Master’s Thesis BA Strategic Innovation Management
Creating Physician-Scientists: The Application of the
Theory of Planned Behavior and Goal Importance
Author:
Daindra W. Utami
Student number: S2934493
Study Program: MSc BA Strategic Innovation Management
Email: d.w.utami@student.rug.nl
Thesis Supervisor: prof. dr. ir. J.M.L. (Jo) van Engelen
Co-Assessor: dr. T.L.J. (Thijs) Broekhuizen
Word count: 12,350
Faculty of Economics and Business University of Groningen
Duisenberg Building, Nettelbosje 2, 9747 AE Groningen, The Netherlands P.O. Box 800, 9700 AV Groningen, The Netherlands
2 Abstract
Background: The number of physician-scientists (doctors who also conduct scientific research on top of their daily clinical practice) all over the world are on the threshold of long-term decline. However, they play an essential role in sustaining the innovativeness of the healthcare sector. Given their critical role, a strategy to counteract their decline needs to be found. This study looks into the psychological theories of behavior, using the Theory of Planned Behavior (TPB) and goal importance, to further study medical students’ motivation in conducting scientific research activities.
Methods: The full study is set up with theoretical research (to build a conceptual foundation) and a mixed method data collection. The mixed method data collection utilizes a questionnaire with open and closed-questions and subsequent interviews. In designing the questionnaire, this study also looks into other potential factors that may influence medical students’ motivation in participating in scientific research activities. This thesis focuses on reporting theoretical research and results of the questionnaire study.
Results: This investigation found significant support for goal importance to predict students’ decision. It is also found that students who participated in extracurricular scientific research activities have a higher motivation to undertake scientific research. This study found no support for the suggestion that previous research experience influences medical students’ motivation to do scientific research. Information sufficiency was found to be an important influencing factor in students’ motivation to participate in scientific research activities. 67.2% students who decided not to participate in the extracurricular research program perceived that they did not have sufficient information about the program. The subsequent interview study will focus on the role of information sufficiency towards medical students’ motivation, but it is beyond the scope of this thesis.
Conclusions: Students will be more likely to participate in scientific research activities if they perceive that it is important to do so. Students’ perception whether they have sufficient information (about extracurricular scientific research activities) give substantial influence to how important it is for them to participate in it.
3 TABLE OF CONTENTS Abstract 2 Table of Contents 3 Acknowledgments 5 Introduction 6
CHAPTER 1: Literature Review 9
1.1 Theory of Planned Behavior 9
1.2 Goal Importance 10
1.3 The Rationale behind the Present Study 11
1.3.1 Other Potential Factors 12
1.3.1a Previous Research Experience 12
1.3.1b Information Sufficiency 12
CHAPTER 2: Research Methods 14
2.1 Research Context 14
2.2 Research Design 14
2.3 Sampling Procedure 15
2.4 Research Instruments 16
2.4.1 Questionnaire & Open Questions 16
2.4.2 Interviews 16
2.5 Study Procedures 17
2.5.1 Questionnaire Procedure 17
2.5.2 Interview Procedure 18
2.6 Measures Used in the Questionnaire 18
2.6.1 Theory of Planned Behavior (TPB) 18
2.6.2 Goal Importance 19
2.6.3 Open Questions 19
2.7 Ethicality of Data Collection and Handling 19
2.8 Data Analysis: Anticipating and Resolving Potential Issues 20
2.8.1 Closed Questions Data Analysis 20
4
CHAPTER 3: Questionnaire Results and Analysis 23
3.1 Respondents Profile 23
3.2 Closed Questions Results 23
3.2.1 Descriptive Analysis 23
3.2.2 Binary Logistic Regression 26
3.2.3 Robustness Check of the TPB 27
3.2.4 MANOVA Analysis 28
3.2.5 Mann-Whitney U Test 30
3.3 Open-questions answers 31
CHAPTER 4: Discussion and Conclusion 35
4.1 Discussion 35
4.2 Conclusion 37
4.3 Managerial (Educational) Implications 38
4.4 Implications for The Subsequent Study 38
4.5 Limitations & Suggestions for Future Research 39
References 41
Appendices 49
Appendix 1 - Questionnaire that has been modified to improve student understanding 49
Appendix 2 - Respondents’ Profile based on Nationalities 51
Appendix 3 - Criteria Fulfillment Test for MANOVA 51
Appendix 4 - Words used for coding 56
5 Acknowledgments
First of all, I would like to extend my profound gratitude to my supervisors who have been involved in the study, prof. dr. ir. Jo van Engelen, Dr. Johanna Schönrock-Adema, and Dr. Joke van der Mark-van der Wouden, for their valuable and constructive suggestions during the planning and the execution of this research. Without their patient guidance, encouragements, and useful feedbacks, it was impossible to accomplish this thesis.
I also would like to acknowledge dr. Thijs Broekhuizen from the Faculty of Economics and Business at the University of Groningen as the co-assessor of this thesis. I am grateful for his very valuable comments and suggestions for the final version of this thesis.
And finally, I would like to use this opportunity to express my sincere gratitude for my parents, Daru Wisaksono and Christina Indrarini, my sister, Anindra Widyaswati; my brother, Ibrahim Wisaksono; and Sylvester Veenstra for their continuous support throughout my study. Without them, I would not be the person I am today and I would not be where I am now.
I, once again, extend a warm thank you to everyone mentioned.
6 Introduction
Social Relevance of The Thesis: What is the problem and its effect on society?
scientists are essential links in the chain of medical research and practice. Physician-scientists are doctors who, on top of their clinical practice, also perform scientific research and contribute to medical sciences. They are uniquely positioned to conduct high-quality and relevant medical research by combining fluency in medical knowledge and privileged access to patient care, which allows them to communicate and collaborate in clinical research with scientists on one side, and with healthcare providers on the other side (Rosenberg, 1999). According to Schön (1983), this position—which allows them to get the best understanding of both scientific (from doing research) and practical (from facing daily problems in medical practice) worlds—can give them the opportunity to conduct much better targeted and relevant medical research.
Another factor is that of innovation in the healthcare sector. The position places the role of physician-scientists as crucial contributors to the development of new healthcare solutions that will improve future patient care, enhance the medical education system, and increase the prosperity of biomedical knowledge. The position also allows them to ensure that new healthcare solutions and innovations can be effectively assimilated in the market (Archer, 2006).
Wyngaarden has brought attention to the importance of the phenomenon of the diminishing number of physician-scientists, and he even goes as far as depicting them as an endangered species (Wyngaarden, 1979). However, up until today, the number of physician-scientists still falls short from what is ideal (Davila, 2016). In the United States, the National Institute of Health (NIH) has reported that although in the last decade the number of physicians is growing, the number of physician-scientists remains stagnant (Rockey, 2014). It shows that the proportion of physician-scientists in the workforce is decreasing and it has been argued that it can eventually lead to a decrease in their numbers (Milewicz et al., 2015). In addition to that, Arab countries also have reported alarming conditions in the development of biomedical knowledge as a result of lack of physician-scientists (Alfaar et al., 2016). Therefore, this thesis is written to address the following problem (societal problem statement): Given the critical role of physician-scientists, it is strategically important for the innovativeness of the healthcare sector to find out a way to counteract the diminishing number of physician-scientists throughout the world.
7
crucial for a robust strategy to increase students’ motivation and promote successful efforts to counteract the diminishing number of physician-scientists throughout the world.
The Netherlands has historically been successful in attracting medical students to do scientific research. A study by Eyk et al. (2013) has shown that performance of Dutch medical students in doing scientific research is satisfactory, with 14.5% of students publishing a research paper during the first three years of their medical study with above average number of citations (> 1.28). This shows that there might be compelling aspects present that motivate students to do scientific research in the Netherlands. Therefore, this Dutch context might be a useful setting to study how to strengthen the scientific research interest of medical students and motivate them to pursue such interests. This thesis addresses the main research question: “How can we effectively increase the medical students’ motivation in doing scientific research activities?”.
Scientific Relevance of The Thesis: Knowledge Gap & Research Questions
Motivation can be defined as “the force that powers people to overcome barriers and achieve a high level of performance” (Larson & Gray, 2015). It strengthens, guides, enhances and maintains a person’s behavior. In general, there are two types of motivation; intrinsic motivation and extrinsic motivation. Intrinsic motivation refers to motivation that is driven by an interest or enjoyment in doing the task itself and exists within the individual. Meanwhile, extrinsic motivation is driven by external factors from outside of the individual (Laming, 2004; Tohidi & Jabbari, 2012).
8
The theory of planned behavior (TPB) explains how human behavior and their decision outcomes are influenced by their attitudes, subjective norms, and perceived behavioral control. The focal argument of this theory is that an individual’s motivation is influenced by factors that are inherent within themselves. While the success of the utilization of the TPB has been proven by many studies in different settings (e.g., Dawson et al., 2015; Jimmieson et al., 2008; Kyle et al., 2014; Murphy, 2014; Sideridis, 2005), it has been argued that the TPB only reflects on reasoned action models and does not account for motivational factors that are related to an individual’s aspirations, goals, or priorities (Orbell et al., 2001; Perugini, 2004).
Ajzen (1991) has stated that the TPB is, in principle, open to new predictors that account for unique variances in human behavior. Therefore, to understand the factors that underlie medical students’ motivation, this study will incorporate a predictor that reflects students’ desire and ambition, and their assessment of the perceived importance of achieving a particular goal. Sideridis (2005) has proposed the inclusion of goal importance as an expansion of the TPB to improve its predictive validity to the prediction of students’ behavior.
Given the relevance of TPB and goal importance in predicting and understanding students’ behavior, motivation, and decision outcomes, this study utilized both constructs as a predictor of medical students’ decision to participate (or not participate) in extracurricular scientific research activities. Therefore, this study draws two further sub-research questions; “How do students who participated in and did not participate in the extracurricular scientific research activities differ with regards to their TPB and goal importance?” and “Do positive TPB and high goal importance translate into a more likely positive intention to participate in extracurricular scientific research activities?”.
9 CHAPTER 1: Literature Review
1.1 Theory of Planned Behavior
The Theory of Planned Behavior (TPB) (Ajzen, 1991) proposes that attitudes, subjective norms, and behavioral controls, which are based on an individual’s beliefs, are the most important determinants of behavioral outcomes. The theory is deemed to represent the intrinsic motivations of an individual based on its focus on an individual’s inherent characteristics and values as predictors of behaviors. It originated as an extension of the Theory of Reasoned Action (TRA) (Fishbein & Ajzen, 1980) which was also developed to study the influence of an individual’s mindset and social circle to their decision-making and behavior.
The predictors from TPB are made up of three kinds of belief. The first TPB predictor is the behavioral belief, and it concerns the likely consequences of performing a behavior. It is postulated to be the underlying influence of an individual’s evaluative reaction to produce a favorable or unfavorable attitude towards an object or behavior (Gardner, 1985). The second TPB predictor is the normative belief, which consists of two kinds of normative values. The first normative value concerns the perceived normative expectations within a social circle, which results in perceived social pressure or subjective norms. It posits that an individual is more likely to perform a behavior when people around them are performing the same behavior, or an individual’s social environment encourages them to do so. The second normative value concerns the social support that an individual receives under difficult circumstances. It posits that an individual is more likely to perform a behavior when they assume that they will receive social support from their family, partners, and closest friends during difficult times. The third TPB predictor is the control belief, which concerns a belief in one's ability to control factors that help master the performance of the behavior and to contain factors that hinder the performance of the behavior. It gives an individual a perceived ease or difficulty in performing a behavior, which is used to evaluate their competence to control and accomplish it.
10
correlations between TPB predictors and behavior ranging from 40% to 66% (Armitage & Conner, 2001; McEachan et al., 2011; Schulze & Wittmann, 2003). The third reason for adopting the TPB is due to the availability of well-developed and validated tools and parameters to study motivations specifically in student settings. This will also allow comparisons of the findings of this research with findings of other research, which will be useful to identify further important aspects that can contribute to students’ motivation to participate in extracurricular scientific research activities.
1.2 Goal Importance
Despite the presence of many studies that support positive results for the TPB as a predictor of human behavior, the sufficiency of using the theory alone to predict behavior is debatable. This is because there is a proportion of variance that influences an individual’s decision that is still unaccounted for by the TPB (Eagly & Chaiken, 1993; Kyle et al., 2014). Ajzen (2011a; 1991) has acknowledged this and concludes that the TPB is, in principle, open to the inclusion of additional predictors if they can capture a significant proportion of variance in decision-making or behavior, after all variables from the TPB have been taken into account.
One criticism of the TPB is that it merely reflects a reasoned decision-making process and does not account for motivational factors that reinforce a behavior (Perugini, 2004; Sideridis, 2005). In other words, even if an individual possesses a positive attitude towards performing a certain behavior, perceives social pressure to perform that behavior, perceives they will receive needed social support, and believes that the performance of the behavior can be easily performed or controlled, it does not mean that they will consider the behavior important to perform (Orbell et al., 2001).
Abraham and Sheeran (1993) have pointed out that the predictive validity of the TPB can be enhanced by including a construct related to goals. Having clear goals is associated with a more positive motivation to perform a behavior relevant to achieving those goals. This is because they represent outcomes that an individual aspires to achieve (Ford, 1992). In addition to that, formulating explicit goals also provide a driving force for an individual to pursue their desired goals (Karoly, 1999). The influence of goals on the behavior and decision-making of an individual has been widely studied in many forms, namely the achievement of goal orientations (Ames & Archer, 1988), goal-directed behavior (Boekaerts et al., 2006), and perceived goal feasibility (Perugini & Conner, 2000).
11
assessment of the perceived importance of achieving a particular goal (Abraham & Sheeran, 1993). It also implies that an individual may not perceive all of their goals as salient or of equal importance. As a result an individual organized their goals, often hierarchically, based on their desire to achieve their goals.
The focal argument of the theory of goal importance is that the relationship between goals and display of performance would be more positive for students who placed higher importance and had stronger willingness to achieve related goals, compared to those who did not (Sideridis & Kaissidis-Rodafinos, 2001).
The inclusion of the theory of goal importance to complement the TPB has been studied by Sideridis and his colleagues (Sideridis, 2002; Sideridis & Kaissidis-Rodafinos, 2001; Sideridis & Padeliadu, 2001; Sideridis & Rodafinos, 2001;). They hypothesized that as a goal becomes more important for an individual, the link between goal and motivation to perform subsequent behavior becomes stronger (Sideridis & Kaissidis-Rodafinos, 2001). In a series of studies which Sideridis and his colleagues conducted on students on various levels (i.e., elementary school, high school, college), they found support for the inclusion of goal importance in the TPB as a direct predictor of motivation and subsequent behavior.
1.3 The Rationale behind the Present Study
The aim of the present study was to assess the contribution of motivation to predict medical students’ decisions in participating in extracurricular scientific research. Based on significant contributions from previous studies, medical students’ motivation will be measured through variables of the TPB and goal importance.
Amongst academics, the attainment of multiple goals involves more than just performing well in a task. It involves having management skills, volition strategies, a good understanding of rules and regulations, and access to a well-established social support network (Boekaerts et al., 2006). This increases the need for self-regulation, mainly because students should be able to keep their regular study and extracurricular activities in balance.
12 Graph 1: Conceptual Framework
1.3.1 Other Potential Factors
This study also tried to look further into other potential factors that affect medical students’ motivation in participating in extracurricular scientific research. These factors have been identified by previous studies as concerning factors that contribute to medical students’ decisions to participate in research activities, albeit not specifically studied in the context of their motivation.
1.3.1a Previous Research Experience
Past behavior and experiences are often found as good predictors of future behavior (Ajzen, 2011). Past studies into medical students’ motivations have also included the influence of past experiences in scientific research as an influencing factor in their decision to be involved in scientific research (Abdel Meguid & Khalil, 2016; Rosenkranz et al., 2014; Seymour et al., 2004). This is because past experiences provide students with a better understanding of what scientific research activities entail. It is thought that (positive) past research experiences enhance the students’ motivation as they have more positive attitudes and experiences, providing them with confidence to conduct scientific research (Conner & Armitage, 1998; Craney et al., 2011; Seymour et al., 2004).
1.3.1b Information Sufficiency
13
14 CHAPTER 2: Research Methods
2.1 Research Context
This study was performed in the University Medical Center Groningen (UMCG), Netherlands. It involves students who are currently studying at the Faculty of Medicine. The regular study program in the Faculty of Medicine at the UMCG is divided into four learning communities; Molecular Medicine, Global Health, Intramural Care, and Sustainable Care. The extracurricular scientific research program involved is the Junior Scientific Masterclass (JSM). The program offers courses and special programs to help students to familiarize themselves with methods and hands-on experience of scientific research. As participation in scientific research activities requires active contributions from students, it will be used in this study as a threshold to define students’ motivation in scientific research (Van der Wouden et al., 2016). On top of that, participation in the program has an extracurricular nature and is not mandatory. Thus, it is assumed that it requires students to have a higher motivation to carry out scientific research activities and to be willing to participate in JSM courses or programs.
The so-called Pilot Project is one of the programs offered at the JSM which allows students to participate directly in a scientific research project, under the supervision of clinical doctors or experienced researchers. Activities in the Pilot Project involve writing a research proposal, finding a supervisor, gathering and analyzing data, and writing a research report. The UMCG has two kinds of honors programs that medical students can pursue. The first one is the university bachelor honors program, where students from across the faculty can join the program. The second one, is the JSM bachelors’ honors program, where they can obtain the honors program by doing extra 60 ECTS on top of their regular bachelors’ medicine study.
2.2 Research Design
This study employs mixed methods of data collection and analysis—both quantitative and qualitative. The study aims to expand knowledge concerning students’ motivation in doing scientific research using a questionnaire with closed-text and open-text responses, and interviews.
15
The qualitative part of the research aims to support the results of the questionnaire, and also to explore other possible factors that are not yet included in the study. This part will comprise of open-text questions in the questionnaire and interviews. Open-text questions in the questionnaire will be asked to explore other factors that may influence their motivation in doing extracurricular scientific research. In the interviews, the aim is to gain a deeper understanding of factors found in the open-text responses to the questionnaire.
Due to time constraints and the deadline of the FEB Master’s Thesis, the interview and its analyses are not going to be included in this thesis. With the intention to publish a scientific journal paper, this study will continue after the deadline of the FEB Master’s Thesis, which has been planned for since the beginning of this study. Further study will be pursued by the master’s student (Daindra W. Utami), with the help and under the supervision of the three current supervisors, prof. dr. ir. Jo van Engelen from the FEB University of Groningen, and Dr. Joke van der Mark-van der Wouden and Dr. Johanna Schönrock-Adema from the UMCG.
2.3 Sampling Procedure
Participants are divided into two groups; students who have decided to participate in scientific research activities (have participated in at least one JSM course or have participated in at least one Pilot Project) (Group 1), and students who have never participated in any JSM programs (Group 2). The questionnaire is made available to all 2nd and 3rd-year students, regardless of their participation in the JSM. This stratification is made in reference to the extent of their motivation to participate in scientific research activities. As mentioned, because a Pilot Project requires higher commitment and effort, it is assumed that this group of students has higher motivation in doing scientific research.
Table 1: Grouping Criteria
Students who have participated in the JSM (Group 1)
Students who have never participated in the JSM (Group 2) Expected (or obtained)
ECTS from the JSM by the end of the semester
≥ 1 ECTS 0 ECTS
or Pilot project Done / Ongoing No pilot project
16 Population
3rd year Students Population
420 students
For interviews, the number of 3-4 informants per category of actor--and less than 20 informants in total-- is targeted, as it is a number that is needed to reach theoretical saturation without resulting in information overload, which will hinder a good understanding of what actually is happening (Kvale, 1996; Zaltman, 1997). However, this criterium is made flexible, depending on the additional knowledge from subsequent interviews. It will also depend on whether the last interviewee brought insights and knowledge that are different from previous interviewees or or that those insights and knowledge reveal something new that is in line with the interests of this study.
2.4 Research Instruments
2.4.1 Questionnaire & Open Questions
To ensure that the questionnaire designed is valid, this research tailored existing, validated questionnaires to our research context (Ajzen, 2002; Kyle et al., 2014; Manstead & Eekelen, 1998; Rosenkranz et al., 2015; Sideridis, 2005). The adaptation of questionnaires from previous studies also allows for comparing the results of this study with those of previous ones. In addition to that, pre-test validations are carried out by asking five medical students to check the questionnaire to ensure that respondents fully understand what is being asked. This is done by asking students to read the questionnaire out loud while filling them out. Any concerns and advice that could help make the questionnaire easier to understand were noted down. After that, adjustments are then made accordingly by keeping in mind the concept of the study. This cycle is repeated twice, to ensure that the questionnaire made is easy to read and understand by students without deviating from the intent of the study.
Open-text questions were added to the questionnaire as a preliminary measure to find out other factors that influenced their motivation in participating in the extracurricular scientific research program. The data resulting from open-text questions was analyzed using ATLAS.ti to ensure a structured coding process.
2.4.2 Interviews
17
better. The second set of questions attempted to employ ZMET to explore unconscious factors that contribute to students’ decisions to participate in the extracurricular scientific research program. By employing this method, it will put interviewees in control more—rather than the interviewer—which is used to explore new factors that are potentially unknown to researchers (Coulter et al., 2001; Zaltman, 1997). Before interviews, students were asked to prepare 8-12 images that represent their thoughts, impressions, or goals associated with scientific research activities and the JSM and will be asked to bring these images to the interview. These images can be anything and can be sourced from anywhere (i.e., internet or personal collections), as long as informants are able to explain the connection between images and the study topic. Each interview will take approximately 60-90 minutes. This part of the study will be executed as planned after the FEB Master’s Thesis defense.
2.5 Study Procedures
2.5.1 Questionnaire Procedure
The students were approached through an email that consists of a link to the online questionnaire and an information letter about the study. The email was sent from a JSM secretary’s email account so that researchers do not have access to contact details and names of students. This process happens automatically via Qualtrics, and researchers cannot find out who has or has not participated in the research. The questionnaire was distributed twice through the UMCG’s student mailing system, one announcement through UMCG’s student association (Panacea) mailing system, two announcements on the JSM’s Nestor page (the university’s online student portal), on Facebook groups of medical students (GNK 2016 and GNK 2017, two announcements on each group), and giving short presentations at the beginning of three introduction lectures.
Upon opening the link to the questionnaire, respondents were faced with a welcome screen which contains a summary of information about the study and a download link to the complete information letter. Respondents were asked for their informed consent to participate in the welcome screen, where respondents can check a tick-box to indicate that they have agreed to participate in the research. Respondents can only continue to the study’s questionnaire after they have given their consent to participate. To avoid undesirable missing data, a force response command was applied to the questionnaire to ensure that students who wish to complete the questionnaire do not forget to fill in any questions.
18
consent letter, and instructions on how to sign up can be sent. The data gathering period for the questionnaire lasted one month (mid-September 2018 – mid-October 2018).
2.5.2 Interview Procedure
Before the interview starts, the interviewer explains the nature of the study, gives an opportunity for respondents to ask further questions about the study, and then respondents are asked to sign the informed consent letter. The interview will be recorded for transcription purposes. It was expected that each interview would take 60-90 minutes.
2.6 Measures Used in the Questionnaire
Questions from previous studies were adapted to measure variables relevant to the TPB, goal importance, and students’ decisions to join (or not join) extracurricular scientific research activities. Changes and adaptations were made by following guidelines presented in the literature to make sure that the intentions of the questions were kept intact (questionnaire can be found in Appendix 1).
2.6.1 Theory of Planned Behavior (TPB)
The TPB addresses beliefs that are domain-specific, and thus measures relating to the TPB are necessarily developed such that they reflect the domain of interest. Questions from past studies, along with guidelines presented in the literature about TPB, were developed for the present study in such a way that they reflect the domain of interest around a student’s decision to participate (or not participate) in extracurricular scientific research activities.
Items measuring behavioral beliefs were designed to examine the extent to which students have a positive stance and positive attitudes towards research activities (i.e., “I like doing scientific research” “To be a good doctor, I think having scientific research skills is important” “Having research skills is important for keeping up to date in the clinical field”) (Ajzen, 2002; Rosenkranz et al., 2015; Sideridis, 2005).
Those measuring normative beliefs were concerned with the extent to which participants believed that they are in a supportive social environment (e.g., “My family and friends will support me during difficult times”, “I will be more likely to join JSM activities if my friends are joining”) (Ajzen, 2002; Manstead & Eekelen, 1998; Rosenkranz et al., 2015).
19
2.6.2 Goal Importance
The theory of goal importance addresses the idea that decisions and behavior are prompted by goal-directed motivations. Questions pertaining to this theory were concerned with the extent to which students believe that obtaining scientific research skills and experience is important and would benefit them in their study and future career (e.g., “Obtaining scientific research skills is one of my main objectives during my medical study”, “I am interested in having a scientific research career after my medical study”). Respondents provide answers for this measure using 1-7 Likert Scale (“Strongly disagree”-“Strongly agree”) (Kyle et al., 2014; Manstead & Eekelen, 1998).
2.6.3 Open Questions
Open questions are incorporated in the questionnaire as an exploratory measure to find out other unidentified factors that can influence students’ decision to participate (or not participate) in extracurricular scientific research activities. Open questions that were included are reasons why they participate (or do not participate) in JSM courses and programs, reasons whether they intend to participate (or not participate) in JSM courses and programs in the future (for Group 2 students), and other factors that may influence (promote or hinder) their decision to participate in the JSM.
2.7 Ethicality of Data Collection and Handling
As mentioned, the online questionnaire is distributed through the JSM secretariat email account to ensure the anonymity of the students who participated in the research. Consequently, researchers have no access to a list of emails or other details of respondents. Responses are kept anonymous, and thus researchers never get to see who filled in what. This also applied to reminder emails to non-responders, researchers do not know to whom these reminder emails are sent as the process happens automatically. Email addresses of students are collected in a different form to disconnect them from the data to further ensure the anonymity of participants.
Responses and other data gathered (e.g., pictures that informants brought to interviews) are kept confidential by the UMCG. Participation of respondents in this research is voluntary, both for the interview and the questionnaire. When respondents do not wish to finish the questionnaire, they can just leave at any point as the system will automatically delete unfinished responses after one week.
20
which indicates they agreed that their responses will be used for the research. For the interview, respondents are asked to sign an informed consent letter. The same letter will also be attached on the email of the invitation to the interview. These steps are taken to make sure that respondents are sufficiently informed about the study and what kind of data that will be expected from them.
For about five months prior to the start of the study, plenty of time and efforts were dedicated to write a research and ethics proposal to be submitted to the ERB (Ethical Review Board) of the NVMO (Nederlandse Vereniging voor Medisch Onderwijs, Netherlands Association for Medical Education) (submission can be found in Appendix 5). By following the highest standards, this research and its instruments have been approved by the ERB of the NVMO. The research proposal submission for this study can be found in Appendix 5.
2.8 Data Analysis: Anticipating and Resolving Potential Issues
2.8.1 Closed Questions Data Analysis
Before proceeding with the analysis, it needs to be checked whether this study has multicollinearity problems. Multicollinearity indicates that there is a high correlation on at least two independent variables within a study, which makes it hard to predict which variable causes what effect to the dependent variable. If it left uncorrected, it can create substantial problems when analyzing the data, as independent variables that suffer from multicollinearity problems include unstable and biased standard errors, which could result in indefensible and misleading statistical interpretations (Grewal et al., 2004; Vatcheva et al., 2016).
There are two ways in which for multicollinearity can be checked; through correlations coefficients and Variance Inflation Factor (VIF). A correlation coefficient indicates how highly correlated independent variables are. In general, the threshold of a correlation coefficient to be assumed suffer from multicollinearity is above 0.8, therefore correlation coefficients should stay below it (Field, 2018). Meanwhile, VIF measures the linear association between an independent variable and all of the other independent variables, and it also quantifies the severity of multicollinearity, should it be present. The general threshold of VIF for multicollinearity is above 10, so the VIF value should stay below it (Grewal et al., 2004).
To answer the first research question, “Do high TPB and high goal importance translate into a more positive intention to do extracurricular scientific research program?”, a binary logistic regression was conducted. This analysis aims to reveal whether the TPB and goal importance can be a significant predictor of students’ decisions to participate in extracurricular scientific research activities.
21
ANOVA tests, and a Mann-Whitney U test were conducted. An independent sample t-test was conducted to see whether the difference in mean scores across students who participated in extracurricular scientific research activities are significantly different from students who did not participate. MANOVA and other follow up tests are conducted to confirm further that these differences (results of Independent Sample t-test) were indeed significant.
Before MANOVA was conducted, tests for the criteria fulfillment of MANOVA is carried out to ensure that the data of this study is eligible to be analyzed using MANOVA (Fields, 2018). The first criterium of MANOVA is that there should be no outliers in the dataset. To identify univariate outliers, this study employed the threshold of 2.2 IQR (Inter-Quartile Range) instead of 1.5 standard thresholds employed by SPSS Statistics, as suggested by Hoaglin & Iglewicz (1987). This is because the study by Hoaglin & Iglewicz (1987) has demonstrated that the threshold of 1.5 can wrongly identify data as outliers when the sample size of the study is less than 300. Multivariate outliers are identified using Mahalanobis Distance with a critical value of chi-square distribution 13.82 (p = 0.001, df = 2) and checked further by looking at the probability value, which is obtained by comparing the Mahalanobis Distance value with Chi-Square and degrees of freedom. In order for an observation to be classified as a non-outlier, the probability value needs to be p > 0.001.
Secondly, the data needs to have a normal distribution. To check this, a normality test is performed. This study followed a suggestion from Ghasemi & Zhadeiasl (2012) and Thode (2002) to focus on the result given by a Shapiro-Wilk test instead of a Kolmogorov-Smirnov one. This is because they have argued that Shapiro-Wilk has a higher power to explain normality in distribution, compared to Kolmogorov-Smirnov. In order for predictors to be assumed has a normal distribution, p-value of the variables should be p > 0.05. Thirdly, the data needs to have homogeneity of variances-covariances and homogeneity of variances. This means that the level of variance across data within a group is more or less constant. This study tests the homogeneity of variances-covariances using a Box Test of Equality of Covariance Matrices and test homogeneity of variances using a Levene’s Test. In order for variables to be assumed to have equal variances, they should have a p-value of p > 0.05. MANOVA will be conducted only to predictors that meet the criteria. If the result of MANOVA is significant, one-way ANOVA is to be conducted to see how this study’s predictors differ across student groups. All statistical analyses of this study were performed using IBM SPSS Statistics software.
2.8.2 Open Questions Analysis
22
data, as expressed by participants in the questionnaire (Maruster & Gijsenberg, 2013; Popping, 2015). On top of that, it also allows for findings to be presented transparently (able to be demonstrated to others), and defensible (justifiable given the objectives of the research) (Thomas, 2006).
Preliminary screening was conducted to find underlying themes and suitable keywords for coding. Coding is essential for a qualitative data analysis as it extracts words and the essential meaning of respondents’ feelings, beliefs, or knowledge (Roth, 1971). The preliminary screenings were conducted on two levels; with all groups combined and with each group separately.
23 CHAPTER 3:
Questionnaire Results & Analysis
3.1 Respondents Profile
Our survey managed to gather 165 responses, of which 55 responses were excluded; 47 to incomplete responses, and 8 responses due to having failed the attention check in the questionnaire, yielding 110 responses that are usable to conduct analysis for this study.
From 110 respondents, 56 are second year students, and 54 are third year students. This shows that out of the population of second year students (459 students), 12.2% responded to the questionnaire. Meanwhile, out of the population of third year students (also 420 students), 11.7% responded to the questionnaire. Based on the study programs, 36 students study Molecular Medicine, 31 students study Global Health, 19 students study Sustainable Care, and 25 students study Intramural Care. Based on gender, 34 students are male, 76 students are female, and one student identifies as a non-binary gender. (For nationalities of respondents, please see Appendix 2)
The number of respondents generated is probably influenced by the fact that the researcher of this study is unable to reach all of the 2nd and 3rd-year students in the Faculty of Medicine at the UMCG. It is expected that a short presentation at introductory lectures will be the main driver for students to fill in the questionnaire, as they directly receive explanations about the study from the researcher and have an opportunity to ask questions straight away. However, as it turns out, only less than half of the students (from the listed populations in Table 1) attended introductory lectures. Researchers are also only able to announce the questionnaire in the online student portal of students who participated in the JSMs. Researchers are also not allowed to spread flyers around the campus or approach students directly to ask them to fill in the questionnaire. Announcements in the Facebook groups are also expected to be not as effective in persuading students to fill in the questionnaire as many messages were posted on the groups every day.
3.2 Closed Questions Results
3.2.1 Descriptive Analysis
24
the interrelatedness of predictors within the test, which will provide further evidence of reliability for this study as a whole (Cronbach & Shavelson, 2004; Santos, 1999; Taber, 2018; Tavakol, 2011).
Table 2 also shows that CA for predictors from the TPB are relatively low. Ajzen (2011a) has explained that the TPB contains random errors and therefore rarely exhibit reliabilities over 0.75. On top of that, the internal consistency of the data can be easily influenced by external factors such as gender and culture (Ursachi et al., 2015). This is probably what caused the low CA values as there are unbalanced representations in the data, based on both gender and nationalities (Appendix 2).
The impact can be seen as well for the predictor normative belief, which has the lowest CA among all predictors. Normative belief represents the influence of social pressure and support from people within the students’ inner social circles. Unbalanced representations are thought to be a cause for this, as social environment may have different meaning and influence across different cultures (Aaker & Maheswaran, 1997).
Table 2: Means, Standard Deviation, and Pearson Correlations of Predictors
Mean SD VIF CA 1 2 3 4
1 TPB Behavioral 4.294 0.678 1.805 0.634 --
2 TPB Normative 4.640 0.748 1.109 0.413 0.235* --
3 TPB Control 4.580 0.825 1.115 0.636 0.316** 0.126 --
4 Goal Importance 3.749 1.382 1.773 0.852 0.641** 0.305** 0.211* --
Note: n = 110; SD = Standard Deviation, VIF = Variance Inflation Factor, CA = Cronbach’s Alpha; predictors were measured with the Likert Scale 1-7; 1 (Strongly disagree), 2 (Disagree), 3 (Somewhat disagree), 4 (Neither agree or disagree), 5 (Somewhat agree), 6 (Agree), 7 (Strongly Agree); all (2-tailed) correlations of 0.15 and over were significant at 0.05(*) level and 0.01(**),
25
about the JSM. From 67 students who did not participate in the JSM, 45 students (67.2%) perceive that they do not have sufficient information.
There is also statistically significant evidence that students who participated in extracurricular scientific research activities scores higher based on behavioral beliefs, normative beliefs, and goal importance. As this study is interested to study further the differences between students who have participated and those who have not, further statistical analysis will be conducted to confirm these differences.
Table 3: Comparison between Motivational Scores and Students’ Characteristics
N TPB Behavioral TPB Normative TPB Control Goal Importance
Gender Male 33 4.477 (0.718) 4.715 (0.814) 4.824 (0.884) 4.230 (1.391) Female 76 4.220 (0.652) 4.624 (0.712) 4.474 (0.787) 3.563 (1.331) Non- binary 1 3.875 3.400 4.600 2.000 t-values 1.833* 0.590 2.058** 2.372** Year of Study 2nd Year 56 4.326 (0.677) 4.625 (0.743) 4.579 (0.896) 3.732 (1.363) 3rd Year 54 4.261 (0.683) 4.656 (0.760) 4.581 (0.754) 3.767 (1.413) t-values 0.496 -0.213 -0.018 -0.130
Previous Research Experience
Yes 14 4.491 (0.579) 4.300 (0.890) 4.300 (0.985) 4.114 (1.109) No 96 42.66 (0.689) 4.660 (0.717) 4.621 (0.797) 3.696 (1.413)
t-values 1.165 -1.841* -1.364 1.059
Perceive had sufficient information
Yes 54 4.369 (0.687) 4.619 (0.817) 4.659 (0.868) 4.203 (1.376) No 56 4.223 (0.667) 4.660 (0.681) 4.503 (0.782) 3.311 (1.249)
26 Have participated in the JSM
Yes (Group 1) 44 4.660 (0.527) 4.880 (0.818) 4.730 (0.747) 4.926 (1.015) No (Group 2) 67 4.058 (0.662) 4.487 (0.661) 4.484 (0.863) 2.994 (1.008)
t-values -5.009*** -2.767*** -1.539 -9.765***(a)
Note: Mean(SD), variables were measured with the Likert Scale 1-7 (Strongly Disagree - Strongly Agree), Group 1 = Students who have participated in JSM activities, Group 2 = Students who have never participated in JSM activities. Significance is denoted with (*) p < 0.1, (**) p < 0.05, (***) p < 0.01. (a) = Equal variances not assumed.
3.2.2 Binary Logistic Regression
The result of binary logistic regression analysis shows that 58.9% variation in students’ decision to participate can be accounted for from behavioral beliefs, normative beliefs, control beliefs, and goal importance. The result of a Hoser & Lemeshow test shows that the conceptual model for this study fits the data (χ2 = 6.079, df = 8, p = 0.638, p > 0.05). The model correctly classified 83.6% of cases where students
decided to participate in scientific research activities and 72.1% cases where students decided not to, giving an overall percentage of correct prediction rate of 79.1%.
However, the insignificant results of the TPB predictors could be due to the fact that the resulting data have normal distribution (see the normality test in Appendix 2). Strömbergsson (2009) have suggested that in cases where predictors are normally distributed, a discriminant analysis may be more useful in analyzing predictions. However, looking from results of Table 2, the variable behavioral beliefs is highly correlated with goal importance (r = 0.641), which yields a significant result. This might imply that at least behavioral beliefs could have an influence towards students’ decision. To check the robustness of the TPB in this study, a separate binary logistic regression analysis (without including the goal importance) is conducted (see section 3.4.3).
Meanwhile, the analysis shows that goal importance is a significant predictor for students’ decisions to participate in scientific research activities (χ2 = 62.690, df = 4, p < 0.0001). It is significant
27 Table 4: Results of Binary Logistic Regression
B S.E. Wald df Sig. Exp(B) 95% C.I.for EXP(B) Lower Upper TPB Behavioral0.199 0.569 0.122 1 0.727 1.22 0.4 3.72 TPB Normative 0.245 0.359 0.466 1 0.495 1.278 0.632 2.584 TPB Control -0.079 0.351 0.051 1 0.822 0.924 0.465 1.838 Goal Importance 1.492 0.314 22.593 1 0.000 4.448 2.404 8.231 Constant -7.998 2.782 8.266 1 0.004 0
Note: Variable(s) entered: TPB Behavioral, TPB Normative, TPB Control, Goal Importance.
3.2.3 Robustness Check of the TPB
This analysis is conducted to investigate the model robustness of the TPB in this study. The result of the Hoser & Lemeshow test shows that the model for this study fits the data (χ2 = 8.353, df = 8, p =
0.400, p > 0.05). The result of the binary logistic regression analysis shows that behavioral beliefs shows that the construct is now significant (Wald = 13.509, p < 0.0001) with OR of 5.44 (at 95% confidence interval of 2.404 - 8.231). This means that the construct explains no additional variance in addition to goal importance. This indicate that behavioral beliefs may have an influence towards students’ decision when the variable goal importance is not considered. The results for normative beliefs (Wald = 3.35, p = 0.067, p > 0.05) and control beliefs (Wald = 0.014, p = 0.907, p > 0.05) are still not significant.
Table 5: Results of Binary Logistic Regression for Robustness Check
B S.E. Wald df Sig. Exp(B) 95% C.I.for EXP(B) Lower Upper TPB Behavioral1.694 0.461 13.509 1 0.000 5.44 2.205 13.425 TPB Normative 0.567 0.31 3.35 1 0.067 1.763 0.961 3.235 TPB Control -0.034 0.294 0.014 1 0.907 0.966 0.544 1.718 Constant -10.375 2.52 16.954 1 0.000 0
28
3.2.4 MANOVA Analysis
Prior to conducting the MANOVA analysis, an assessment is conducted to ensure that variables actually meet the criteria and can be analyzed using MANOVA. The results show that only predictors from the TPB (i.e., behavioral beliefs, normative beliefs, and control beliefs) have normal distributions and therefore are eligible for MANOVA. For more details about this criteria test, please see Appendix 3.
The result of the MANOVA analysis shows that there is a statistically significant difference in students’ decision to participate in extracurricular scientific research activities based on their motivation (as measured by TPB) (F (3, 106) = 9.625, p < 0.0001, Wilk’s 𝛬 = 0.786, partial η2 = 0.214).
Table 6: Multivariate Test Result
Effect Value F Hypothesi s df Error df Sig. Partial Eta Squared Noncent. Parameter Observed Power (c) Intercept Pillai's Trace 0.989 3180.598 (b) 3 106 0.000 0.989 9541.795 1 Wilks' Lambda 0.011 3180.598 (b) 3 106 0.000 0.989 9541.795 1 Hotelling's Trace 90.017 3180.598 (b) 3 106 0.000 0.989 9541.795 1 Roy's Largest Root 90.017 3180.598 (b) 3 106 0.000 0.989 9541.795 1 G1G2_Dummy Pillai's Trace 0.214 9.625(b) 3 106 0.000 0.214 28.874 0.997 Wilks' Lambda 0.786 9.625(b) 3 106 0.000 0.214 28.874 0.997 Hotelling's Trace 0.272 9.625(b) 3 106 0.000 0.214 28.874 0.997 Roy's Largest Root 0.272 9.625(b) 3 106 0.000 0.214 28.874 0.997
Note: Computed with Design: Intercept + G1G2_Dummy, (b) = Exact statistic, (c) = Computed using alpha = 0.05
29
0.189) and normative beliefs (F (1, 108) = 7.657, p = 0.007, p < 0.05, partial η2 = 0.66). Meanwhile, it is
found that there is no significant evidence that students who decide to participate and not participate in extracurricular scientific research activities differ in control beliefs (F (1, 108) = 2.369, p = 0.127, p > 0.05, partial η2 = 0.021). Therefore, it can be concluded that these results confirm the t-tests previously
conducted.
Table 7: Result of Univariate One-Way ANOVA for Behavioral Beliefs
Source Type III Sum of Squares
df Mean Square
F Sig. Partial Eta Squared Corrected Model 9,434 (a) 1 9.434 25.087 0.000 0.189 Intercept 1991.32536 1 1991.325 5295.150 0.000 0.980 G1G2_Dummy 9.43445542 1 9.434 25.087 0.000 0.189 Error 40.6151184 108 0.376
Total 2078.57813 110 Corrected Total 50.0495739 109
Note: (a) R Squared = .189 (Adjusted R Squared = .181)
Table 8: Result of Univariate One-Way ANOVA for Normative Belief
Source Type III Sum of Squares
df Mean Square
F Sig. Partial Eta Squared Corrected Model 4,035 (a) 1 4.035 7.657 0.007 0.066 Intercept 2297.33965 1 2297.340 4359.809 0.000 0.976 G1G2_Dummy 4.03492676 1 4.035 7.657 0.007 0.066 Error 56.9090732 108 0.527
Total 2429.2 110
Corrected Total 60.944 109
30
Table 9: Result of Univariate One-Way ANOVA for Control Belief
Source Type III Sum of Squares
df Mean Square
F Sig. Partial Eta Squared Corrected Model 1,593a 1 1.593 2.369 0.127 0.021 Intercept 2223.461 1 2223.461 3305.687 0.000 0.968 G1G2_Dummy 1.59336203 1 1.593 2.369 0.127 0.021 Error 72.642638 108 0.673
Total 2381.64 110
Corrected Total 74.236 109
Note: (a) R Squared = .021 (Adjusted R Squared = .012)
3.2.5 Mann-Whitney U Test
To find out whether students who decide to participate and not participate in extracurricular scientific research activities are different based on goal importance—which does not have a normal distribution—a Mann-Whitney U test is conducted. The result shows that students who decided to participate in extracurricular scientific research activities have statistically significant higher goal importance compared to students who decided not to participate (U = 283.5, p = 0.000, p < 0.05). This result confirms the result of the t-test. Therefore, it can be concluded that students who participated in scientific research activities show that they have higher goal importance than those that did not participate.
Table 10: Result of Mann-Whitney U Test
G1G2_Dummy N Mean Rank Sum of Ranks
Goal Importance 0 (Not participate) 67 38.2 2561.5
1 (Participate) 43 82.4 3543.5
Total 110
Test Statistics (a)
31 Mann-Whitney U 283.5
Wilcoxon W 2561.5
Z -7.096
Asymp. Sig. (2-tailed) 0
Note: (a) Grouping Variable: G1G2_Dummy
Based on the results of the statistical analyses, it can be concluded that students who have participated and have never participated in extracurricular scientific research activities have significantly different levels of motivation (based on behavioral beliefs, normative beliefs, and goal importance). It is also found that the results from predictor control beliefs are insignificant across the analyses.
The results from the analyses also show that students who had higher goal importance were significantly more likely to participate in extracurricular scientific research activities. However, there are inconclusive results as to whether students who score higher on the TPB predictors were more likely to participate in them.
3.3 Open-questions answers
The answers to open questions were analyzed through a coding process. From the preliminary screening, underlying themes and keywords were identified and used for coding. For more information about the keywords used for coding, please see Appendix 4. Properties explain the underlying theme and categorization of codes.
Table 11: Codes for Open Questions
Codes Properties Examples of sentences used by participants
Interest and enjoyment in doing scientific research (Applied in Group 1 & Group 2)
Happiness obtained from satisfying (research) curiosity and activities in-between
Group 1:
- “I enjoy (doing) the Pilot Project offered by the JSM...”
- “I like the extra challenge (that JSM provides) on top of my medicine study” - “There are many undiscovered parts in (clinical) surgery that still requires to a further look into”
Group 2:
32
- “I don’t think I would like it (doing scientific research)”
Improving scientific research skills
(Only applied to Group 1)
Willingness to do extra effort in order to become a better scientist
“I believe (joining JSM programs) will improve my skills”
“To improve my research skills is an important factor for me”
“(I) Want to know more about research in psychiatry”
Extrinsic benefit factors (Only applied to Group 1)
Willingness to do effort in to obtain extrinsic benefit
- “I am intending to apply for a scholarship … and (joining) JSM will make a
noteworthy distinction”
- “Because (joining JSM) will makes it easier to do an MD/PhD project” Information
(Applied to Group 1 and Group 2)
The level of information that a student has about the extracurricular scientific research program
Group 1:
- “There needs to be more information how JSM courses work”
- “I wish I knew more about JSM in my first year”
- “I am not sure what it (JSM Honors Degree) is”
Group 2:
- “JSM (courses and programs) are not very clear to me”
- “I do not know what (Pilot Project) is” - “I do not think I have enough information to decide (whether or not to) participate in the JSM”
Course organization (Only applied to Group 1)
Factors related to the program management of extracurricular scientific research activities
G1: “I participated so far because the (JSM) courses can be integrated into my schedule”
“When schedules don’t work … sometimes the lecturer and students don’t know what to do”
Other extracurricular commitments
(Only applied to Group 2)
Activities that students participated in, but not included in the regular curriculum, and it could be outside of the university’s boundaries
“My free time already goes to playing in two orchestras”
“I play sports … which may sometimes get in conflict with JSM (activities)”
Burden from regular study (Only applied to Group 2)
Factors that arises as a result of the student’s regular study, in a way that it makes student feel that they cannot take other extra
“Stress (from regular study)”
“(I am) Already struggling to keep up with the regular study”
33 responsibilities
Table 12 was made to summarize the answers between students of Group 1 and Group 2, which is beneficial in comparing the difference in answers between the two groups. It can be seen that the difference between their level of motivation was reflected within their answers. A new factor that arises from students is identified as information sufficiency issues. Students who have participated in the extracurricular scientific research activities mention information availability to be an issue. Meanwhile, students who have never participated mentioned that they do not have enough information about them.
Answers from open questions by students of Group 1 show that they have stronger motivations and more defined goals for why they are participating in extracurricular scientific research activities. It reveals that they participated because they like doing scientific research and seek to access support facilities provided by the scientific research program (e.g., access to a good lab, research funds, etc.). Some answers also reveal that they are participating in extracurricular scientific research activities because they would like to pursue an MD/PhD program after they finish their bachelors’ study. Students of Group 1 identify that lack of flexibility in the deadline of research project intake, course scheduling problems, and more complete information about JSM courses and available research project as factors that hinder them from being more active in JSM activities.
Answers to open questions by students of Group 2 shows that they do not have strong motivations and clear goals to participate in JSM activities. Most answers also reveal that students in this group are not interested in doing research activities, do not know enough about the program, and they are worried that participating in JSM activities will add too much stress and burden to their regular study.
However, it is also identified that misinformation is a big issue for the participation decision for students in Group 2. From the answers to open questions, it was found that there are a lot of students who have incorrect information about JSM programs. For example, some students plan to participate in JSM courses during their masters. This is, however, not possible as JSM courses are aimed only at bachelor’s students. Another example is also that some students believe that JSM courses are only for students with high study performance and possess advanced statistic knowledge and strong research experience. However, this is not necessarily true as all students have the same opportunity to participate in JSM courses.
Table 12: Answer Summaries from Open Questions
Questions Group 1 Group 2
What is the main reason why did you (G2: not) participate in the
- Want to learn about scientific research, interesting research
34
JSM? subject
- Want to obtain the (honors) degree, distinction to obtain future scholarship
- Expand future career possibilities
activities, too busy with regular study
Is there any other factor(s) that hinder you to participate (G1: more) actively in the JSM?
- Too many rules and regulations, clashes in schedules with regular study programs
- Communication with the JSM is difficult, out-of-date information on the JSM’s website
- Too low study points compared to the effort
- Assumptions that it will take a lot of time and compete with regular study programs and schedules
- High expectation to complete the course, worry about stress - Unclear sign-up and course
35 CHAPTER 4: Discussion and Conclusion
This study is drawn with the main aim to find a way to increase medical students’ motivations to participate in scientific research activities. This master’s thesis focuses on identifying factors that influence medical students’ motivation in participating in scientific research activities using the TPB and goal importance, which is studied quantitatively. This chapter aims to discuss the outcomes of the questionnaire study and its implications, the plan of the subsequent study, and also limitations of this study.
4.1 Discussion
To answer the research question “Do positive TPB and high goal importance translate into a more positive intention to participate in extracurricular scientific research program?”, this study utilized a binary logistic regression analysis for closed questions. The question aimed to reveal whether the TPB and goal importance are substantial predictors to students’ decisions. The results only show goal importance to be a significant predictor of students’ decision. It suggests that when students have a concrete goal of what they want to achieve (or what they can obtain) by participating in scientific research activities, and if they perceive obtaining such goals as important, the likelihood of them participating in scientific research activities increases.
The answers to the open questions also reflect the relevance of goal importance as a predictor of students’ decision to participate in extracurricular scientific research activities. The results show that students who have participated (Group 1) have clearer goals as compared to what they would like to achieve or benefit from by participating in extracurricular scientific research activities. It also shows that students of Group 1 have more positive attitudes towards scientific research and consider themselves to have the ability to perform their regular study and extracurricular scientific research activities in balance. These might be the aspects that explain why medical schools in the Netherlands have satisfactory scientific research performance.
36
predicting students’ decision (Strömbergsson, 2009). The subsequent study will further investigate this result.
To answer the research question “How do students who participate in scientific research activities differ—with regards to TPB and goal importance—from students who do not participate in scientific research activities?”, this study utilized independent sample t-test, MANOVA, One Way ANOVA, and Mann-Whitney U analyses. The result of an independent sample t-test shows that students who participated in extracurricular scientific research activities score higher based on behavioral beliefs, normative beliefs, and goal importance compared to students who did not participate. The analysis, however, found no evidence that students’ control beliefs differ throughout the participation group. These results are further supported by the result of MANOVA (followed up with univariate one-way ANOVA tests) and a Mann-Whitney U analysis, which also found that students who participated in extracurricular scientific research activities have a higher motivation to do so based on their behavioral beliefs, normative beliefs, and goal importance. These results imply that students of Group 1 have a more positive attitude towards scientific research activities, perceive that they have support from their social surroundings, and that participation in extracurricular scientific research activities has more benefit to them.
This study, however, found no evidence to support that control beliefs as substantial predictors of students’ decisions and also found none to suggest that students who participated in extracurricular scientific research activities have different motivational scores based on control beliefs. This finding is also in line with the study of Kyle et al. (2014) who conclude that control beliefs may not have served as a sufficient proxy measure of control abilities. However, open question answers of students from Group 2 reflect that they do not perceive that they are able to perform extracurricular scientific research activities and regular study in balance, which reflect the influence of control beliefs to students’ decision. This finding warrants further investigation which will be pursued in the subsequent study.
As another aim of this study is to determine factors that influence students’ motivation to participate in scientific research activities, turning to factors that have been previously explored by other studies. In particular, the influence of students’ previous research experience and the sufficiency of information that students have. This study found that students’ previous research experience only had a substantial influence on the normative belief of students. However, considering there is a noteworthy difference in the number of respondents who had previous research experience compared to those who had not, the result may be not valid. Another factor studied is the influence of perception of having sufficient information.
