Chapter 3 Study 1: Motivation and Anxiety: A Two-‐Year Study of Change in Medical Students
“Power, time, gravity, love. The forces that really kick ass are allinvisible.”
– David Mitchell, Cloud Atlas
Study 1 was designed to investigate how medical students’ motivation orientationand anxiety levels changed over the course of their two-‐year preclinical undergraduate curriculum, following a high stakes admissions procedure. To address the first two hypotheses of the research, the current study sought to examine the differing baseline levels of motivation and their rates of change between different age groups, sexes, ethnicities, admission schemes, and modes of entry into medical school. The study did so by gaining serial quantitative data from participants through questionnaires investigating their dominant motivation based upon self-‐ determination theory (Deci & Ryan, 1985). The questionnaires asked participants to identify demographic data in line with the study hypotheses.
This study also addressed the third hypothesis of the research by measuringanxiety levels through serial responses to a questionnaire on test anxiety. In addition to the quantitative data gathered from this questionnaire, qualitative data were obtained from free text responses of participants in order to provide additional insight into the different anxieties experienced by medical students. In line with the third hypothesis, this study also scrutinised the relationship between motivation, motivation change, and anxiety.
Previous studies had postulated a continuous change in motivation until a steady state was achieved (Otis et al., 2005; Weiner, 1990). Others had shown that baseline differences in motivation were seen between younger and older cohorts(Kronqvist et al., 2007; Newble & Entwistle, 1986; Perrott et al., 2001). Others still had shown differences between the sexes where female students had more evolved approaches to learning than their male counterparts (Inglehart, 1987; Perrott et al., 2001). Similar differences were once thought to have existed in minority students (Ogbu, 1978), but this has more recently been attributed to a lack of recognition of differing cultural influences such as group orientation, stereotype threat, and self-‐ handicapping (Andriessen, Phalet, & Lens, 2006; McInerney et al., 1997; Trueba, 1988). It was hypothesised that similar results in the baseline data of age,sex, and ethnicity would be seen. However, the current study ventured beyond the existing literature by using serial measurements.
Accordingly, this study addressed several previously unanswered questions: firstly, whether motivation and anxiety in medical students differed at their time of entry to medical school; secondly, whether changes in motivation and anxiety were seen over time; thirdly, if specific differences in these changes existed between different demographic groups; and finally, the relationship between motivation change and anxiety.
The study population were medical students enrolled in the Bachelor of Medicine/Bachelor of Surgery (MBChB) programme at the University of Auckland. The participants were the 194 of the 213 (91.1%) students enrolled in MBChB2 in 2011 who were approached and completed the questionnaires. This cohortwas longitudinally sampled at five time points during the next two academic years. At the conclusion of the study, at the end of the second semester in 2012, 210 of the original 213 students remained enrolled in the programme (an attrition rate of 1.4%).
This second-‐year group was deliberately sampled to reflect a student population that was fledgling in its motivational development, having come through a widely disseminated and established competitive selection procedure (Faculty of Medical and Health Sciences, 2014). The faculty provides attenuated schemes for the prospective medical school cohort to redress ethnic and societal imbalances in the medical school cohort. The first is the Rural Origin Medical Preferential Entry Scheme (ROMPE) which provides an avenue for students from regional or rural New Zealand with a view to potentially redressing workforce shortages in these areas. The Māori and Pacific Admission Scheme (MAPAS) is a second attenuated admission scheme, which resulted from a faculty commitment to the principles of theTe Tiriti o Waitangi (Treaty of Waitangi)16, which seeks to increase the number of Māori and Pacific graduates in the health profession.
Although medical school cohorts rarely reflect established population norms for ethnicity and gender (Poole, 2009), no significant differences existed with respect to ethnicity and gender numbers between this cohort and those year cohorts directly preceding and following.
The participants were sampled on three occasions in 2011 (28 February, 2011, at the beginning of semester one; 19 May, 2011, just prior to semester one examinations; 13 October, 2011, midway through semester two at a time of no particular examination or assignment burden), and on two occasions in 2012 (27 February, 2012, at the beginning of semester one; 12 September, 2012, midway through semester two, at a time where students had beenincreasingly exposed to the clinical environment). These times were chosen specifically to coincide with expected times of higher or lower stress. A sampling time of high stress was deliberately sought to examine the effects of stress on motivation orientation, with the hypothesis that higher levels of anxiety would have detrimental effects on motivation(Mosely et al., 1994). For example, the second sampling time was before semester one examinations; this was postulated to be a time of particularly high stress for the respondents and was chosen accordingly. The questionnaires were initially distributed in paper form for the sampling times in 2011, and an online option was also available for the third sampling time in 2011. For subsequent sampling in 2012, the questionnaires were only delivered online (see Table 11 for response rates). Sampling was identified by a brief presentation given by the author to the class at the beginning of a lecture deliberately chosen to reflect one that would typically have high attendance, such as those on examination tips, recapitulate lectures, or those given by lecturers who consistently received teaching awards for excellence17. Potential participants were strongly advised to complete the questionnaires thoughtfully in their own time so as not to disrupt their learning, or the teaching of the lecturer who followed the author’s brief presentation. This introduction was followed up with two reminders: an email to the entire class through a bulk email message system, and a notice on the students’ shared class portal (CECIL). Both of these reminders were facilitated through third parties (a class representative, and a faculty staff member respectively) so that at no time other than the presentations did the author personallyapproach potential participants to complete the questionnaire.
Once the questionnaires had been delivered to the class—either physically, electronically as a Portable Document Format (PDF) version to their shared class portal, or later with an invitation to a survey website (http://www.surveygizmo.com)—sampling remained open for two weeks to allow time for the participants to complete this in their own time. With physical questionnaires participants were asked to leave these in specifically labelled drop boxes outside their lecture theatre or in a secure box in the Student Centre on the Grafton Campus of the University of Auckland. The author regularly cleared these sites. Once the questionnaire moved to an online version, completed questionnaires were stored electronically and downloaded once the sampling time had closed.
Ethical approval was obtained for this study from the University of Auckland Human Participants Ethics Committee (reference 2011/041) on 22 February, 2011, for the duration of the study (not exceeding three years).
The questionnaires were delivered either as a single, double-‐sided piece of A4 paper, or through a questionnaire on a webpage. Paper versions of the questionnaires can be found in Appendix A.
Academic Motivation Scale.
On five occasions over 2011(MBChB2) and 2012 (MBChB3), participants were invited to complete thecollege version of the Academic Motivation Scale (AMS; Vallerand et al., 1992). Participants answered each question based on a five-‐point Likert scale. The gradations were as follows: 1 = strongly disagree, 2 = disagree, 3 = undecided, 4 agree, 5 = strongly agree. Participants did not have to answer every question, and when the questionnaire moved to online delivery, participants could complete and finalise the questionnaire with an incomplete response.
The wording of several questionswas changed to more closely reflect a medical school curriculum and a New Zealand setting (e.g., “Because with only a high-‐school degree I would not find a high-‐paying job later on” was changed to “Because with NCEA18 I would not find a high-‐ paying job later on”).
At the beginning of the questionnaire, participants were asked to enter their unique University of Auckland student identificationnumber so that their serial results could be tracked over time. If this was not completed in the paper format the responses were not accepted; in the online format, this box had to be completed to allow the entire response to be submitted. On the initial sampling time, participants were also asked to indicate their current age, sex, ethnicity19, and their mode of entry into medical school. If these fields were incomplete, the author obtained this information in an anonymised fashion from the Medical Programme Directorate based on their unique student identification number.
The preamble to each question was the same:“Why I chose to come to medical school…” The purpose of the questionnaire was to determine the motivation orientation of the participants on the self-‐determination theory spectrum as proposed by Deci and Ryan (1985). The questionnaire was made up of 27 questions with four questions corresponding to each of amotivation (e.g., “Honestly, I don’t know; I really feel that I am wasting my time in medical school”), external motivation (“In order to obtain a more prestigious job later on”), introjected motivation (“To prove to myself that I am capableof completing my medical degree”), and identified motivation (“Because I think that auniversity education will help me better prepare for the career I have chosen”). Eleven questions corresponded to intrinsic motivation (“Because I experience pleasure and satisfaction while learning new things”). Reliability coefficients are noted in Table 10.
Exploratory factor analysis.
The original description of this scale by Vallerand and colleagues suggested satisfactory levels of internal consistency (mean alpha value = .81), and stability (re-‐test correlation = .79). A seven-‐factor structure was proposed with sub-‐scales for amotivation, external regulation, introjected regulation, identified regulation, intrinsic motivation to know, intrinsic motivation towards accomplishments,and intrinsic motivation to experience stimulation. This has more recently beenreinvestigated for construct validity (Fairchild, Horst, Finney, & Barron, 2005) and the original structure has again been supported. A priori predictions of relationships between scores were also supported. In line with the view that self-‐determination theory was not so much continuous as contiguous, a specific finding was highlighted in Fairchild and colleagues’ work which showed a lack of support for an inter-‐ subscale simplex pattern, which the authors reflected as “a limitation in its [the AMS] theoretical foundations [of a motivation continuum]” (p. 354).
Intrinsic motivation to know describes a pleasure or satisfaction while learning or trying to understand something new; intrinsic motivation towards accomplishments describes the pleasure or satisfaction when one attempts to accomplish or create something; and intrinsic motivation to experience stimulation describes the sensation from the act of engagement in an activity (e.g., flow).
The AMS was reinvestigated for the current study using exploratoryfactor analysis. However, it should be noted that there are fewer responses than the 300 that would be typically required for a robust factor analysis; correlation coefficients fluctuate more so in small samples than in large ones (Field, 2009). A principal components analysis was used because the primary purpose was to identify and compute composite scores for the factors underlying theAMS. A correlation matrix was analysed for each question in the AMS to ensure that cases of major multicollinearity and singularity were not present. The determinant of theR-‐matrix was 1.73 × 10−6. The Kaiser-‐Meyer-‐Olkin measure was .665 (above the commonly recommended value of .60). Bartlett’s test of sphericity was significant (χ2 (351) = 1050.15, p < .001); this was a measure of the null hypothesis and a significant result indicated that theR-‐matrix was not an identity matrix. Given these overall indicators, factor analysis was deemed to be suitable with all 27 items. No items were removed. The average of the communalities was .547, which fell outside of the Kaiser criterion and justified the use of the scree plot to determine the number of factors (Field, 2009).
A scree plot showed a leveling off of eigenvalues after five factors. There were eight factors with eigenvalues of greater than 1.0. However a five-‐factor structure was preferred because of the aforementioned leveling off on the scree plot,an insufficient number of primary loadings for the subsequent three factors, and a difficultytheoretically interpreting the sixth, seventh, and eighth factors. Initial eigenvalues indicated that the first five factors explained 20%,
14%, 8%, 7%, and 6% of the variance respectively, fora cumulative explanation of variance of 55%.
An oblique rotation (δ = 0) was used to optimise the factor structure which allowed the relative importance of the five factors to be equalised. An oblique rotation was used because of the belief that factors should be correlated to one another (Field, 2009). The pattern matrix is shown in Table 2.
Table of Contents
List of Tables
List of Figures
List of Appendices
Chapter 1: INTRODUCTION
The University of Auckland Medical School Learning Environment
Purpose of the Research
Significance of the Research
Design of the Research
Chapter 2: LITERATURE REVIEW
Overview of Motivational Theories and Research
Early Measures of Motivation
Motivation Towards Future Events
Locus of Control
The Role of the Learning Environment
Self-‐Efficacy and Self-‐Concept
Motivation Theory as Applied to Medical Education
Selection of Medical Students
Curriculum Delivery Methods
The Learning Environment
Qualitative Research in Motivation Including Medical Education
Anxiety and Motivation Theory
Levels of Stress and Anxiety in Medical Students
Changes in Stress Across the Medical School Journey
Opportunities for Further Research
Chapter 3: STUDY 1
Motivation Scores Across the Study Participants
Anxiety Scores Across the Study Participants
Free Text Box Responses
Chapter 4: STUDY 2
Results for the Entire Study Participants
Academic Outcome Differences Between Different Demographic Groups
The Relationships Between Motivation,Anxiety,and Academic Outcome:MBChB2
The Relationships Between Motivation,Anxiety,and Academic Outcome:MBChB3
Chapter 5: STUDY 3
Settings for interview
Data collection and analysis
Results and Discussion
Learning of Medicine
Psychosocial Aspects of Medicine
Chapter 6: DISCUSSION
Summary of Results
Unique Findings of the Current Research
Multiple Motivation Orientations in Self-‐Determination Theory
Why are Māori and Pacific Students LessIntrinsically Motivated?
The MAPAS Identity
The Role of Ethnicity in Explanations of Variance in Academic Outcome
The Anxiety Hangover
Unique Relationships of Anxiety to Medical Students
Limitations of the Current Research
Educational Implications of the Current Research and Future Directions for Research
GET THE COMPLETE PROJECT
Motivation Changes in Medical Students During Two Years of the Preclinical Curriculum