Introduction
The relevant hypothesis to be tested is that female participants would have a higher PIUQ score in comparison to male participants. In this regards, the independent variable is the gender while the PIUQ score is the dependent variable. This is because gender tends to influence the average PIUQ score as stated by the hypothesis.
Analysis and Results
The given data corresponds to quantitative data regarding PIUQ score for the two genders. The appropriate measurement scale used is ratio since an absolute zero may be defined for the given values. In order to indicate whether the given variables are independent or dependent, correlation analysis has been performed using SPSS whose relevant output is represented in Table 1.Based on the results of the correlation analysis, it is apparent that the degree of correlation between the data for male and females is statistically insignificant at 5% level of significance considering the fact that requisite p value is 0.62 (Hillier, 2016). This is not surprising since the two genders have difference in preferences and social media behaviour.
Histogram has been used as the appropriate tool to determine the distribution of the given two variables i.e. PIUQ scores of the two genders. These are indicated in Graph 1 and Graph 2 shown in the Appendix. Neither of the above distributions resemble a normal distribution which is bell shaped. Also, as per Central Limit Theorem, a minimum size of 30 observations is required for the given sample to be assumed as normal. Considering that the sample size for the two variables is significantly lower than 30, hence it would be appropriate to conclude that the given variables are not normally distributed. Owing to the non-normal distribution for the two independent variables, the non-parametric test ought to be used (Medhi, 2016).
Descriptive Statistics
The requisite descriptive statistics for the two variables have been computed using SPSS and indicated in Table 2. The various measures of central tendency for the sample PIUQ scores for the two genders. It is evident that the mean and median scores of females is greater than the corresponding values of males. This tends to lend some support to the hypothesis which ought to be tested. However, a potential concern is that the sample size chosen for the two variables is quite small (Eriksson and Kovalainen, 2015).It is evident from the above measures of dispersion that the extent of variation found amongst female respondents PIUQ score is higher than the corresponding values for male respondents. This may be related to a large amount of intra female variation with regards to the impact of social media.
Inferential Statistics
Since there are two independent and non-normal variables whose mean values ought to be compared, hence the appropriate non-parametric test to be deployed is Mann Whitney U Test. The requisite hypotheses are defined as follows (Hair et. al., 2015).
Null Hypothesis (H0): There is no statistically significant difference between average PIUQ score for male and female.
Alternative Hypothesis (H1): The average PIUQ score for females is significantly greater than the average PIUQ score for males.
Level of significance = 5% (0.05)
The requisite output obtained from SPSS is presented in Table 3. The relevant results are presented as follows.
U =40.00, p value =0.193, two tailed
Since the p value derived above exceeds the 5% level of significance assumed for this test, hence the available evidence does not warrant the rejection of null hypothesis. As a result, the alternative hypothesis cannot be accepted (Medhi, 2016). This, it may be concluded that there is no statistically significant difference between average PIUQ score for male and female.
Discussion
The result of the inferential statistics clearly highlights that the average PIUQ of the two genders does not vary in statistically significant manner. In terms of the psychological theory, it would imply that the negative impact of internet does not seem to be gender specific and has similar relevance for both the genders. The current theories tend to reflect that the long term impact of social media is more prominent for females in comparison to males. There have been various studies in the recent years which tend to lend support the theory of females being more impacted by social media than males especially in the context of negative impact (Ducharme, 2019). An example of one such study is a study by University of Essex using 14,000 test subjects (age =14 years) which came to the conclusion that depression related effects arising from social media tend to be more prominent for females in comparison to males (Walton, 2019). The given study does not validate the existing viewpoint and instead indicates that gender is an insignificant variable with regards to extent of adverse influence in regards to social media.
It is imperative to critically analyse the given experimental design in context of the above discussion. Unlike similar studies in the past which have used a very large sample, the given study has used only a small sample of 11 observations for each gender which is clearly not representative of the population (Hair et. al., 2015). For the US, the population is quite diversified and a sample size running in thousands would be required to faithfully capture the American population. Owing to sample not being representative, it is likely that the results obtained from the given study would lack reliability (Hillier, 2016). Also, for the boys and girls which have been included in the sample, there has not been any matching owing to which it is likely that the compared scores for the two participants belonging to different genders may be arising from other factors besides gender (Eriksson and Kovalainen, 2015). Some of these factors may be related to demographics, physical attributes, amount of wealth, level of self-esteem, educational standards etc. It is quite possible that these extraneous factors may be driving the scores which are being compared. As a result, in future studies, it is imperative that these factors must be matched for the respondents from the two genders so that the influence of the gender can be captured in terms of PIUQ score. Additionally, another improvement in the current research design is the need for a much larger sample size to be preferably selected using stratified sampling. This would ensure a representative sample is obtained which would provide reliable results. Further research should focus on understanding the various factors besides gender having a significant impact on scores so that a better understanding of role of gender may be developed (Medhi, 2016).
References
Ducharme, J. (2019) , Social Media Hurts Girls More Than Boys, Retrieved from https://time.com/5650266/social-media-girls-mental-health/
Eriksson, P. and Kovalainen, A. (2015) Quantitative methods in business research. 3rd ed. London: Sage Publications.
Hair, J. F., Wolfinbarger, M., Money, A. H., Samouel, P., and Page, M. J. (2015) Essentials of business research methods. 2nd ed. New York: Routledge.
Hillier, F. (2016) Introduction to Operations Research.6th ed.New York: McGraw Hill Publications.
Medhi, J. (2016) Statistical Methods: An Introductory Text. 4th ed. Sydney: New Age International
Walton, A. (2019), Social Media (Again) Shown To Be Worse For Girls’ Mental Health Than Boys’, Retrieved from https://www.forbes.com/sites/alicegwalton/2019/01/05/social-media-again-shown-to-be-worse-for-girls-mental-health-than-boys/#4d6e40af5057
