Jeralyn Jones

April 15, 2018

Research question

Does expression of affection on social media make couples feel happier in their relationship?

Data collection

Couples filled out a questionnaire on Facebook to the users who have the relationship status as "In a relationship”, “Engaged to” or “Married to”. The technique of simple random sampling is used to select a random sample of 30 participants.

Data

The data obtained regarding the happiness level and affection is given below:

Happiness_Don’t flaunt / Happiness_Flaunt
-3 / -2
-3 / -3
-3 / -3
0 / 1
3 / -3
3 / -3
3 / -3
2 / 3
-3 / 1
3 / 3
2
3
1
-3
3
-3
-3
0
3
-3

Hypothesis testing

Assumptions

1)For the application of parametric test it is required that the sample is normally distributed.

From the histogram given above, it is evident that it is approximately bell-shaped indicating that the distribution is normal. Hence the assumption of normal distribution for application of the parametric test is satisfied.

2)With the help of Levene test it is important to check the assumption of equality of variance.

Test and CI for Two Variances: Happiness_Don’t flaunt, Happiness_Flaunt

Method

Null hypothesis σ(Happiness_Don’t flaunt) / σ(Happiness_Flaunt) = 1

Alternative hypothesis σ(Happiness_Don’t flaunt) / σ(Happiness_Flaunt) ≠ 1

Significance level α = 0.05

Statistics

95% CI for

Variable N StDev Variance StDevs

Happiness_Don’t flaunt 10 2.898 8.400 (2.402, 4.349)

Happiness_Flaunt 20 2.645 6.997 (2.328, 3.332)

Ratio of standard deviations = 1.096

Ratio of variances = 1.200

95% Confidence Intervals

CI for

CI for StDev Variance

Method Ratio Ratio

Bonett (0.851, 1.640) (0.725, 2.691)

Levene (0.754, 1.530) (0.569, 2.342)

Tests

Test

Method DF1 DF2 Statistic P-Value

Bonett — — — 0.470

Levene 1 28 0.13 0.725

Null Hypothesis, ho: there is no significant difference in the variance of two groups. The alternative hypothesis, h1: there is a significant difference in the variance of two groups. With (F=0.13, p>5

%), I fail to reject the null hypothesis and conclude thatthere is no significant difference in the variance of two groups.

Hypothesis

Step 1: hypothesis

The null hypothesis, Ho: there is no significant difference in the mean happiness level between those who don't flaunt and those who flaunt their love on social media.

The alternative hypothesis, h1: the mean happiness level for those who flaunt their love and social media is greater than those who don't flaunt.

Step 2: level of significance

The level of significance Alpha is taken as 5%.This is an indication that I am 95% confident about my results.

Step 3:test statistic

I use T-test for independent samples with equal variances to test my hypothesis.

Two-Sample T-Test and CI: Happiness_Don’t flaunt, Happiness_Flaunt

Two-sample T for Happiness_Don’t flaunt vs Happiness_Flaunt

N Mean StDev SE Mean

Happiness_Don’t flaunt 10 0.20 2.90 0.92

Happiness_Flaunt 20 -0.45 2.65 0.59

Difference = μ (Happiness_Don’t flaunt) - μ (Happiness_Flaunt)

Estimate for difference: 0.65

95% upper bound for difference: 2.45

T-Test of difference = 0 (vs <): T-Value = 0.61 P-Value = 0.728 DF = 28

Both use Pooled StDev = 2.7291

T= 0.61

Step 4: P-value

P value is the probability that null hypothesis is true.P-Value = 0.728

Step 5: Decision rule

The decision rule is to reject the null hypothesis if P-value is less than Alpha, 5%. Else if P-value is greater than 5%, I fail to reject the null hypothesis.

Step 6: conclusion

Since P value is greater than 5%, I fail to reject the null hypothesis and conclude thatthere is no significant difference in the mean happiness level between those who don't flaunt and those who flaunt their love on social media.

Conclusion

With (t=0.61, p>5%), I can say that there is no significant difference in the mean happiness level between those who don't flaunt and those who flaunt their love on social media.