Discussion

There were many possible factors as to why the results differed so greatly from the earlier conducted studies. One of which is the mode of data collection. An experiment has more control over the variables and is able to produce more definitive results. Moreover, there are too many other factors that also affect self-esteem that we came across during the literature review phase of our project, such as family upbringing, social environment, culture, religion and mental health. It was therefore impossible to control all of these extraneous variables and that might have led to our results showing no relationship between the two factors.

This study also lacked the resources to conduct an experiment to work with the variables in a controlled environment. Conducting an experiment on such a scale would also demand commitment from subjects which was not feasible. Thus, a questionnaire was chosen as the most suitable way to conduct the study.

An anonymous questionnaire, only allowed for data collection at the point in time  when the respondent answered to the survey. There was no baseline measured to compare changes in the Rosenberg Self-Esteem Scale Score overtime. There were also limitations to the complexity of the format when constructing the questionnaire which hindered us from effectively controlling the multitude of extraneous variables that have an influence on self-esteem.

The large inclusion criteria of our research question resulted in having no access to a list of all members of the target population. Causing the use of a non-probability sampling method (ie. snowball sampling) which does not allow the generalisation of the findings to all members of the target population. The response rate of the survey was relatively low compared to the target population. Hence, results were not normalised. The conclusion of the findings were only valid for the available sampling frame, the respondents who answered the survey.

Due to the mode at which the questionnaire was distributed, via a link on Whatsapp, some members of the target population were automatically excluded if they did not own any electronic device capable of connecting to the Internet, had no access to the internet or did not own an account on any major social media platform.