Methodology

Our group decided to collect the data via a survey/ questionnaire as opposed to conducting an experiment for a plethora of reasons.

1. Firstly, limited time and resources prevented us from conducting an experiment which would have been the ideal method of study. Conducting an experiment on such a scale also demands commitment from the subjects which we felt was not appropriate for us to do so. Hence, we decided to settle for a questionnaire.
2. Due to the large inclusion criteria defining our target population (young adults aged 20-25), it was not possible for a list of all members of our target population to be made available to us. As such, we were compelled to do snowball sampling. Since we had to do snowball sampling, a questionnaire that is simple to fill out plays well in this form of sampling and a large sample/ number of response covering geographical areas outside of our school would be compensation enough.
3. Subjects also have a greater sense of anonymity filling out a questionnaire compared to conducting an experiment which is pertinent in collecting accurate data regarding self-esteem due to the sensitive nature of the topic (ie. respondents might be embarrassed to admit/ reveal they have low self-esteem if the researcher is able to trace the data collected back to them and so they alter their answer)
4. As mentioned earlier, a questionnaire allows us to expand our reaches beyond the parameters of our school and collect data from respondents who are exposed to different environments. Coupled with the ease of filling out a short questionnaire taking no more than 3 minutes and barely any open-ended questions that require careful consideration to minimise deterring any potential respondents, there could potentially be a high response rate which would be useful in normalising and reducing the impact of variables related to social factors such as an individual’s upbringing, culture, etc. on the outcome if the sample obtained is large enough.
5. A questionnaire is also not biased towards or against any particular group of potential participants as opposed to an experiment. We also wanted to collect data from those who did are not considered to have exercised regularly and this group of people may not want to commit to our study if we are conducting an experiment as they might find it as too much of a disruption to their usual lifestyle.

 

Figure 1

The respondents were required to read a short introductory paragraph (Fig. 1) before attempting the questions. This is to obtain an informed consent from the people who were taking part in our questionnaire. In our introductory paragraph, we included the the name of members who were conducting the study, purpose of the study, an estimate of survey completion time, information of their right to withdraw from the survey and an email to contact for any inquiry. This is to ensure that the potential respondents were well informed about the study and knows how it would be conducted. Hence, they would be able to make voluntary and rational decision to withdraw if they felt uncomfortable at any point of time during the process of answering the questionnaire.

After having constructed our questionnaire, we sent the link out to our available social circles as our group decided to carry out snowball sampling. As mentioned earlier, we were compelled to do snowball sampling as our inclusion criteria of young adults aged 20 to 25 entailed a large target population and so  a list of all members of the target population could not possibly be made available to us. Snowball sampling played well with a questionnaire as our mode of collecting data and obtaining a large sample size would compensate for having chosen a non-probability based sampling method. Snowball sampling is non-probability based because not everybody within the target population has been given an equal opportunity to be  selected and chance to respond to the questionnaire. As such, snowball sampling cannot guarantee representativeness, meaning to say that we cannot conclusively say that our findings would apply to everybody within the target population. However, non-representativeness does not concern us too much since our study is exploratory and merely wishes to establish if exercise does have any influence on self-esteem regardless the mechanism involved or how it does affect self-esteem.

The Questions

1. Age – confirm that they are in our target population of young adults aged between age 20-25, those who did not fall under this criteria would be excluded from the survey. (Carnas, 2014)
2. Gender –Gender is one of our extraneous variables. Research has shown that different genders have different responses to exercise. As such, we want to analyse and compare the results of males and females, to see if the difference in gender affects self esteem. (McKelvie, 2005), (Ivey, et al., 2000).

 

As discussed under the previous section,(Conceptualisation & Operationalisation), these questions were targeted at eliminating participants which did not fall under the definition of our Independent Variable. A specific rationale for the questions is as follows:

3. Have you been exercising regularly in your leisure time in the last 4 weeks? Yes/No – This helped exclude participants that did not exercise regularly. (Medicine & Science in Sports & Exercise, 2011)

4. How many sessions of exercise do you do in one week?
This helped to exclude the participants that exercised less than once a week .(Medicine & Science in Sports & Exercise, 2011),  (June, 2015)

5. What is the total number of minutes of exercise you do in one week? (please enter your numerical answer in minutes)
This data was the data that we intended to use directly as our independent variable data if the participants qualified through questions 3 and 4.

 

6. On a rate perceived exertion (RPE) scale from 1-10, how much effort do you usually exert during your above mentioned exercise? Also as mentioned in the previous section, through our Literature Review we had established that intensity of exercise affected self-esteem. (Sonstroem, 1984). It aimed at excluding participants that do not exercise at moderate or vigorous intensity.

 

Through our research, we realised that many other factors also play a role in affecting an individual’s self-esteem. It was not possible for us to eliminate all the extraneous factors and it would have been inappropriate to disregard these variables. Hence, we decided to consider them in the survey to evaluate possible relationships during our data analysis phase. A detailed breakdown for each of the questions is as follows:

7. Of the following, which form of exercise do you engage the most? Aerobic, anaerobic, mixed
This enabled us to take into account type of exercise affecting exercise (Sonstroem, 1984), (McKelvie, 2005). Some research even states that aerobic exercise can help improve self esteem better than anaerobic exercise. (Scully, Kremer, Meade, Graham & Dudgeon, 1998) As such, we would like to see if this extraneous factor has affected our data results. 

8.Do you exercise alone or with others?
Research has shown that presence of a peer, or running in a group, results in greater liking of exercise, hence can make them feel better about themselves. (Carnas, 2014), (Dembeck, 2011).  This may affect the self esteem of the individual.

9. What is your greatest motivation for you exercising? The reason for exercising is an extraneous variable that also needs to be taken into account, as this may affect the results of an individual’s self esteem. (Carnas, 2014), (Dembeck, 2011), (Basich,2006).

 

10. Rosenberg Scale
Once again, as mentioned under the section, Conceptualisation & Operationalisation, the Rosenberg Scale Proven to be an accurate scale to measure self esteem. Research has shown that it high reliability and validity.  (Rosenberg, 1965)

 

 

Our Complete Survey:

Ethical Considerations

1. Other than age and gender, the questionnaire did not collect any other identity markers to assure the respondents of their anonymity. This is especially important for us as this establishes mutual trust that in ensuring their anonymity, respondents can answer the questions as honestly and accurately as possible.
2. Due to the sensitivity nature of the topic Self-esteem, the purpose of the questionnaire was emphasized that our group merely wishes to investigate if there is any association between exercise and self-esteem and there was no malice in any of the questions posed. An email address was also provided should any of the respondents really have any concerns to be addressed with us.
3. It was re-emphasized in the questionnaire that respondents could simply discontinue answering the questionnaire should they feel uncomfortable to do so.
4. Lay terminology was used so as not to compromise the respondent’s comprehension of the questionnaire which also would influence the quality of respondent’s answer. Elements of the questionnaire such as visual aids (ie. tables) and even the mode of answering were intended to facilitate a smooth navigation through the questionnaire by the respondent.
5. Due to the mode of data collection, there could not have been any interaction between us the researcher and respondent that would influence the answers to the questionnaire. In addition, initiation of communication is only one way; from respondent to interviewer. Even with the data collected, there is no way for the interviewer to trace the respondent. These ensure the anonymity of the respondents.