Week 5 ANOVA Exercises | Complete Solution

Week 5 ANOVA Exercises

Research Question: Is there a difference in the overall satisfaction of women based on the number of housing problems (no problems, 1 problem, 2 or more problems)?
Using Polit2SetA data set, run an ANOVA using Overall Satisfaction, Material Well-Being (satovrl) as the dependent variable and Housing Problems (hprobgrp) (this is the last variable in the data set) as the independent variable.
Follow these steps when using SPSS:
1.    Open Polit2SetA data set.
2.    Click on Analyze, then click on Compare Means, then One-way ANOVA.
3.    Move the Dependent Variable (Overall Satisfaction “satovrl”) in the box labeled “Dependent List” by clicking the arrow button. The dependent variable is a continuous variable.
4.    Move the Independent Variable (Housing Problems “hprobgrp”) into the box labeled “Factor.” The hprobgrp is a categorical variable coded as (1= no hoursing problem, 2=one housing problem, 3=two or more housing problems).  
5.    Click on the Options button (right side of box) and click on Descriptives and Homogeneity of Variance and then click on Continue.
6.    Click on Post Hoc (right side of box). Click on Tukey and then click on Continue.
7.    Click on OK.
Oneway
[DataSet1] /Users/arnoldresto/Downloads/Polit2SetA (12).sav
Descriptives
Overall satisfaction, material well-being  
    N    Mean    Std. Deviation    Std. Error    95% Confidence Interval for Mean    Minimum    Maximum
                    Lower Bound    Upper Bound        
No Housing Problem    367    12.71    2.353    .123    12.47    12.95    4    16
One Housing Problem    264    11.97    2.588    .159    11.66    12.28    4    16
Two or More Housing Problems    304    10.57    2.594    .149    10.28    10.86    4    16
Total    935    11.80    2.658    .087    11.63    11.97    4    16

Test of Homogeneity of Variances
Overall satisfaction, material well-being  
Levene Statistic    df1    df2    Sig.
2.109    2    932    .122

Post Hoc Tests

Multiple Comparisons
Dependent Variable:   Overall satisfaction, material well-being  
Tukey HSD  
(I) Housing Problems    (J) Housing Problems    Mean Difference (I-J)    Std. Error    Sig.    95% Confidence Interval
                    Lower Bound    Upper Bound
No Housing Problem    One Housing Problem    .739*    .202    .001    .27    1.21
    Two or More Housing Problems    2.139*    .194    .000    1.68    2.59
One Housing Problem    No Housing Problem    -.739*    .202    .001    -1.21    -.27
    Two or More Housing Problems    1.401*    .210    .000    .91    1.89
Two or More Housing Problems    No Housing Problem    -2.139*    .194    .000    -2.59    -1.68
    One Housing Problem    -1.401*    .210    .000    -1.89    -.91
*. The mean difference is significant at the 0.05 level.

Homogeneous Subsets

Overall satisfaction, material well-being
Tukey HSDa,b  
Housing Problems    N    Subset for alpha = 0.05
        1    2    3
Two or More Housing Problems    304    10.57        
One Housing Problem    264        11.97    
No Housing Problem    367            12.71
Sig.        1.000    1.000    1.000
Means for groups in homogeneous subsets are displayed.
a. Uses Harmonic Mean Sample Size = 306.055.
b. The group sizes are unequal. The harmonic mean of the group sizes is used. Type I error levels are not guaranteed.

Assignment: Through analysis of the data and use of the questions below, write one to two paragraphs summarizing your findings from this ANOVA.
1.    What is the total sample size?
Answer: 935
2.    How many women were in each of the different hprobgrp groups?
3.    What are the mean (SD) overall satisfaction scores for each group?
4.    Interpret the Levene’s statistic. (Hint: Is the assumption of homogeneity of variance met? Are equal variances assumed or not assumed?)
5.    What is the value of the F statistic, number of degrees of freedom, and the p value?
6.    Is there a significant difference in the overall satisfaction level of women in each of the hprobgrp groups?
7.    Interpret the post hoc test. When interpreting the post hoc test indicate the mean and standard deviation for each group and indicate which group was signifantly higher or lower from the other. If there is no difference between two groups, indicate that as well.