Question

ID	Affiliation	Location	Education	Confidence
1	1	3	0	72
2	1	3	5	65
3	0	4	5	66
4	0	1	4	78
5	0	3	1	81
6	1	2	5	81
7	1	1	2	83
8	1	3	3	74
9	0	4	0	78
10	0	2	2	85
11	0	1	1	85
12	1	3	5	69
13	1	2	0	69
14	1	3	2	79
15	1	4	1	82
16	1	1	5	74
17	0	3	0	85
18	0	4	0	68

In the previous item, we used the Mann-Whitney test rather than an independent t-test. Why might we Mann-Whitney rather than the t-test?

Original question- A sample of nurses with affiliation to private hospitals (affiliation = 0) and to university hospitals (affiliation = 1) was asked to rate their confidence in making the right decisions based on their level of ongoing inservice professional development. Use a Mann-Whitney U-test to determine if the distribution of confidence in each group is the same. Be sure to always write the null and alternate hypotheses, so that the decision is made in the correct direction. Also, conduct all as two-tailed tests at α = 0.05.

Answer 1

In the previous item, we used the Mann-Whitney test rather than an independent t-test. Why might we Mann-Whitney rather than the t-test?

The measurement level of confidence is in the level of ordinal. For t-test level of measurement should be atleast in the level of interval. Therefore Mann-Whitney test used.

Original question- A sample of nurses with affiliation to private hospitals (affiliation = 0) and to university hospitals (affiliation = 1) was asked to rate their confidence in making the right decisions based on their level of ongoing inservice professional development. Use a Mann-Whitney U-test to determine if the distribution of confidence in each group is the same. Be sure to always write the null and alternate hypotheses, so that the decision is made in the correct direction. Also, conduct all as two-tailed tests at α = 0.05.

Ho: there is no difference in the distribution of confidence in not affiliated and affiliated groups

H1: there is a difference in the distribution of confidence in not affiliated and affiliated groups

Calculated test statistic z=0.98, P=0.326 which is > 0.05 level of confidence. Ho is not rejected.

We conclude that the distribution of confidence in each group is the same.

Wilcoxon - Mann/Whitney Test
	n	sum of ranks
	8	87.5	Affiliation 0
	10	83.5	Affiliation 1
	18	171	total
		76.00	expected value
		11.21	standard deviation
		0.98	z, corrected for ties
		.3264	p-value (two-tailed)
No.	Label	Data	Rank
1	Affiliation 0	66	2
2	Affiliation 0	78	9.5
3	Affiliation 0	81	12.5
4	Affiliation 0	78	9.5
5	Affiliation 0	85	17
6	Affiliation 0	85	17
7	Affiliation 0	85	17
8	Affiliation 0	68	3
9	Affiliation 1	72	6
10	Affiliation 1	65	1
11	Affiliation 1	81	12.5
12	Affiliation 1	83	15
13	Affiliation 1	74	7.5
14	Affiliation 1	69	4.5
15	Affiliation 1	69	4.5
16	Affiliation 1	79	11
17	Affiliation 1	82	14
18	Affiliation 1	74	7.5