable.

Results

Exploring fatigue as a moderator for the test anxiety-test performance relationship The aim of this study was to inspect a possible relation between test anxiety and test performance, as moderated by fatigue. This was further inspected through a motivational model. Due to fatigues close association to sleepiness and chronic pain, they were included as control

(12)

12

variables. Mean, standard deviation, and range of total scores were reported for all variables of interest (See Table 2).

Table 2

Means, standard deviations (SD) and range of obtained total scores for all variables

M SD Range of obtained total scores Total (N) Minimum Maximum Test anxiety 2.06 .55 1.00 3.80 163 Performance motivation 2.01 .66 1.00 4.20 163 Test performance 2.26 .62 1.00 3.00 163 Fatigue 2.86 .74 1.05 4.50 163 Motivation index 3.32 .93 1.00 5.00 163 Sleepiness 2.89 .88 1.20 5.40 163 Chronic pain 4.51 1.40 1.14 7.71 163

Note. N = population; M = Mean; SD = Standard deviation.

To accurately address a predictive relation, and identify a possible moderator,

associations first had to be identified. Using Pearson’s correlation test, an initial analysis was performed on sleepiness, chronic pain, and fatigue (See Table 3). The index was included to be able to measure motivational fatigue, separate from the rest of the indexes.

A second correlation analysis was performed on test anxiety, performance motivation, test performance, fatigue, and the motivation index (See Table 4). There was no significant correlation between test anxiety and performance motivation. However, a strong negative correlation was found between test anxiety and test performance, whereas a strong positive

(13)

13

correlation was found between fatigue and the motivation index. There was no significant correlation between performance motivation and test performance, although there was a strong positive correlation between performance motivation and fatigue, as well as the motivation index. There was a strong negative correlation between test performance and fatigue, however no such significance was found between test performance and the motivation index. Of note is the fact that the items on cost/benefit significantly correlated with both fatigue and the motivation index.

Table 3

Sleepiness, chronic pain, and fatigue; Pearson correlations

Variables Sleepiness Chronic pain Fatigue

Sleepiness -

Chronic pain .541** -

Fatigue .571** .479** -

(14)

14

Table 4

Test anxiety, performance motivation, test performance, fatigue, and motivation index; Pearson correlations

Variables Test anxiety Performance motivation Test performance Fatigue Motivation index Test anxiety - Performance motivation -.086 - Test performance -.340** -.095 - Fatigue .511** .260** -.212** - Motivation index .350** .225** -.134 .753** - Note. *p < .05. **p < .01. ***p < .001.

To measure a possible moderator, a regression analysis was performed, using test anxiety as the independent variable, test performance as the dependent variable, and fatigue as the moderator (See Table 5). The motivation index was not included as a possible moderator as no associations were found (See Table 4). The results showed that the overall model explained 13% of the variations in test performance, with a total R-square of .13, and an adjusted R-square of .11. The analysis indicated that there was a predictive, negative association between test anxiety and test performance. No predictive association was found between fatigue and test performance. For there to be an interaction effect, both the independent variable and the moderator variable must have a predictive association with the dependent variable. However, since the moderator variable, fatigue, did not predict test performance, no interaction effect could be found. Thus, fatigue does not function as a moderator for the test anxiety-test performance relationship.

(15)

15

Table 5

Test performance predicted by test anxiety and fatigue

Predictor Β P 95% CI Fatigue -.041 .570 -.084 .160 Test anxiety -.333 .001*** -.527 -.154 Test anxiety * Fatigue -.054 .232 -.031 .121

Note. *p < .05. **p < .01. ***p < .001. 95% CI = Confidence interval.

To summarize, sleepiness positively correlated with chronic pain and with fatigue, suggesting that the presence of a disturbed sleep pattern is associated with both high levels of chronic pain and high levels of fatigue. Furthermore, chronic pain positively correlated with fatigue. This would suggest that high levels of chronic pain are associated with high levels of fatigue.

Furthermore, high levels of test anxiety predict lower scores on exams, whereas fatigue does not have a predictive relation to test performance.

Discussion

In this study we argued that performance was defined by pre-existing anxiety, which is especially common in students, and that this could be observed in a test anxiety-test performance relationship. We argued that fatigue, also common in students, moderates this relationship. In contrast with previous literature on fatigue (Ardebil, Fekri, Goleyakh, & Moradzadeh, 2010; Beckne, 1995; Bitsika, Sharpley, & Bell, 2009; Kawada, 2017; Law, 2007; Luntamo, Sourander, Santalahti, Aromaa, & Helenius, 2012; Menting, Tack, & Knoop, 2017; Nagane, 2004; Palmer, et al., 2014; Van Der Slot, et al., 2012), this study delved into a cost/benefit balance, more

(16)

16

formally known as motivational fatigue. Although no significant interaction effect was found between test anxiety and fatigue, the associations it has with test anxiety, test motivation, and test performance is evidence enough that an interaction exists. Thus, further research is needed on these variables, and on students’ health. There is at least one plausible explanation for fatigues role in the student environment; rather than fatigue predicting test performance, test performance may instead precede fatigue. That is to say, motivation to study, and how well you perform may have an accumulative effect, predicting fatigue.

Not entirely surprising, an association was found between test anxiety and fatigue, in line with the findings of previous literature (Ackerman & Heggestad, 1997; Ardebil, Fekri, Goleyakh, & Moradzadeh, 2010; Bitsika, Sharpley, & Bell, 2009; Hembree, 1988), tying fatigue to general anxiety, state anxiety, chronic anxiety, and transitory anxiety. It is rather evident then, that fatigue must be associated with anxiety. In addition, test anxiety predicted test performance, in line with the findings of previous literature (Ashcraft, 2002; Ashcraft & Krause, 2007; Dutke & Stöber, 2001; Eysenck, Derakshan, Santos & Calvo, 2007; Sommer & Arendasy, 2015).

Interestingly, the self-developed items on cost/benefit were strongly associated with the fatigue scale, and its motivation index. This supports the writings of Boksem, et al. (2006) and Tops, et al. (2004).

Although the results have some very interesting implications, further research may have to be performed due to the limitations that are present. As noted by Van Damme, et al. (2017); “simple self-report measures will be insufficient to measure fatigue as described” (p. 8). Thus, future studies on the subject may want to take on a different approach, if not the mixed-method approach provided by Van Damme, et al. (2017). Furthermore, test performance can be more accurately measured, using actual reports rather than relying on participant memory. This would

(17)

17

not only increase the accuracy of the results, but also decrease the attrition ratio. Of note is the works of Dutke and Stöber (2001), wherein participants responded to a survey on test anxiety, followed up by an actual test. Moreover, researchers have found a comorbidity between fatigue and depression (Corfield, Martin, & Nyholt, 2016). It has been reported that a person with depression is ten times more likely to suffer from symptoms of fatigue. Thus, future research addressing fatigue may want to control for depression. In addition, future research may want an even larger sample, in order to control for fatigue.

Conversely, strengths can also be reported for our study. Firstly, although some of the scales have been adjusted for this study, it is worth noting that validity and reliability has been established for all but one scale. Furthermore, as previously reported, anonymity functions as a deterrent for socially desirable responses (Mitchell & Jolley, 2010). Thus, it can be assumed that the responses were sincere. More importantly, the sample is arguably representative of the Swedish student population. As reported by Universitetskanslerämbetet (UKÄ, 2017), 60% of the student population is female, and the remaining 40% male, as opposed to our sample of 74% females, and 25% males (before exclusions). Additionally, in line with previous literature, there was a strong positive correlation between all variables, including the motivation index.

To better understand the test anxiety-test performance relationship that is observably prevalent in students (Ashcraft, 2002; Ashcraft & Krause, 2007; Dutke & Stöber, 2001; Eysenck, Derakshan, Santos & Calvo, 2007; Sommer & Arendasy, 2015) it is necessary that interaction effects are explored. That very concept was explored within this study, and although no

interactive relationship was found with fatigue as a moderator, there may still be other variables that moderate the relationship. Furthermore, fatigue shouldn’t be excluded as a variable in its’ entirety. Based on the strong associations that fatigue had with test anxiety, performance

(18)

18

motivation, and test performance, it can be speculated that fatigue may instead be a result of poor grades. The associations we found, support previous findings on the prevalence of test anxiety and fatigue in students, and gives clear implications for their ties, in addition to the ties they have with test performance. This would mean that universities and other educational organizations may have to consider how exams are presented, as well as provide information and physical and mental support to students. Overall, it becomes rather evident that further research is necessary, not only on the relationship between test anxiety, test performance and fatigue, but also on students’ health, and how to tackle these issues.

(19)

19

References

Ackerman, P. L., & Heggestad, E. D. (1997). Intelligence, Personality, and Interests: Evidence for Overlapping Traits. Psychological Bulletin, 121, 219-245. doi:

10.1037/0033-2909.121.2.219

Ardebil, M. D., Fekri, Y., Goleyakh, M. H., & Moradzadeh, M. (2010). P01-129 - A comparative study on the prevalence of fatigue among students with trait-state anxiety in India and Iran. European psychiatry, 25, 337-337. doi: 10.1016/S0924-9338(10)70334-5

Ashcraft, M. H. (2002). Math anxiety: Personal, educational, and cognitive consequences. Current Directions in Psychological Science, 11, 181–185. doi: 10.1111/1467-8721.00196 Ashcraft, M. H., & Krause, J. A. (2007). Working memory, math performance and math anxiety.

Psychological Bulletin & Review, 14, 243–248. doi: 10.3758/BF03194059

Beckne, R. (1995). Studieavbrott i svensk skola : en uppföljningsstudie av en årskull elever från 10-26 års ålder = [Dropouts in the Swedish school] : [a follow-up study of an age group of students from 10-26 years of age]. Stockholm, Sweden: Diss. Stockholm : Univ. Benson, J., & EI-Zahhar, N. (1994). Further refinement and validation of the Revised Test

Anxiety Scale. Structural Equation Modeling, 1(3), 203-221. Retrieved from http://www.tandfonline.com/loi/hsem20

Benson, J., Moulin-Julian, M., Schwarzer, C., Seipp, B., & El-Zahhar, N. (1992). Cross

validation of a revised test anxiety scale using multi-national samples. In K. A. Hagvet and T. B. Johnsen (Eds.), Advances in test anxiety research, 7, 62–83. Lisse, The Netherlands: Swets & Zeitlinger.

Bitsika, V., Sharpley, C. F., & Bell, R. (2009). The contribution of anxiety and depression to fatigue among a sample of Australian university students: Suggestions for university

(20)

20

counsellors. Counselling Psychology Quarterly, 22, 243-255. doi: 10.1080/09515070903216929

Boksem, M. A. S., Meijman, T. F., & Lorist, M. M. (2006). Mental fatigue, motivation and action monitoring. Biological Psychology, 72, 123-132. doi:

10.1016/j.biopsycho.2005.08.007

Chung, K.-F., Yu, B. Y.-M., Yung, K.-P., Yeung, W.-F., Ng, T. H., & Ho, F. Y.-Y. (2014). Assessment of fatigue using the Multidimensional Fatigue Inventory in patients with major depressive disorder. Comprehensive Psychiatry, 55, 1671-1678. doi:

10.1016/j.comppsych.2014.06.006

Corfield, E.C., Martin, N.G, & Nyholt, D.R (2016) Co-occurrence and symptomatology of fatigue and depression. Comprehensive Psychiatry, 71, 1-10. doi:

10.1016/j.comppsych.2016.08.004

Dutke, S., & Stöber, J. (2001). Test anxiety, working memory, and cognitive performance: Supportive effects of sequential demands. Cognition and Emotion, 15, 381–389. doi: 10.1080/0269993004200231

Ericsson, A. & Mannerkorpi, K. (2007). Assessment of fatigue in patients with fibromyalgia and chronic widespread pain. Reliability and validity of the Swedish version of the MFI-20. Disability and Rehabilitation, Vol.29,1665-1670. doi: 10.1080/09638280601055782 Eysenck, M. W., Derakshan, N., Santos, R., & Calvo, M. G. (2007). Anxiety and cognitive

performance: Attentional control theory. Emotion, 7, 336–353. doi: 10.1037/1528-3542.7.2.336

Forsbrand, M., Grahn, B., Hill, J. C., Peterson, I. I., Post Sennehed, C., & Stigmar, K. (2015). THU0631-HPR Validation of the Swedish Version of Start Back Tool Against the Short

(21)

21

Version of the Örebro Musculoskeletal Pain Screening Questionnaire in Patients with Back and/or Neck Pain in Primary Health Care. Annals of the Rheumatic Diseases, 74, 1319. doi: 10.1136/annrheumdis-2015-eular.4622

Fürst, C. & Åhsberg, E. (2001). Dimensions of fatigue during radiotherapy. An application of the Multidimensional Fatigue Inventory. Support Care Cancer, 9, 355-360. doi:

10.1007/s005200100242

Gürses, A., Kaya, Ö., Doğar, C., Günes, K., & Yolcu, H. H. (2010). Measurement of secondary school students’ test-anxiety levels and investigation of their causes. Procedia - Social and Behavioral Sciences, 9, 1005-1008. doi: 10.1016/j.sbspro.2010.12.276

Hembree, H. (1988). Correlates, causes, effects, and treatment of test anxiety. Review of Educational Research, 58, 47-77. doi: 10.3102/00346543058001047

Hodapp, V. & Benson, J. (1997). The multidimensionality of test anxiety: A test of different models, Anxiety, Stress, & Coping, 10, 219-244. doi: 10.1080/10615809708249302 Kaida, K., Takahashi, M., Åkerstedt, T., Nakata, A., Otsuka, Y., Haratani, T., & Fukasawa, K.

(2006). Validation of the Karolinska sleepiness scale against performance and EEG variables. Clinical Neurophysiology, 117, 1574-1581. doi: 10.1016/j.clinph.2006.03.011 Kawada, T. (2017). Poor sleep and fatigue in patients with multiple sclerosis. Journal of the

Neurological Sciences, 373, 143. doi: 10.1016/j.jns.2016.12.047

Kecklund, G., & Åkerstedt, T. (1992). The psychometric properties of the Karolinska Sleep Questionnaire. Journal of Sleep Research, Suppl. 1, 113. Retrieved from

http://onlinelibrary.wiley.com/journal/10.1111/(ISSN)1365-2869

Kim, H.-Y. (2014, March 3). KBT behandling av olika former av prestationsångest. Retrieved from https://www.psykolog.nu/psykoterapi/behandling-av-prestationsangest/

(22)

22

Kruglanski, A. W., & Higgins, E. T. (2017). Introduction to the Motivation Science special issue. Motivation Science, 3, 175-178. doi: 10.1037/mot0000067

Law, D.W. (2007). Exhaustion in university students and the effect of coursework involvement. Journal of American college health, 55, 239-245. doi: 10.3200/JACH.55.4.239-245

Linton, S. J., & Halldén, K. B. A. (1998). Can we screen for problematic back pain? A screening questionnaire for predicting outcome in acute and subacute back pain. The Clinical Journal of Pain, 14(3), 209-215. Retrieved from

http://journals.lww.com/clinicalpain/pages/default.aspx

Linton, S. J. (1999). Örebromanualen för screening av patienter med muskuloskeletala besvär. Tidig identifiering av patienter I riskzonen för kronisk smärta. Retrieved from

http://www.fhvforskning.se/index.php/metoder-verktyg/item/74-orebro-musculoskeletal-pain-screening-questionnaire-ompsq

Luntamo, T., Sourander, A., Santalahti, P., Aromaa, M., & Helenius, H. (2012). Prevalence changes of pain, sleep problems and fatigue among 8-year-old children: Years 1989, 1999, and 2005. Journal of Pediatric Psychology, 37, 307-318. doi: 10.1093/jpepsy/jsr091 McIlroy, D., Bunting, B., & Adamson, G. (2000). An evaluation of the factor structure and

predictive utility of a test anxiety scale with reference to students’ past performance and personality indices. British Journal of Educational Psychology, 70, 17–32. doi:

10.1348/000709900157949

Menting, J., Tack, C.J., & Knoop, H. (2017). Prevalence and correlates of pain in fatigued patients with type 1 diabetes. Journal of Psychosomatic Research, 95, 68-73. doi: 10.1016/j.jpsychores.2017.02.010

(23)

23

Mitchell, M., & Jolley, J. (2010). Research design explained. Belmont: Wadsworth Cengage Learning.

Morosanova, V. I., & Fomina, T. G. (2017). Self-regulation as a mediator in the relationship between anxiety and academic examination performance. Social and behavioral sciences, 237, 1066-1070. doi: 10.1016/j.sbspro.2017.02.156

Nagane, M. (2004). Relationship of subjective chronic fatigue to academic performance. Psychological reports, 95, 48-52. doi: 10.2466/pr0.95.1.48-52

Nonclercq, O. & Berquin, A. (2012). Predicting chronicity in acute back pain: Validation of a French translation of the Örebro Musculoskeletal Pain Screening Questionnaire. Annals of Physical and Rehabilitation Medicine, 55, 263-278. doi: 10.1016/j.rehab.2012.03.002 Nordin, M., Åkerstedt, T., & Nordin, S. (2013). Psychometric evaluation and normative data for

the Karolinska Sleep Questionnaire. Sleep and Biological Rhythms, 11, 216-226. doi: 10.1111/sbr.12024

Palmer, L. K., Economou, P., Cruz, D., Abraham-Cook, S., Huntington, J. S., Maris, M., Makhija, N., Welsh, T., & Maley, L. (2014). The relationship between stress, fatigue, and cognitive functioning. College Student Journal, 48(1), 198-211. Retrieved from

http://www.projectinnovation.com/home.html

Putwain, D. W., Connors, L., & Symes, W. (2010). Do cognitive distortions mediate the test anxiety–examination performance relationship? Educational psychology, 30, 11-26. doi: 10.1080/01443410903328866

Ruokolainena, O., Haapea, M., Linton, S., Korniloff, K., Häkkinen, A., Paananen, M. & Karppinen, J. (2016). Construct validity and reliability of Finnish version of Örebro

(24)

24

Musculoskeletal Pain Screening Questionnaire. Scandinavian Journal of Pain, 13, 148-153. doi: 10.1016/j.sjpain.2016.06.002

Schneider, R. A. (1998). Reliability and validity of the Multidimensional Fatigue Inventory (MFI-20) and the Rhoten Fatigue Scale among rural cancer outpatients. Cancer Nursing, Vol.21(5), 370-373. Retrieved from

http://journals.lww.com/cancernursingonline/pages/default.aspx

Shen, J., Barbera, J., & Shapiro, C. M. (2006). Distinguishing sleepiness and fatigue: focus on definition and measurement. Sleep Medicine Reviews, 10, 63-76. doi:

10.1016/j.smrv.2005.05.004

Smets, E. M. A., Garssen, B., Bonke, B., De Haes, J. C. J. M. (1995). The Multidimensional Fatigue Inventory (MFI). Psychometric qualities of an instrument to assess fatigue. Journal of Psychosomatic Research, 39(3), 315-325. Retrieved from http://www.jpsychores.com/ Sommer, M., & Arendasy, M. E. (2015). Further evidence for the deficit account of the test

anxiety–test performance relationship from a high-stakes admission testing setting. Intelligence, 53, 72-80. doi: 10.1016/j.intell.2015.08.007

Tian, J., & Hong, J.-S. (2013). Application of the Chinese version of the MFI-20 in detecting the severe fatigue in cancer patients. Supportive Care in Cancer, 21, 2217-2223. doi:

10.1007/s00520-013-1783-x

Tops, M., Lorist, M. M., Wijers, A. A., Meijman, T. F. (2004). To stress or relax: neurochemical aspects of activity and rest. Gedrag en Organisatie, 17, 32–42. Retrieved from

https://www.researchgate.net/publication/286787519_To_stress_or_relax_Neurochemical_ aspects_of_activity_and_rest

(25)

25

Universitetskanslersämbetet. (2017). Universitet och högskolor. Årsrapport 2017. Retrieved from http://www.uka.se/download/18.667441fa15af4fb7b21126b3/1497612948871/arsrapport-2017.pdf

Van Damme, S., Becker, S., & Van Der Linden, D. (2017). Tired of pain? Toward a better understanding of fatigue in chronic pain. PAIN, 0, 4. doi:

10.1097/j.pain.0000000000001054

Van Der Slot, W. M. A., Nieuwenhuijsen, C., Van Den Berg-Emons, R. J. G., Bergen, M. P., Hilberink, S. R., Stam, H. J., & Roebroeck, M. E. (2012). Chronic pain, fatigue, and depressive symptoms in adults with spastic bilateral cerebral palsy. Developmental Medicine & Child Neurology, 54, 836-842. doi: 10.1111/j.1469-8749.2012.04371.x

Åkerstedt, T., Ingre, M., Broman, J. E., & Kecklund, G. (2008). Disturbed sleep in shift workers, day workers, and insomniacs. Chronobiology International, 25, 333–348. doi:

(26)

26

Appendix 1 Information page

(27)

27

(28)

28

(29)

29

(30)

30

(31)

31

(32)

32

(33)

33