Student Test Anxiety and Learning Strategy
Increased anxiety is often felt when individuals encounter life-events that may instigate recurring fears or worries, such as a test environment. Previous studies have found that female students tend to report more test anxiety compared to their male peers. Learning strategies such as rehearsal help learners acquire knowledge at a surface level by retaining information through repetition. The purpose of this study was to explore gender differences in self-reported levels of test anxiety and to investigate whether these differences and reported rates of test anxiety are related to the reported use of rehearsal as a learning strategy. We hypothesised students who report higher levels of test anxiety also report higher levels of rehearsal-strategy and that female students report higher levels of test anxiety compared to male students. Two thousand, and eighty four participants responded to the Motivated Strategies for Learning Questionnaire. 1269 (67.4%) were women and 613 (32.6%) were men. Participants ages ranged from 16 to 89 years with a mean of 21.28 years (SD=1.53). Participants who reported higher levels of test anxiety also tended to report higher levels of rehearsal as a learning strategy and females tended to report higher levels of test anxiety compared to males. Comparatively high incidences of anxiety and stress have been reported in a global sample including students during the COVID-19 pandemic. Managing student’s levels of test anxiety should be a key aim for university policy that integrates measures to mitigate hazards to mental health including education on appropriate learning strategies.
One in four people will experience a mental health concern in their lives; suggesting that many students are likely to be affected. Mental illness is argued to top the list globally of primary reasons for burden of disease. It is therefore of utmost importance that academic institutions encourage methods of study that are likely to limit symptoms of anxiety and other mental health disorders amongst student populations, as well as providing appropriate support.
Anxiety has been long studied in comprehending the function of emotion related to performance, usually noted by feelings of tension, worried thought patterns, and associated physiological manifestations, such as high blood pressure (APA, 2013). Increased anxiety is often felt when individuals encounter life-events that may stimulate recurring fears or worries, such as a test environment. Test anxiety is often due to fears that are triggered by potential critical evaluation resulting in knock on negative behavioural, emotional or physical responses (Zeidner, 1998). Previous studies have found that female students tend to report more test anxiety compared to males (Núñez-Peña et al., 2016). Biological, psychological, and environ-mental variables affect how individuals respond in a test situation exhibiting varying levels of anxiety. These include differing styles of learning strategy (e.g., self-efficacy, motivation and self-regulation) and also gender (Schnell et al., 2015).
Learning strategies such as deep-processing and monitoring allow individuals to modulate their own study, knowledge acquisition and academic achievement (Flavell, 1981; Paris & Lindauer, 1982). Students who have modified their processes meta-cognitively, motivationally, and behaviourally to be more involved in their learning by adopting such strategies are more likely to achieve success (Camahalan, 2006). Failure to uptake these approaches can mean diminished achievement and/or heightened anxiety levels through the testing process (Fulk and Brigham, 1998).
Self-regulation is noted as “an active, constructive process whereby learners set goals for their learning and then attempt to monitor, regulate, and control their cognition, motivation, and behaviour, guided and constrained by their goals and the contextual features in the environment” (Pintrich, 2000, p. 453). Learning strategies such as rehearsal help learners assume knowledge at a more surface level by remembering subject matter through repetition such as a learner who listens to an on-line lecture multiple times or repeatedly reviews lecture notes and lists to memorise key terms and theories.
There is growing support in literature connecting specific learning strategies with assessment-related anxiety (Sotardi & Brogt, 2020) and researchers have shown rehearsal to be positively related with anxiety compared to other strategies (Credé & Phillips, 2011). Interestingly, studies exploring correlation between rehearsal and educational achievement have not found a strong significant relationship (Puzziferro, 2008). Meta-analysis of three papers did not find a significant association between rehearsal and achievement. Since this strategy has not been found to be linked with improved academic achievement, it is possible that students adopting this strategy experience a higher level of test anxiety due to preceding academic results and a potential last minute preparatory approach. Bidjerano (2005) found that female students tend to more frequently use rehearsal learning strategy compared to males. Rehearsal as a learning strategy is thought to be a superficial strategy that does not provide robust learning (Pintrich, 2000).
Since female students experience a higher level of test anxiety, we were motivated to examine any correlation between test anxiety and rehearsal as a self-regulated learning strategy. Studies examining gender differences in the wider discourse of self-regulated learning report conflicting findings with some studies concluding no significant disparities while others show such divide (Bidjerano, 2005; Susilowati et al., 2020). Based on these points, the purpose of this study is to explore gender differences in self-reported levels of test anxiety amongst study participants. Furthermore, our research aims to investigate whether rates of test anxiety are related to the reported use of rehearsal as a learning strategy. We hypothesise those who report higher levels of test anxiety also report higher levels of rehearsal-strategy and that female participants report higher levels of test anxiety compared to males.
Two thousand, and eighty four participants voluntarily participated in the study. Opportunity/convenience sampling was used primarily with advertising on social media. The percentage of non-binary participants tends to be small, with approximately 3% of adolescents identifying as non-binary (Cameron & Stinson, 2019). Therefore we only compared data of female and male participants in this study. This decision was made after identifying the lack of literature on non-binary people regards to our research topic and the assumption that the collected sample size would not allow for a meaningful comparison with the inclusion of those of non-binary gender (Cameron & Stinson, 2019). Only data of participants reported as over the age of 16, who fully answered all questions were included to remove erroneous data and in accordance with the British Psychological Society Code of Ethics and Conduct and APA Code of Ethics. After exclusions, there were a total of 1882 participants, 1269 (67.4%) were women and 613 (32.6%) were men. Participants’ ages ranged from 16 to 89 years with a mean of 21.28 years (SD=1.53).
The data for our research was collected using a pre-existing Motivated Strategies for Learning Questionnaire (MSLQ) (Pintrich et al., 1991) and Experimentum software (DeBruine, 2019). The MSLQ uses a 7-point Likert scale and participants score each item by choosing 1-7, representing strongly disagree to strongly agree (Pintrich et al., 1991). Participants were asked to answer 81 questions, but our analysis only focussed on five items to measure test anxiety and four items for rehearsal strategy. An example of an MSLQ question on test anxiety is: ‘When I take a test I think about how poorly I am doing compared with other students.’ An example of an MSLQ question on rehearsal strategy is: ‘When studying for this course, I read my class notes and the course readings over and over again.’ Additionally, participants provided demographic information including gender, age, year of study, employment status and accommodation status.
Participants were invited to take part via a social media advert, they clicked on the provided link, provided consent by reading participant information and using an on-line check box. Questions were then answered and the questionnaire submitted on-line. Demographic questions were answered firstly followed by the MSLQ, for which questions were randomised for each participant.
Design and data analysis
The dependent variable measured are the means of test anxiety and rehearsal strategy sub-scale scores from the MSLQ in male and female students. Data were expected to meet the assumptions for a monotonic relationship, with a positive correlation between dependent variables. Based on distribution of data, Spearman’s correlational analysis was performed.
The independent variable is the gender of the student reported as male or female. Data were screened before analyses. For our between-subjects design we ran a two-sided, independent t-test to examine whether there is a difference between anxiety scores of male and female students. All data wrangling, visualisation and analysis were conducted using the R programming environment (Version 4.0.2, R Core Team, 2020). Data were prepared utilizing tidyverse library (Version 1.3.0) (Wickham et al., 2019), broom library (Version 0.7.0) (Couch et al., 2020), car library (Fox & Weisberg, 2019), lsr library (Version 0.5) (Navarro, 2015), pwr library (Version 1.3-0) (Champely, 2020) and psych library (Version 2.0.7) (Revelle, 2020).
Means, medians, standard deviations, min/max and skew of our correlation coefficients are reported in Table 1. The MSLQ data was treated as interval data, it was determined that variables were independent. A histogram and confidence envelope qq plot visualisations were used to test for normal distribution of anxiety and rehearsal scores. Normal distribution could not be confirmed for anxiety test scores nor rehearsal scores based on visualisations, therefore a Shapiro Wilk test was used to confirm residuals were not distributed normally, w=0.98586, p<.01. A scatter-plot was used to confirm a monotonic relationship with a small, positive correlation between the anxiety scores and rehearsal scores noted (Garson, 2012).
A Spearman’s rank correlation was conducted since normal distribution was not confirmed. We found a small, significant, positive correlation between test anxiety (M = 4.61, SD = 1.41) and rehearsal strategy (M = 4.30 , SD = 1.28, r (828604847) = .25, p = >.01. The results confirm the hypothesis that participants who report higher levels of test anxiety also tend to report higher levels of rehearsal as a learning strategy.
For our between-subjects design, data was examined to determine whether there is a difference between anxiety scores for male and female participants. The large sample size data was considered to be interval and the variables independent and therefore meets the assumptions for parametric statistics . To assess for homogeneity of variance and if the assumption of normal distribution of residuals was met , data visualisation was used in the form of a violin-box plot. Outliers were not noted.
A one-sided independent-samples Welch’s t-test found that females (M = 4.83, SD = 1.36) had significantly higher anxiety scores than male participants (M = 4.14, SD = 1.41, t(1171.86) = 10.16, p < .01). The calculated effect size for our t-test was 0.6, therefore a medium significant difference in anxiety scores between females and males was noted, with females tending to report higher test anxiety scores. A one-sided t-test was performed in line with our hypothesis where females were expected to report higher test anxiety scores based on previous literature (this differs from pre-registration where a two-sided t-test was noted in error). Our results confirm the hypothesis that female participants tended to report higher levels of test anxiety compared to male participants.
The results confirm our hypotheses that: 1. participants who report higher levels of test anxiety also tend to report higher levels of rehearsal as a learning strategy and 2. female participants tend to report higher levels of test anxiety compared to male participants. Only a small positive correlation was found between test anxiety and rehearsal strategy. Further analyses utilizing parametric statistical tests based on the large sample size available could be conducted to examine the relationship between the variables. The rehearsal scale examines how much students rely on memorising class materials and is considered to be a very basic learning strategy (Anthony et al., 2013). It is usually used to integrate new information into short-term memory, however does not lead to building cognitive connections between new information and existing knowledge like other strategies, e.g. meta-cognition. It therefore, may not be surprising that students scoring highly on rehearsal, potentially relying predominantly on this strategy, may suffer more from test-anxiety.
Our results demonstrate a medium effect size of higher test-anxiety scores in female students compared to males. Students exhibiting higher levels of test-related anxiety benefit most by adopting self-regulatory processes to aid achievement. Since females tend to report higher levels of test anxiety, gender-sensitive interventions may be useful as a means of reducing test anxiety in females as well as males. Future studies are necessary to determine if females tend to adopt rehearsal as a learning strategy compared to males. Nonetheless, due to the growing necessity for on-line learning, there is a critical demand to comprehend how students can best adopt learning strategies for on-line academic proficiency (Broadbent and Poon, 2015). Self-regulated learning strategies like meta-cognition and critical thinking have previously been found to be positively correlated with higher levels of attainment in on-line settings, however with smaller effect sizes than results from prior in-person learning. Conversely, rehearsal and other strategies such as organisation and elaboration were found to be less effective on-line learning strategies. Face to face peer learning has been dramatically compromised in 2020 which may also account for higher levels of anxiety as well as stressful world events. It is important to investigate the relationship between cognitive performance and face to face interaction since neuronal health has been previously found to be related to social interaction (Fuchs, 2014). Comparatively high incidences of anxiety and stress were reported in a global sample including students during the COVID-19 pandemic (Xiong et al., 2020).
In education, literature suggests from 15% to 22% of students exhibit high test anxiety (Thomas et al., 2017). Research has shown that a relationship exists between test anxiety and increased risk for future anxiety and depression (Leadbeater et al., 2012). Since brain development undergoes critical changes throughout childhood and teenage years with the adolescent brain uniquely defined by adaptability until the 20s (Giedd, 2015), mitigating the impact of anxiety-inducing circumstances is of key importance with widespread public health and societal implications.
There is growing evidence that early life events can affect long term mental health in humans by influencing regulation of the genome by epigenetic pattern and therefore genetic output leading to the genesis of personality traits and disorders. (http://www.frontiersin.org/people/u/535686 et al., 2018) This would suggest that personality and precursors for personality disorders can be shaped by early life events including stressful experiences. Adverse childhood experiences have long-term effects on development, including increased probability for the development of personality traits, disorders and psychopathology. Research examining the biologic origins of these associations has revealed adverse childhood experiences can lead to abnormalities within the stress response system and function of the hypothalamic pituitary adrenal (HPA) axis in both children and adults (https://www.ncbi.nlm.nih.gov/pubmed/?term=Parade%20SH%5bAuthor%5d&cauthor=true&cauthor_uid=27691985 et al., 2016). It is therefore, critically important that test-environments, especially for children and adolescents, do not act in a compounding manner to the current mental health epidemic.
Associated risk factors for increased levels of stress and anxiety in a recent study included female gender, age less than 40 years, psychiatric illnesses, unemployment, student status, and regular exposure to social media/news concerning COVID-19 (Xiong et al., 2020). Students should be advised of the strategies that will aid their learning and achievement, with mindful integration of supportive measures to help them cope with difficulties they may face throughout their academic tenure. Introducing early intervention programmes for mental illness can help reduce the impact of ill mental health on individuals, businesses, the economy and wider society.
Although our hypotheses were supported by our results, there are some limitations with our study design. Firstly, correlation does not equal causation therefore caution must be applied in translating results from this study to wider populations. Some researchers have proposed that less capable students experience high test anxiety (Sommer and Arendasy, 2014). This could in part account for the relationship between test anxiety and the ‘shallow’ effectiveness of utilizing rehearsal as a learning strategy. Further, only 32.6% of our sample participants were male, therefore further studies are necessary with a more equal distribution of males and females. Our correlational analysis could be re-run utilizing parametric testing due to large sample size available (Lehman, 2012). In addition, previous studies have called for the need to improve the validity of the MSLQ itself, as a valid measure of self-regulated learning strategies (Anthony et al., 2013). The timing that the MSLQ was delivered for our study, close to the start of term could account for some misrepresentations in the data, since students had not yet experienced exams or full course evaluations.
Due to unpredicted societal changes encountered during 2020, the subject of test-anxiety holds more importance than ever before. The COVID-19 pandemic has brought upon an unprecedented threat to mental health across the globe and priority must be given to the prevention of mental disorders (e.g. anxiety, PTSD, and suicide). Managing student’s levels of test anxiety should be a key aim for university policy that integrates measures to mitigate hazards to mental health through learning support, student education on most helpful learning strategies, appropriate course design and university policy that carefully allows for social interaction, gender-sensitive approaches may be helpful. Further research on self-directed learning techniques and in particular the importance of peer to peer learning is required to help develop such strategies.