1.
Non-parametric tests.
Kruskal-Wallis H-test
When the assumptions of the One-way ANOVA are not met, it is important to use the Kruskal- Wallis H-test to compare means amongst 2 or more groups. We would use the Kruskal-Wallis test to compare outcomes of a continuous response variable on levels of the dependent variable where the question of interest is to test whether there is a statistically significant difference in observations across the levels. The null hypothesis is stated that there is no statistical difference across the levels. The assumptions of the test are; - independence of observations, there are two or more levels of the independent variable and the dependent variable is on a continuous or ordinal scale.
Sign test.
If we …show more content…
When testing for independence using the chi-square test, the expected frequencies of each cell should be greater or equal to 5. We use the test to address the question of whether two categorical variables are related.
Simple random sample, to test for homogeneity across the categories of a categorical variable drawn from two different populations, we use the chi-square test of homogeneity with an assumption that the samples are simple random samples from each population. The research question is to test whether the counts across the categories of the variable from the two populations are equal.
3.
Assumptions are made to support the fundamental procedures of a process making. Many times the statistical assumptions are meant to validate the statistical methods used in data analysis, this method of justifying a process is very punishing. Therefore, to lighten the weight in an analytic process, we would like to believe that there exist requirements that are different from the assumptions and which can easily be …show more content…
Chi-Square suspect.
8.
20% of the cells have expected counts that are less than 5.
9.
Fisher 's test of Independence.
Hypothesis
H0: The proportion of smokers are the same at the two levels of hxdepression.
Ha: The counts of smokers differs across the levels of hxdepression.
Output.
Contingency table results:
Rows: hxdepression
Columns: smoker 1 2 Total
No 4 9 13
Yes 11 49 60
Total 15 58 73
Fisher 's exact test:
P-value = 0.4477
The result is p= 0.448, presenting enough evidence supporting the null hypothesis, the null hypothesis is retained. There is no significant difference in smoking proportions between those who have had depression issues and those who never have.
10.
a.
Males and females work in different departments; it would be interesting to conduct a research where we ask both genders the department of the selected they would prefer to work in.
b.
Does the proportion of both genders differ across the departments?
c.
There are two variables; - Gender with two levels, male or Female and Departments with say 4 levels.
d.
Statistical significance would imply a rejection of the null hypothesis of homogeneity; we would conclude that proportion differs across the
Non-parametric tests.
Kruskal-Wallis H-test
When the assumptions of the One-way ANOVA are not met, it is important to use the Kruskal- Wallis H-test to compare means amongst 2 or more groups. We would use the Kruskal-Wallis test to compare outcomes of a continuous response variable on levels of the dependent variable where the question of interest is to test whether there is a statistically significant difference in observations across the levels. The null hypothesis is stated that there is no statistical difference across the levels. The assumptions of the test are; - independence of observations, there are two or more levels of the independent variable and the dependent variable is on a continuous or ordinal scale.
Sign test.
If we …show more content…
When testing for independence using the chi-square test, the expected frequencies of each cell should be greater or equal to 5. We use the test to address the question of whether two categorical variables are related.
Simple random sample, to test for homogeneity across the categories of a categorical variable drawn from two different populations, we use the chi-square test of homogeneity with an assumption that the samples are simple random samples from each population. The research question is to test whether the counts across the categories of the variable from the two populations are equal.
3.
Assumptions are made to support the fundamental procedures of a process making. Many times the statistical assumptions are meant to validate the statistical methods used in data analysis, this method of justifying a process is very punishing. Therefore, to lighten the weight in an analytic process, we would like to believe that there exist requirements that are different from the assumptions and which can easily be …show more content…
Chi-Square suspect.
8.
20% of the cells have expected counts that are less than 5.
9.
Fisher 's test of Independence.
Hypothesis
H0: The proportion of smokers are the same at the two levels of hxdepression.
Ha: The counts of smokers differs across the levels of hxdepression.
Output.
Contingency table results:
Rows: hxdepression
Columns: smoker 1 2 Total
No 4 9 13
Yes 11 49 60
Total 15 58 73
Fisher 's exact test:
P-value = 0.4477
The result is p= 0.448, presenting enough evidence supporting the null hypothesis, the null hypothesis is retained. There is no significant difference in smoking proportions between those who have had depression issues and those who never have.
10.
a.
Males and females work in different departments; it would be interesting to conduct a research where we ask both genders the department of the selected they would prefer to work in.
b.
Does the proportion of both genders differ across the departments?
c.
There are two variables; - Gender with two levels, male or Female and Departments with say 4 levels.
d.
Statistical significance would imply a rejection of the null hypothesis of homogeneity; we would conclude that proportion differs across the