STATISTICAL ANALYSIS FOR THE STUDY “Relationship Between Social Media Use and Self Esteem: The moderating role of social support (Rev. 2)”
Research objectives
This research intends to uncover the connection between self-esteem and social media use, especially the negative effects of social media on individuals of self-esteem and the role of social support in moderating these effects. By investigating the following research questions, this study seeks to shed light on the intricate dynamics of online behaviour and its impact on psychological well-being:
- Is there a relationship between self-esteem and social media use?
- Does social media use predict self-esteem?
- Does social support moderate the relationship between self-esteem and social media use?
Summary of statistical methods to be used per objective
Is there a relationship between self-esteem and social media use?
🔹 Statistical Method: Pearson or Spearman Correlation Coefficient
🔹 Purpose: To determine the strength and direction of the linear relationship between self-esteem and social media use.Does social media use predict self-esteem?
🔹 Statistical Method: Simple Linear Regression
🔹 Purpose: To examine whether levels of social media use significantly predict variations in self-esteem.Does social support moderate the relationship between self-esteem and social media use?
🔹 Statistical Method: Moderation Analysis via Multiple Linear Regression (with interaction term)
🔹 Purpose: To assess whether the relationship between social media use and self-esteem changes depending on the level of social support.
Data
Descriptive Statistics
Socio-demographic Characteristics of respondents
Data shows that from 391 valid samples, the median age is 26 yrs old, where majority are female (60%), heterosexual/straight (85%), college graduates (59%), single(74%) and are employed(59 %).
Table 1. Socio-Demographic Characteristics
|
|
|---|---|
| Characteristic | N = 5471 |
| Age | 26.0 (22.0, 30.0) |
| Sex | |
| Female | 327 (60%) |
| Male | 217 (40%) |
| (Missing) | 3 |
| Gender | |
| Bisexual | 20 (3.7%) |
| Gay | 12 (2.2%) |
| Heterosexual/ Straight | 466 (85%) |
| Lesbian | 1 (0.2%) |
| Prefer not to say | 47 (8.6%) |
| Transgender | 1 (0.2%) |
| Educational Attainment | |
| Bachelor's Degree/College Graduate | 320 (59%) |
| Elementary graduate/High school undergraduate | 5 (0.9%) |
| High School Graduate/College undergraduate | 151 (28%) |
| Post graduate | 67 (12%) |
| Vocational course | 4 (0.7%) |
| Civil status | |
| Divorced | 1 (0.2%) |
| Married | 132 (24%) |
| Separated | 7 (1.3%) |
| Single | 406 (74%) |
| Widower | 1 (0.2%) |
| Employment Status | |
| Employed | 291 (53%) |
| Self-employed | 32 (5.9%) |
| Student | 190 (35%) |
| Unemployed | 34 (6.2%) |
| 1 Median (Q1, Q3); n (%) | |
Rosenberg Self Esteem
The table below shows summary statistics on scores from the Rosenberg Self Esteem Scale Items. The average total score for all respondents is 29, implying that respondents have moderate to average self-esteem.
A sample interpretation guide for the RSES scores
| Score Range | Interpretation |
|---|---|
| 10–15 | Very Low Self-Esteem |
| 16–25 | Low to Moderate Self-Esteem |
| 26–30 | Moderate Self-Esteem (Average) |
| 31–40 | High Self-Esteem |
| Table 4. Summary statistics on Rosenberg Self Esteem Scale items | |
|---|---|
| Characteristic | N = 547 |
| On the whole, I am satisfied with myself. | |
| Median (Q1, Q3) | 3.00 (3.00, 4.00) |
| Mean (SD) | 3.31 (0.59) |
| Min, Max | 1.00, 4.00 |
| At times, I think I am no good at all. | |
| Median (Q1, Q3) | 2.00 (2.00, 3.00) |
| Mean (SD) | 2.35 (0.80) |
| Min, Max | 1.00, 4.00 |
| I feel that I have a number of good qualities. | |
| Median (Q1, Q3) | 3.00 (3.00, 4.00) |
| Mean (SD) | 3.22 (0.59) |
| Min, Max | 1.00, 4.00 |
| I am able to do things as well as most other people. | |
| Median (Q1, Q3) | 3.00 (3.00, 4.00) |
| Mean (SD) | 3.21 (0.59) |
| Min, Max | 1.00, 4.00 |
| I feel I do not have much to be proud of. | |
| Median (Q1, Q3) | 3.00 (2.00, 3.00) |
| Mean (SD) | 2.73 (0.82) |
| Min, Max | 1.00, 4.00 |
| I certainly feel useless at times. | |
| Median (Q1, Q3) | 3.00 (2.00, 3.00) |
| Mean (SD) | 2.68 (0.86) |
| Min, Max | 1.00, 4.00 |
| I feel that Im a person of worth, at least on an equal plane with others. | |
| Median (Q1, Q3) | 3.00 (3.00, 4.00) |
| Mean (SD) | 3.17 (0.60) |
| Min, Max | 1.00, 4.00 |
| I wish I could have more respect for myself. | |
| Median (Q1, Q3) | 2.00 (1.00, 3.00) |
| Mean (SD) | 2.10 (0.90) |
| Min, Max | 1.00, 4.00 |
| All in all, I am inclined to feel that I am a failure. | |
| Median (Q1, Q3) | 3.00 (2.00, 3.00) |
| Mean (SD) | 2.87 (0.87) |
| Min, Max | 1.00, 4.00 |
| I take a positive attitude toward myself. | |
| Median (Q1, Q3) | 3.00 (3.00, 4.00) |
| Mean (SD) | 3.37 (0.63) |
| Min, Max | 1.00, 4.00 |
| Rosenberg Self Esteem Scale Sum | |
| Median (Q1, Q3) | 29.0 (26.0, 32.0) |
| Mean (SD) | 29.0 (4.5) |
| Min, Max | 14.0, 40.0 |
| Rosenberg Self Esteem Scale Mean | |
| Median (Q1, Q3) | 5.27 (4.73, 5.82) |
| Mean (SD) | 5.27 (0.83) |
| Min, Max | 2.55, 7.27 |
Objective 3: Does social support moderate the relationship between self-esteem and social media use?
To test whether social support ( as measured by Multidimensional scale of perceived social support (MSPSS) ) moderated the effect of social media integration on self esteem, we conduct a multiple linear regression including the interaction effect of mspss_overall_mean with smuis_overall_mean on rses_mean.
| Table 11. Self Esteem ~ Social Media Use X Social Support | |
|---|---|
| Linear Model Summary | |
| (1) | |
| (Intercept) | 5.264 (0.035)*** |
| (0.035) | |
| scale(smuis_overall_mean) | -0.124 (0.035)*** |
| (0.035) | |
| scale(mspss_overall_mean) | 0.244 (0.037)*** |
| (0.037) | |
| scale(smuis_overall_mean) × scale(mspss_overall_mean) | 0.039 (0.031) |
| (0.031) | |
| Num.Obs. | 547 |
| R2 | 0.082 |
| R2 Adj. | 0.077 |
| AIC | 1305.5 |
| BIC | 1327.1 |
| Log.Lik. | -647.765 |
| F | 16.119 |
| RMSE | 0.79 |
| RSES overall mean as measure of Self Esteem. | |
| SMUIS overall mean as measure of Social Media Use. | |
| MSPSS overall mean as measure of Perceived Social Support | |
Note: All continuous predictors were standardized (z-scores) using scale(). This means that each unit increase reflects a 1 standard deviation change from the mean. |
|
| Predictor | Estimate (β) | p‑value | Interpretation |
|---|---|---|---|
| Intercept | 5.264 | < 2 × 10⁻¹⁶ | The baseline self‑esteem score is 5.259 when social media integration and perceived support are average—suggesting average well‑being. |
| SMUIS (social media integration) | –0.1239 | 0.000505 | Each 1 SD increase in social media integration corresponds to a 0.1239 SD drop in self‑esteem. This means that reliance on social media routines/emotional engagement may subtly decrease self‑esteem. |
| MSPSS (perceived social support) | +0.245 | 9.23× 10⁻¹¹ | Each 1 SD boost in perceived support links to a 0.245 SD rise in self‑esteem. This implies stronger friendships, family, or partner support increases self-esteem. |
| SMUIS × MSPSS interaction | +0.0392 | 0.205 | Non‑significant moderation—social support doesn’t buffer or moderate the impact of social media integration on self‑esteem. |
A visual interaction plot, showing how the slope of social media use → self-esteem changes across low, mean, and high levels of social support.
Social media apps, use and time spent
Majority of the respondents use Facebook (97%), Messenger (97%),Youtube (81%), Instagram (75%),and TikTok (74%).
Majority of the respondents also use social media apps/sites several times per day (90%) with at least 3-4 hours per day (35%). The site with most time spent on is facebook.