|
|
|
|
|
A Factorial Design and MANOVA This case study is divided into two separate parts. The first part requires the use of a factorial design to determine if coffee drinkers are more alert towards the end of the working day than non-coffee drinkers. The case also examines the extent to which obesity may be a factor. The second part requires that you present a set of data in an appropriate manner that allows for interpretation of the data based on an ANOVA with a Covariate and a MANOVA Case assignment due by the end of this Module. PART I A Factorial Design A) You wish to determine if coffee drinkers perform better near the end of the working day than non-coffee drinkers. You also want to know if obesity is also a factor in their ability to negotiate the maze. You develop a simple maze for the study's subjects to negotiate in timed trials. The maze requires that the subjects complete a simple task at various points before being allowed to continue. The maze and required tasks are designed to test performance abilities and one's physical abilities should not effect performance. Before you begin this part of the Case Assignment please review the Table below. Construct a SPSS database for the study. Remember that each person participating this study falls into only one of four combinations for only TWO factors (Obese or Not Obese and Coffee drinker or Non-Coffee drinker). Therefore, there are four different conditions being measured and evaluated which are a combination of only two factors. (Hint: There are only three variables for this study (Time to complete the maze, coffee and obesity. Time is the dependent variable. Coffee and obesity are factors and are grouped variables. They are also called a coding variable in SPSS. Under "Value" click on grey box and open "Define Labels" window. Code the grouping factor of coffee as 1 = Coffee Drinker and 2 = Non-Coffee Drinker. Click on the "Add" button then repeat for Obesity. Go to: http://web.umr.edu/~psyworld/between_subjects.htm for an explanation in factorial designs. Remember, this case requires only a 2 x 2 factorial and not the 2 x 3 factorial discussed on the above linked site. Also, it may help to simplify your data before running the analysis. Once the database is set up please complete the tasks listed after the data table below: B) Data Table for the 2x2 Factorial Design Data. Data presented in number of seconds over 2 minutes for subjects to complete the maze and tasks. Time for Groups of Employees to Complete Maze
1.
Research Question 1: ______________________________ H0: __________________
.
2. Present the data in a meaningful graph with appropriate
explanation/description. 3.
Discuss why Post hoc tests cannot be run with a 2 x 2 factorial design.
4. Describe and interpret your results. PART II Two-Way
ANOVA with Covariate 5.
Utilizing the Ugandan database perform Two-Way ANOVAs with meaningful
variables (factors) as follows: 6.
Properly interpret the SPSS output for each Two-Way ANOVA performed. Properly
discuss how the use of Age at First Intercourse as a covariate had an effect on
the interpretation of the effects of the independent variables on the dependent
variable. 7. MANOVA
- Run a MANOVA with Age at First Marriage and Age at First Intercourse as
the dependent variables. Use the same independent variables that were used
for the ANOVAs used above. 8. Discuss the
differences and similarities between the MANOVA and the ANOVAs. (When you complete this Part of
the case please go to Part III and respond to questions 7 and 8.) Review
the following before conducting the Two-Way ANOVA with Covariate Two-Way ANOVA is an extension of
the one-way ANOVA and has two independent variables instead of one. There are three sets of
hypothesis with the two-way ANOVA, hence the SPSS output will have three levels
of significance. The first two null hypotheses essential state that the
population means for each factor (independent variables) are equal. The
third null hypothesis states that there is no interaction between the two
factors. To perform a Two-Way-ANOVA
do the following steps:
1) Go to Analyze » General Linear Models (GLM) » Univariate
2) Place your Dependent Variable in the Dependent Variable box
and your Factors in the Fixed Factors box. For the Ugandan database,
a dependent variable should be "Age at First Marriage" and Factors could be Ways
to Avoid AIDS, Type of Residence, Religion, among others. 3)
Place the variable "Age at First Intercourse" in the Covariate box. 4) You and then select various Post Hoc tests, descriptive
statistics, and estimates of effect size. 5) Reading
the Output. Review the following
before conducting the MANOVA To perform a MANOVA
do the following steps:
1) Go to Analyze » General Linear Models (GLM) » Multivariate
2) Place your Dependent Variables in the Dependent Variable box
and your Factors in the Fixed Factors box. For the Ugandan database,
a dependent variables should be "Age at First Marriage" and "Age at First
Intercourse". Factors could be Ways to Avoid AIDS, Type of Residence, Religion,
among others. 3) You and then select various Post Hoc tests, descriptive
statistics, and estimates of effect size. 4) Reading
the Output. Case assignment due by the end of this
Module | ||||||||||||||||||||||||||||||||||||||||||||||||
|
|
