Step1: Objective

The study aimed to investigate factors influencing daytime sleepiness (totsas) in university
staff. It examined how gender, age, physical fitness (fitrate), and depression scores (depress)
contribute to predicting sleepiness.

Step 2: Methodology

A multiple regression analysis was conducted to determine the unique contribution of each
predictor variable to daytime sleepiness while controlling for others.

Step 3: Assumption Tests

Normality of residuals was confirmed through a histogram (Figure 1) and Shapiro-Wilk test
(W=0.942, p< 0.001). Homoscedasticity was supported by a random scatter in residuals
versus predicted values (Figure 2). Linearity and multicollinearity assumptions were also
satisfied.

1. Normality: Normality of residuals was confirmed through a histogram

Figure 1

Histogram of Residuals Showing Normality of Residuals

2. Multicollinearity was assessed via Pearson correlations, Correlations among predictors

were below (0.8), indicating no significant multicollinearity concerns in the model.

Table 1
Correlation Matrix Assessing Multicollinearity Among Predictor Variables

Predictor Gender Age Physical Fitness HADS Depression

Gender — .13 -.00 -.05
Age .13 — -.15 .04
Physical Fitness -.00 -.15 — -.23*
HADS Depression -.05 .04 -.23* —

Note. *p < .05.

3. Homoscedasticity: The scatterplot of residuals versus predicted values demonstrates

homoscedasticity, as the residuals appear randomly distributed with no clear pattern or
funnel shape.

Figure 2
Scatterplot of Residuals vs Predicted Values Showing Homoscedasticity

Step 4: Report Findings

Report Findings

The regression analysis indicated a significant model, F (4, 89) = 8.10, p < .001, with an R² of
.267, suggesting that 26.7% of the variance in totsas was explained by the predictors.

Table 2

Regression Coefficients for Predictors of Daytime Sleepiness

Variable B β SE
Constant 25.57 - 5.39
Gender (Male) -4.90** -.25 1.78
Age -0.12 -.17 0.07
Physical Fitness -0.04 -.01 0.57
HADS Depression 1.41*** .41 0.32

R² .267
Note: Gender coded as 1 = male, 2 = female. **p < .01, ***p < .001.

Results

In Step 1 of the hierarchical regression, gender and age were entered as predictors, explaining
10.4% of the variance in daytime sleepiness (R2=.104, F (2,93) =5.37, p=.006). In Step 2,
physical fitness and depression scores were added to the model, significantly increasing the
explained variance to 26.7% (R2=.267, ΔR2=.163, F (2,89) =10.14, p <.001). Depression
(B=1.41, p<.001) was a significant predictor, contributing strongly to the additional variance,
while physical fitness (B=−0.04) was not significant. These results suggest that depression is
a key factor influencing daytime sleepiness, even after accounting for gender and age.

Question2 a Objectives

The objective of the study was to evaluate the impact of time and diet type (high-fiber vs.
low-fiber) on bodyweight over three time points: baseline, mid-treatment (V1), and post-
treatment (V2). The study also aimed to assess whether weight changes differed significantly
between the two diet groups.

Methodology

A repeated measures ANOVA was used to analyze changes in bodyweight over time within
and between diet groups. This method was chosen because it accounts for repeated
measurements of the same individuals, reducing variance from inter-subject variability and
allowing for the evaluation of both within-subject (time) and between-group (diet) effects.

Assumption Test

The assumptions for repeated measures ANOVA were assessed:

1. Normality: Bodyweight data at each time point were approximately normally

distributed.

Table 4
Tests of Normality for Bodyweight Measurements by Diet Group

Variable Diet Group Kolmogorov-Smirnov Shapiro-Wilk

Statistic p Statistic p
Weight Baseline Low-Fiber .211 .072 .905 .112
High-Fiber .179 .200* .946 .460
Weight_V1 Low-Fiber .210 .074 .899 .091
High-Fiber .144 .200* .919 .187
Weight_V2 Low-Fiber .208 .080 .895 .080
High-Fiber .137 .200* .929 .268
Note. p=.200p = .200p=.200 indicates the lower bound of true significance due to the

Lilliefors Significance Correction.

2. Sphericity: Sphericity Mauchly’s test indicated a violation of the sphericity

assumption (𝑝 =. 047). Therefore, the Greenhouse-Geisser correction was applied to
adjust the degrees of freedom for the repeated measures ANOVA. in this where is it
Greenhouse-Geisser correction.

Table 5

Mauchly’s Test of Sphericity for Time

Effec Mauchly’s Approx. df Sig. Epsilo Greenhouse- Huynh- Lower-

t W Chi- n Geisser Feldt bound
Square
Time .797 6.124 2 .047 .831 .909 .500

3. Descriptives Statistics: Bodyweight decreased slightly over time in both groups, with

the high-fiber group maintaining a consistently higher weight at all points.

Table 4

Descriptive Statistics for Bodyweight by Diet Group at Baseline, 4 Weeks (V1), and 8 Weeks
(V2)
 Time Point Diet Group M SD N
Baseline Low-Fibre 82.15 9.43 15
High-Fibre 84.39 11.19 15
(V1) Low-Fibre 81.67 9.78 15
High-Fibre 83.77 11.32 15
(V2) Low-Fibre 81.19 9.89 15
High-Fibre 83.32 11.06 15
Note. M = Mean; SD = Standard Deviation.

Step 4: Report Findings

Table 5

Repeated Measures ANOVA Results for Bodyweight

Effect and Source Type III Sum of df Mean F p Partial

Squares Square η²
Within-Subjects
Time 15.435 1.66, 9.284 6.831 .00 .196
46.55 4
Time × Diet Group 0.086 1.66, 0.052 0.038 .94 .001
46.55 1
Error (Time) 63.266 46.553 1.359 — — —
Between-Subjects
Diet Group 104.544 1 104.544 0.32 .57 .011
6
Error 9147.073 28 326.681 — — —
Pairwise
Comparisons (Time)
 Baseline - V1 — — — — .04 —
*
Baseline - V2 — — — — .00 —
*
V1 - V2 — — — — .05 —
2
Note. p<.05*

The repeated measures ANOVA revealed a significant main effect of time on bodyweight, F
(1.66,46.55) =6.83, p=.004, ηp2=.196, indicating that bodyweight significantly changed over
the three time points (Baseline, V1, and V2). Pairwise comparisons showed significant
reductions in bodyweight from Baseline to V1 (p=.04) and from Baseline to V2 (p=.00), but
the change from V1 to V2 was marginally non-significant (p=.052).

The interaction effect between time and diet group was not significant, F (1.66,46.55) =0.038,
p=.941, ηp2=.001, suggesting that weight changes over time were similar for both the high-
fiber and low-fiber diet groups. Additionally, the main effect of diet group on bodyweight
was also not significant, F (1,28) =0.32, p=.576, ηp2=.011, indicating no significant
difference in overall weight between the two diet groups. These results suggest that while
bodyweight reduced over time, the high-fiber diet did not result in significantly greater
weight loss compared to the low-fiber diet.

Figure 4

Estimated Marginal Means of Bodyweight Across Time for Low-Fiber and High-Fiber Diet
Groups
Question 3: Comparison of Perceived Stress Between Smokers and Non-Smokers

Step 1: Objective

The objective of this analysis was to examine whether there is a significant difference in
perceived stress levels between smokers and non-smokers. The study aimed to determine
whether smoking status influences the perceived stress scores in the participants.

Step 2: Methodology
An independent samples t-test was employed to compare the mean perceived stress levels
between two independent groups: smokers and non-smokers. The independent samples t-test
is appropriate because it allows for comparing the means of two distinct, unrelated groups.
This method helps to identify whether the differences in perceived stress between smokers
and non-smokers are statistically significant.

Descriptive statistics for perceived stress by smoking status were calculated. Additionally,
assumption tests (normality and equality of variances) were performed to verify the
assumptions necessary for the t-test.

Step 3: Assumption Test

Before conducting the t-test, the assumptions were checked:

1. Normality:
The Shapiro-Wilk test was used to assess normality. Given the large sample size (n =
430), the assumption of normality was met as the test supported the distribution of the
data.

Table 7

Tests of Normality for Total Perceived Stress by Smoking Status

Smoking Kolmogorov- Statisti df Sig. Shapiro- Statistic df Sig.

Status Smirnov c Wilk
Smoker .072 84 .200* .989 84 .714 84 .989
(YES)
Non- .076 346 .000 .991 346 .043 346 .043
Smoker
(NO)

Step 4: Report Findings

The descriptive statistics for perceived stress by smoking status

Table 7

Descriptive Statistics for Perceived Stress by Smoking Status

Smoking Status N Mean SD

 Smoker 84 27.08 5.765

Non-Smoker 346 26.65 5.894

Note. SD = Standard Deviation.

The mean perceived stress score for smokers (M = 27.08, SD = 5.765) was slightly higher
than that for non-smokers (M = 26.65, SD = 5.894).

Table 8

Independent Samples t-Test for Perceived Stress

Test t df p (2- Mean Std. Error 95% CI for

tailed) Difference Difference Mean
Difference
Equal 0.611 428 .542 0.436 0.714 [-0.967, 1.839]
variances
assumed
Equal 0.619 128.4 .537 0.436 0.704 [-0.958, 1.829]
variances not 8
assumed

The results of the independent samples t-test are shown in Table 8. The perceived stress levels
of smokers and non-smokers did not differ statistically significantly, t (428) = 0.611, p = .542.
To perform independent sample t-test Assumption checks: Levene’s test (p = .694) met,
Shapiro-Wilk supported normality by sample size.
The key statistics from the t-test, including the mean difference, standard error, and 95%
confidence interval, are presented in Table 8 The results show that the difference in perceived
stress between smokers and non-smokers was not statistically significant (p = .542).