ASSIGNMENT 5 XXXXXXXXXX
Assignment 5: Analysis of Variance
CES 714 Advanced Statistics
Assignment 5 – 20 points Analysis of Variance
Please, conduct Two-Way ANOVA with provided file for Assignment 5. Please, submit your SPSS and Word files when you submit your assignment.
DEFINITION OF ANOVA: 2-Way Factor Analysis
DV = DEPRESSIONsubDASS from DASS Scale4
DEPR ESS is made up of the following individual questions in the DASS scale questionnaire:
, 5, 10, 13, 16, 17, 21C
SIV1 (or Factor 1) = Sex_Gender (level = Male, Female, Trans)
IV2 (or Factor 2) = PWI_health_cat (level = low, medium, high)
1. State the purpose of your study that would be linked to your Two-Way ANOVA research question. Read class notes on how to state the purpose statement (2 points).
ReseaRQ1 = Does levels of depression differ? (Only comparing mean scores across Gender groups)
Ho: There is no significant difference in mean DEPRESSIONsubDASS score between Genders
Ha: There is a significant difference in mean DEPRESSIONsubDASS score between Genders
y Gender (male, female, trans)?
RQ2 = (Only comparing mean scores across Health groups)
Ho: There is no significant difference in mean DEPRESSIONsubDASS score between Health groups
Ha: There is a significant difference in mean DEPRESSIONsubDASS score between Health groups
Does levels of depression differ by Health?
RQ3 = (Determine if mean scores across gender groups differ by health groups)
Ho: There is no significant interaction in means DEPRESSIONsubDASS score between Gender and Health groups
Ha: There is a significant interaction in means DEPRESSIONsubDASS score between Gender and Health groups
2. Describe your sample in a sentence format (write at least3 sentences), include demographics table for this assignment – see example provided in the assignment 5 folder (1 point).
3. State research question and two hypotheses (null and directional) for Two-Way ANOVA. Be careful to use a Two-Way ANOVA research language (3 points).
RQ1 = Does depression differ by Gender? (Only comparing mean scores accross Gender groups)
Ho: There is no significant difference in mean DEPRESSIONsubDASS score between Genders
Ha: There is a significant difference in mean DEPRESSIONsubDASS score between Genders
RQ2 = Does depression differ by Health (Only comparing mean scores across Health groups)
Ho: There is no significant difference in mean DEPRESSIONsubDASS score between Health groups
Ha: There is a significant difference in mean DEPRESSIONsubDASS score between Health groups
RQ3 = Is there an interaction between Gender and Health
(Determine if mean scores across gender groups differ by health groups)
Ho: There is no significant interaction in means DEPRESSIONsubDASS score between Gender and Health groups
Ha: There is a significant interaction in means DEPRESSIONsubDASS score between Gender and Health groups
4. State how your data meets the assumptions for Two-Way ANOVA. Use graphs, descriptive statistics, and additional tests (e.g., test for normality) to support your conclusions. Please, read the example of assumptions write up for that provided in assignment 5 folder (2 points).
** Tried to do log transformations for Sex/Gender variable to no avail. There is a violation of normality for levene’s test
a. Interval level DV
. Normally distributed DV in each grouping
· The distribution is not normal according to shapiro-wilkes, and kolmogrov smirnov
· Tried to transform with log (n+1) and ln(n+1) and failed
c. Equal variance in groupings (Levene’s test)
d. * meet the equal variance assumption since Levene’s test is not significant
3. Equal Variance in groupings (Levene's test)n assignment 5 folder (2 points)
5. Run Two-Way ANOVA statistics and past* Paste tables in here
· The main factors are both significant (p <.05)
a. Gender (F(1,189) = , p = .022)
. PWI HealthCat (F(2,189) = 9/198, p < .001
· Interaction of the factors (PWI HealthCat x Gender) is not significant (p => .05)
· (F(2,189) = 1.398, p = .250)
6. Run a post-hoc analysis and explain what post-hoc test you chose to run and why, paste your output (2 points).
· Since PWI HealthCat is significant, we ran a posthoc. We did not run a posthoc for Gender since there are only 2 groupings and there is no need.
PWI HealthCat:
Low - High: p < .001
Medium - High: p = .001
Low - Medium: not significant
7. Report your findings in APA, write them up in a paragraph format (3 points).
8. State practical significance of your test by reporting eta squared and what it means (2 points)
Gender: η2 = .028 (small because between .01 and .06)
PWI Health Cat: η2 = .089 (medium because between .06 and .14)
Cohen XXXXXXXXXXhas provided benchmarks to define small (η2 = 0.01), medium (η2 = 0.06), and l
Effect size interpretation for η2
large (η2 = 0.14) effects.
Small (η XXXXXXXXXXto .06)
medium (η2 = 0.06 to .14)
large (η2 = 0.14 or greater)
9. State whether or not you retained your null hypothesis or rejected it (1 point).
Main Factors and Interactions:
· For gender we rejected the Ho: (F(1,189 = , p = .022)
· For PWI Health Category we rejected the Ho; (F(2,189) = 9.198, p < .001)
· For the interaction we accepted the Ho :
PostHoc:
· For the pairs Low - High (p < .001) and Medium - High (p = .001) we reject the Ho:
· * For the pair low - high we accepted the Ho:
10. Compute the power of your test, you may use this calculator http:
www.datavis.ca/online/powe
or find another one and explain what it means, see the example of power reporting in the assumptions and power write up file in the Assignment 5 folder (2 points).
** Did not know how to do.