CHAPTER 3 – METHODOLOGY
This chapter describes the methodology used and procedures followed to carry out this study. The following sub-sections include the rationale for the study and the proposed research questions derived from the literature review. This follows participant information, a description of the research design and the measure used, how the data were collected and analysed, and finally ethical issues.
The aim of this study was to explore relationships between personality traits and health-related behaviours by adapting the study in the Maltese context since such studies lacked significantly. A quantitative approach was chosen as it allowed a wider collection of data from the University of Malta student population. Furthermore, compared to qualitative research, quantitative research concerns with quantifying phenomena and therefore it provided a more accurate type of measure (Langdridge & Hagger-Johnson, 2009), which could easily compare to previous literature. In fact, this study aimed to replicate previous literature by using similar data collection tools.
Based on the literature review, the following research questions were drawn in an attempt to address some of the limitations encountered;
Are there significant differences between low, medium, and high levels of personality traits (Openness, Conscientiousness, Extraversion, Agreeableness, Neuroticism) and health-promoting behaviours?
Are there significant differences between low, medium, and high levels of personality traits (Openness, Conscientiousness, Extraversion, Agreeableness, Neuroticism) and health-deterring behaviours?
What is the relationship between (1) and (2) and demographic factors?
How do previous studies and the current personality-health study compare?
These research questions were used as a starting point for this study, and in an attempt to answer them, the researcher chose a student population, which was available for the study. In addition, a cross sectional survey-based study was selected because it allowed application to the local context.
The sample in this study included 576 University of Malta students, which varied from under-graduate to post-graduate courses, both full-time and part-time. The participants included 166 males (29%) and 410 females (71%), between 17 and 62 years, with a mean age of 24 years. Students came from various faculties and courses. The gender representation at the time of data collection was of 4642 males (42%), and 6425 females (58%).
A questionnaire was designed consisting of three main sections. The first section was constructed for participants to fill in demographics (Section A) which included questions about age, gender, and nationality. The following section comprised a personality inventory with statements that identified specific personality traits (Section B). The final section included a range of questions related to health-related behaviours (Section C). The questionnaire can be found in Appendix A, while the following sections provide a detailed description of how final two sections were designed.
Personality traits (Section B) were measured using the Big Five Inventory (BFI) devised by John et al. (1991). The main reason for choosing this inventory was that it presented an average alpha reliability of .83, which was remarkable for being a short scale measure. Reliability was similar to other personality measures like the NEO Five-Factor Inventory (NEO-FFI) developed by Costa and McCrae (1992) and the Trait Descriptive Adjectives scales (TDA) developed by Goldberg (1992), with a .81 and .84 alpha reliability, respectively. Likewise, convergent validity within these three personality measures was also similar. The BFI and TDA showed an overall convergence of .80, followed by the BFI and NEO-FFI with .77 and the TDA and NEO-FFI with .68 (John, Naumann, & Soto, 2008).
Another reason for choosing this instrument was because it provided a brief inventory with only 44 short phrases and with relatively accessible terms (John et al., 2008). Considering the population was students, this inventory was preferred over similar instruments because it was deemed more time efficient, thus it reduced the possible lack of response due to a lengthy questionnaire. Indeed, short scale instruments, like the BFI, not only save testing time, but avoid subject boredom and fatigue (Burisch, 1984).
Health-related behaviours (Section C) were measured by adopting questions from the study ‘Healthy Students Healthy Lives’ by Cefai and Camilleri (2009), because it was designed for Maltese students, and therefore provided a useful guideline for questions about health behaviours, and social wellbeing. Although questions were only used as a guideline, the test retest reliability produced a Pearson correlation ranging from 0.783 to 0.964, making the study significantly robust.
Health-related behaviour questions in the current study focused on; general health habits like food consumption and exercise, smoking habits, alcohol and drug-use, sexual behaviour, and social wellbeing. The questions were mostly Likert scale format and varied from four to six options. This allowed a distinction between considered health-promoting behaviours and health-deterring behaviours.
After combining the questionnaire, a pilot study was carried out to gain feedback about any difficult questions or ambiguities. This was conducted with ten university students, during which students were encouraged to ask questions and give feedback about the structure and the type of questions administered. A systemic table was used to record notes about issues being pointed out. Feedback and suggestions from the pilot study were later discussed with the supervisor. This allowed modifications in the questionnaire by; rewording of the questions, including additional options, and including definitions for terms like ‘junk food’ and ‘binge drinking’.
The finalised questionnaire comprising a total of 83 items and was distributed via Google Documents, a web-based survey provider, which allows the survey to be accessible online. In order to reach university students, the questionnaire was sent by email through the university registrar’s office, which reached the entire university population, about 11,067 students. By sending an email to all students, ensured randomness to some degree because every student received the questionnaire in the same time-frame and had equal amount of time to respond. The survey was available online for four weeks, to ensure a wider chance of participation, after which the online link was deactivated. A response rate of 5.2% was achieved. Interestingly, over half the sample had completed the questionnaire on the first day of dissemination.
The data collected was analysed using the SPSS version 20. Results for personality traits and demographics were analysed with results for health-related behaviours. Descriptive and inferential statistics were carried out using one way anova and independent sample t-tests, while the Pearson R test was used for the correlation coefficient.
There were no major ethical concerns with this study. Participants were briefed before they were directed to the survey link. Similar to an informed consent, the briefing section stated that participation was voluntary, responses were anonymous, and to avoid any harm, participants could stop the survey at any time. Participants were ensured that all data collected was treated with full confidentiality and was only accessed by the researcher and the supervisor.
Overall, this chapter presented the methodology used to carry out this study. Although this study aimed at replicating previous literature, a number of factors needed to be well planned and rationalised prior to the data collection. This was important not only to address the research questions, but also to ensure the questionnaire fitted well in the Maltese context. The following chapter presents the results derived from the data collected.