Question

 

Postpartum Depression

In this assignment we will look at women with birth depression and how estrogen treatment can affect depression. The data is contained in the data depression footprint.

Prior to treatment, all women were assessed after the Edinburgh Postnatal Depression Scale (EPDS). High EPDS scores indicate severe depression. Patients were randomly assigned to either receive placebo (group = 0, n = 18) or estrogen (group = 1, n = 22) as treatment. The EPDS score was again measured after 3 months of treatment.

In this task you should choose whether to perform a regular two-sample T-test / paired t-test, or if you should also perform a non-parametric test (Wilcoxon-Mann-Whitney test / Wilcoxon pairing test) when we To check if treatment has an effect. To determine this, check the assumptions that are the basis for performing a t-test. For a two-sample t-test, make sure that the two samples are normally distributed and for a paired t-test, be sure to check that the difference between the paired data is normally distributed. Do not use log-transformation in this task, even if it is possible.

1- Give a descriptive description of the variables EPDS prior to treatment initiated for each of the treatment groups (placebo and estrogen). Write a brief summary of what you find. Is there a significant difference between the two treatment groups in EPDS score before receiving treatment?

2- Is there a significant difference between the two treatment groups in 3 months of treatment?

3- Just look at those who have received estrogen as treatment. Have they had a significant change in EPDS score from before treatment to after 3 months of treatment?
4- Now look at those who have received a placebo. Have they had a significant change in EPDS score from before treatment to after 3 months of treatment?
5- Check the change from before treatment to after 3 months of treatment for the placebo and estrogen groups. Perform a test and check whether estrogen has had an effect on the EPDS score.

SPPS FILE

BEFORE TREATMENT

18
22
17
15
20
27
28
25
16
26
19
22
16
21
20
22
21
25
21
27
15
24
15
17
20
18
28
21
18
27
19
20
21
23
24
25
28
23
22
23

AFTER TREATMENT

21
24
23
14
17
20
20
28
20
28
14
26
18
22
13
23
18
26
10
13
15
15
17
21
8
14
9
26
9
12
9
9
11
27
10
11
10
13
8
11

GROUP

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

Answer 1

Result:

The difference between before and after is calculated as variable change. The normality test for before score for two groups shows that data follows normal distribution. The normality test for after score for two groups shows that data follows normal distribution. The normality test for difference of score for two groups shows that data does not follows normal distribution.

Tests of Normality
	group	Kolmogorov-Smirnova			Shapiro-Wilk
		Statistic	df	Sig.	Statistic	df	Sig.
before	placebo	0.133	18	0.200*	0.958	18	0.570
	estrogen	0.093	22	0.200*	0.963	22	0.563
after	placebo	0.097	18	0.200*	0.958	18	0.563
	estrogen	0.197	22	0.027	0.801	22	0.001
change	placebo	0.230	18	0.013	0.912	18	0.094
	estrogen	0.245	22	0.001	0.879	22	0.012
*. This is a lower bound of the true significance.
a. Lilliefors Significance Correction

Give a descriptive description of the variables EPDS prior to treatment initiated for each of the treatment groups (placebo and estrogen). Write a brief summary of what you find. Is there a significant difference between the two treatment groups in EPDS score before receiving treatment?

Group Statistics
	group	N	Mean	Std. Deviation	Std. Error Mean
before	placebo	18	21.11	3.924	0.925
	estrogen	22	21.77	3.878	0.827

Independent Samples Test
		Levene's Test for Equality of Variances		t-test for Equality of Means
		F	Sig.	t	df	Sig. (2-tailed)	Mean Difference	Std. Error Difference	95% Confidence Interval of the Difference
									Lower	Upper
before	Equal variances assumed	0.000	0.985	-0.534	38	0.596	-0.662	1.239	-3.170	1.847
	Equal variances not assumed			-0.533	36.275	0.597	-0.662	1.241	-3.177	1.854

There is no significant difference between the two treatment groups in EPDS score before receiving treatment, t=-0.534, P=0.596.

Is there a significant difference between the two treatment groups in 3 months of treatment?

Group Statistics
	group	N	Mean	Std. Deviation	Std. Error Mean
after	placebo	18	20.83	4.630	1.091
	estrogen	22	13.09	5.380	1.147

Independent Samples Test
		Levene's Test for Equality of Variances		t-test for Equality of Means
		F	Sig.	t	df	Sig. (2-tailed)	Mean Difference	Std. Error Difference	95% Confidence Interval of the Difference
									Lower	Upper
after	Equal variances assumed	0.048	0.828	4.816	38	0.000	7.742	1.608	4.488	10.997
	Equal variances not assumed			4.890	37.882	0.000	7.742	1.583	4.537	10.948

There is a significant difference between the two treatment groups in EPDS score after receiving treatment, t=4.816, P=0.000.

Just look at those who have received estrogen as treatment. Have they had a significant change in EPDS score from before treatment to after 3 months of treatment?

Paired Samples Statisticsa
		Mean	N	Std. Deviation	Std. Error Mean
Pair 1	before	21.77	22	3.878	0.827
	after	13.09	22	5.380	1.147
a. group = estrogen

Paired Samples Testa
		Paired Differences					t	df	Sig. (2-tailed)
		Mean	Std. Deviation	Std. Error Mean	95% Confidence Interval of the Difference
					Lower	Upper
Pair 1	before - after	8.682	7.253	1.546	5.466	11.898	5.614	21	0.000
a. group = estrogen

There is a significant change in EPDS score from before treatment to after 3 months of treatment, t=5.614, P=0.000.

Now look at those who have received a placebo. Have they had a significant change in EPDS score from before treatment to after 3 months of treatment?

Paired Samples Statisticsa
		Mean	N	Std. Deviation	Std. Error Mean
Pair 1	before	21.11	18	3.924	0.925
	after	20.83	18	4.630	1.091
a. group = placebo

Paired Samples Testa
		Paired Differences					t	df	Sig. (2-tailed)
		Mean	Std. Deviation	Std. Error Mean	95% Confidence Interval of the Difference
					Lower	Upper
Pair 1	before - after	0.278	4.240	0.999	-1.831	2.386	0.278	17	0.784
a. group = placebo

There is no significant change in EPDS score from before treatment to after 3 months of treatment, t=0.278, P=0.784.

Check the change from before treatment to after 3 months of treatment for the placebo and estrogen groups. Perform a test and check whether estrogen has had an effect on the EPDS score.

Group Statistics
	group	N	Mean	Std. Deviation	Std. Error Mean
change	placebo	18	0.2778	4.23994	0.99936
	estrogen	22	8.6818	7.25315	1.54638

Independent Samples Test
		Levene's Test for Equality of Variances		t-test for Equality of Means
		F	Sig.	t	df	Sig. (2-tailed)	Mean Difference	Std. Error Difference	95% Confidence Interval of the Difference
									Lower	Upper
change	Equal variances assumed	3.879	0.056	-4.340	38	0.000	-8.40404	1.93624	-12.32375	-4.48433
	Equal variances not assumed			-4.564	34.723	0.000	-8.40404	1.84120	-12.14294	-4.66514

The change from before treatment to after 3 months of treatment for the placebo and estrogen groups is significant, t=-4.340, P=0.000.