📝 Final-year thesis: social media and academic performance

Research question: Is social-media addiction related to CGPA, and does study engagement matter more? Answer: The 5-item social-media-addiction scale had excellent internal consistency, Cronbach's α = .91. The 4-item study-engagement scale had good internal consistency, Cronbach's α = .87. A multiple linear regression showed that the predictors together explained 34.9% of the variance in CGPA, F(3, 156) = 27.89, p < .001. Significant predictors were Social-media addiction score (β = −.25, p = .020), Study engagement score (β = .37, p < .001); daily social-media hours did not add significantly (p = .273).

Open this example in StatLab   Dataset and codebook (CSV, Excel, .sav)

What is the situation?

The classic final-year project: 160 (fictional) university students answered a questionnaire with a 5-item social-media-addiction scale (sm1–sm5), a 4-item study-engagement scale (st1–st4), daily social-media hours and their CGPA. The data are synthetic — produced by a seeded random process for teaching; they describe no real people or organisations.

What is in the dataset?

160 cases.

How is it analysed, step by step?

1. Descriptives — Chapter 4 opens with a profile of the respondents. (guide)
   Daily social-media hours: M = 3.57, SD = 1.52, range 0.5–7.8, skewness −0.12
   CGPA: M = 3.10, SD = 0.39, range 2.05–4, skewness −0.23
2. Frequencies — Demographic breakdown. (guide)
   Gender: Male 89 (55.6%), Female 71 (44.4%)
   Year of study: Year 1 44 (27.5%), Year 2 42 (26.3%), Year 3 42 (26.3%), Year 4 32 (20.0%)
3. Reliability (Cronbach's alpha) — Reliability of the social-media scale. (guide)
   Cronbach's α = .911 (5 items, n = 160); corrected item-total correlations .75–.81.
4. Reliability (Cronbach's alpha) — Reliability of the study-engagement scale. (guide)
   Cronbach's α = .865 (4 items, n = 160); corrected item-total correlations .68–.73.
5. Compute variable — Build the scale scores. (guide)
   New variable sm_addiction = MEAN(sm1, sm2, sm3, sm4, sm5).
6. Compute variable (guide)
   New variable study_eng = MEAN(st1, st2, st3, st4).
7. Correlations — Relationships between the study variables (Research Question 1). (guide)
   Social-media addiction score – Study engagement score: r = −.35, p < .001
   Social-media addiction score – Daily social-media hours: r = .79, p < .001
   Social-media addiction score – CGPA: r = −.47, p < .001
   Study engagement score – Daily social-media hours: r = −.31, p < .001
   Study engagement score – CGPA: r = .49, p < .001
   Daily social-media hours – CGPA: r = −.43, p < .001
8. Linear regression — Which variables predict CGPA together (Research Question 2). (guide)
   R² = .349, adjusted R² = .337, F(3, 156) = 27.89, p < .001, Durbin-Watson = 1.89.
   (Constant): B = 2.975, SE = .155, t = 19.14, p < .001
   Social-media addiction score: B = −.105, SE = .045, β = −.25, t = −2.34, p = .020, VIF = 2.75
   Study engagement score: B = .161, SE = .030, β = .37, t = 5.35, p < .001, VIF = 1.14
   Daily social-media hours: B = −.030, SE = .027, β = −.12, t = −1.10, p = .273, VIF = 2.66

How would I write it up (APA style)?

The 5-item social-media-addiction scale had excellent internal consistency, Cronbach's α = .91. The 4-item study-engagement scale had good internal consistency, Cronbach's α = .87. A multiple linear regression showed that the predictors together explained 34.9% of the variance in CGPA, F(3, 156) = 27.89, p < .001. Significant predictors were Social-media addiction score (β = −.25, p = .020), Study engagement score (β = .37, p < .001); daily social-media hours did not add significantly (p = .273).

Every number on this page was computed by StatLab's statistics engine from this dataset when the page was built, and the same analyses are part of the reference validation against scipy/statsmodels (433 reference test cases, 36,554 values checked against scipy/statsmodels/numpy, 0 failures). “Open this example” loads the data and runs these steps live in your browser.