I need help Please let me know if Im on the right track I ha

I need help! Please let me know if I’m on the right track. I had to create a research question, and conduct the 5 step Hypothesis testing. Also to create a mock SPSS dataset and run appropriate analysis.

I also have to include a discussion of the statistical analysis that I plan to perform. Why is it the correct test? What is the test statistic from the output that you have generated?

Research Question: To what extent does socioeconomic status (income) impact the number of new deaths by firearm suicide in males ages 15-34?

Step One: Populations

Population 1: set of suicide males socioeconomic impact

Population 2: set of suicide males who did not have socioeconomic impact

Non-Directional Null Hypotheses

There is no correlation between SES and number of new deaths by firearm suicide in males ages 15-34.

Non-Directional Alternative Hypothesis:

There is a correlation between SES and number of new deaths by firearm suicide in males ages 15-34.

Step Two: Characteristics of comparison distribution

Dependent Variable: cumulative incidence of new deaths by firearm suicide

Independent Variable: SES (income)

The variable is ordinal as we can order the SES of the people into various categories such as low, medium and high.

Step Three: Level of significance

In the non-directional hypothesis, a two tailed test was used because there could have been a difference in either direction.

The common level of significance of 0.05 is used.

I will compare the type of SES young males and its impact on the cumulative incidence of new deaths by firearm suicide in each state. I selected three different classes to compare – low income versus medium income versus high income. The cumulative incidence of new suicide deaths by firearm in males for each state will fall into 1 of 3 categories (17 records per group): Low income, medium income, high income. The data are listed below:

Low Income

Medium Income

High Income

17

11

15

19

14

4

14

14

12

16

19

12

8

11

11

12

10

16

4

3

11

10

12

11

9

8

5

12

18

15

13

14

10

3

17

15

14

9

10

9

14

12

13

6

9

8

6

10

9

14

9

Does this make sense???

Low Income	Medium Income	High Income
17	11	15
19	14	4
14	14	12
16	19	12
8	11	11
12	10	16
4	3	11
10	12	11
9	8	5
12	18	15
13	14	10
3	17	15
14	9	10
9	14	12
13	6	9
8	6	10
9	14	9

Solution

I think you are going the right way. But please note the following points -

1) You are having three different classes of income to compare. An F test would be required to test the mean of all three classes of the socio economic deaths and non socio economicc deaths.

2) The test would give the result as whether the mean of the three samples are significantly different from each other.

3) But your null hypothesis says that there is a correlation between SES and number of deaths. So, the result that the means are not significaantly different from each other would not lead to conclusion that there is a correlation between SES and number of deaths.

4) I would suggest that take any two classes at a time and run a one tailed t-test assuming the alternate hypothesis to claim that -

a) the number of deaths due to low income is greater than number of deaths in medium income.

b) the number of deaths due to low income is greater than number of deaths in high income.

c) the number of deaths due to medium income is greater than number of deaths in high income.

5) This would give a better idea of the correlation between the number of deaths and income.

6) An alternative option would be to reduce the number of classes from 3 to 2 and then run a single one tailed t test claiming the alternate hypothesis to be true.