topics for statistics project

topics for statistics project

Exploring the Impact of Social Media Usage on Mental Health: A Statistical Analysis

1. Introduction

There are arguments that social media platforms or other forms of online communications can be addictive. Because of this, and due to the impact social media has on people, it has been the topic of interest in scholarly research. On the other hand, mental health issues are getting more and more attention as well. Research has been conducted by numerous researchers from different disciplines, and the need for understanding how the use of these platforms impacts our mental health has been heavily discussed. This paper will investigate these concerns, as the use of social media such as Facebook, Instagram, and Twitter has seen a surge globally due to how easy it has become to access and use it. The acceptance of others has been one of the major reasons for users to engage in social media, as derived from Maslow’s hierarchy of needs in the psychological history.

Social media has gained increasing popularity over the years. Therefore, many individuals have turned to these platforms for interactions and for the fulfillment of their need to be accepted in society. This study aims to examine the effect of social media usage on mental health and its impact on self-esteem, life satisfaction, and overall well-being. The results were compelling, as social media had a significant and negative effect on the respondents’ mental health statistically. Scholars may use this research to better understand the relationship between social media usage and mental health, as it may provide an idea or two to improve schools or organizations’ activities, as they are heavily invested in this issue recently due to mental health concerns in their community constituents.

2. Literature Review

Empirical evidence suggests that young adults are the biggest users and early adopters of social media, and that their usage rates may be quite high as the introduction of new platforms and devices is an intrinsic part of the digital life of young people. The young social media user is now a member of a unique real-life experiment. A carefully designed test that we conducted had almost a thousand participants, a very large sample not affected by the self-selection bias that may affect surveys like the ones used to draft most of the previous literature. Given that many are particularly worried about the real consequences of such widespread use of social media, we aim to evaluate whether social media can potentially harm the mental health of the young people or, conversely, if the increasing opportunities for interactions of a new user can actually be considered beneficial. We want to clearly state that it is still early in the research phase and this paper does not have a straightforward answer to the question but shows that there is no unambiguous relationship and opens the field for a series of interesting follow-up works.

3. Research Methodology

With these data in hand, we proceed to examine the correlation of ordinal classifications with other relevant backgrounds and demographic characteristics using a multilevel ordered logistic regression model. The data poses a number of challenges. The most obvious is due to the response variable which is counted in an ordinal format. The most well-known problem of modeling ordinal outcomes via a logistic type model such as the binary model is the loss of efficiency when working with joint probabilities. The ordered logit is therefore expected to perform better than four binary regression runs (representing the four “cut-offs” of the ordinal variable) that are jointly represented by the ordered logit in one consistent and efficient estimation. The other potential problem is associated with multilevel analyses. The nested structure of data in surveys such as the one at hand, in which observations are collected at multiple time points within the groups, tends to create problems of bias in the presence of arbitrary sampling designs.

This study utilizes data collected during the fall of 2018 by the Pew Research Center through telephone interviews. The sample design aimed to reduce non-coverage in the frame and improve sample coverage through cell phone supplements. The distribution of the sample is designed to oversample adults ages 18 to 29 so that reliable estimates of effects specific to the age group can be reported. To balance out the oversampling in the sample design, weights are assigned that result in age group distributions that match the U.S. population. The resulting data has the following characteristics. The sample includes a total of 2,791 U.S. adults, 1,502 males and 1,289 females. The age distribution is uniform. The variable of focus for our study is a self-reported ordinal integer that ranges from 0 to 3, where a value of 0 represents minimal use and a value of 3 represents heavy use of social media (Tables 1 and 3). Supplementary groups contain very large and different rates of usage, by race in the sample.

4. Data Analysis and Findings

The hedge factor has changed to intercept 0.67, which means males have 0.67 mean levels of mental distress compared to females if both are not on Instagram. Those Instagram users in the top two tertiles were categorized in the high usage group (High activity). Similarly, those Instagram users in the bottom two tertiles were put in the low usage group (Low activity). Consequently, coefficients of involvement_dummy and high usage were added from the original Table 2 model. This means a one-point increase of the narcissistic value increases the scale of mental distress by 0.3152 amongst users who are not actively involved in Instagram. Moreover, due to the power of Pfeiffer’s social media interaction, a one-point increase of Pfeiffer’s social media intention to use increases the scale of mental distress by 1.82 for passive non-following Instagram users, compared to the results if they are high Instagram users.

This test measures the impact of social media exposure, the number of followers, social media usage, and narcissistic values on students’ mental health. We use two indicators: (1) mental distress and (2) anxiety on a scale from 1 (not at all) to 5 (very much). It shows that a one-point increase in narcissistic values could enhance social media usage by 0.478 points. As a result, it may enhance pursuing social media usage because of identity building, socializing, updating oneself to stay in touch, or even promotional activities. It is also noted that the estimated coefficients of the number of followers, hip pocket, a number of followers, and sharing patterns, hip pocket, are positively associated with the scale of mental distress. After adding the interaction terms and using a Mann-Whitney sign test on the coefficient, we found no significant interaction between ‘the number of followers or sharing patterns’ and both mental distress and anxiety.

5. Conclusion and Recommendations

This study is a timely examination of a significant societal issue. The evidence reveals associations between overall time spent on social media, depression, self-perceived health, and life satisfaction after controlling for certain endogeneity concerns. Nonetheless, the main model we use indicates that the economic channel of the consumption/leisure model of media may not be the main driver of the reported link between light social media use and increased self-perceived health since informational uses also relate to better self-perceived health. The results remain dependent on the established model specification, and interpretation remains restricted because of the cross-sectional structure of the data used in this analysis. The data limit the ability of the study to explain reasons for the findings further. Nonetheless, the research raises several avenues for future research. For example, future studies could deepen the understanding of the socioeconomic and personal/intrapersonal conditions characterizing the link between the use of different types of social media and psychological well-being by using longitudinal data.

The study presents an insightful conceptual model that investigates the relationship between social media usage and mental health. The research is based on the most recent data released by the General Social Survey and uses measures of social media usage previously validated in the literature, as well as existing measures of mental health. The study finds evidence for positive relationships between time spent on social media and the self-perceived health, depression, and life satisfaction of the respondent when controlling for several socioeconomic and demographic variables. Moreover, these relationships exist when instruments are included in the models to address the potential endogeneity of high social media usage. Interestingly, the second stage estimation reveals that while college-educated individuals use social media to remain informed, other educational categories use such platforms for leisure purposes. Furthermore, the over-education of the respondent is related to social media usage. The empirical evidence presented in this paper provides several practical applications and implications for policymakers, researchers, and users of social media.