Assignment: Planning ANOVA/ANCOVA
Given the variables you have identified in the hypothetical SPSS database in your research project, consider posing a question for your study compatible with the use of a factorial ANOVA or a factorial ANCOVA for an analysis of the selected data.
In this project task, write a 2 to 3 page proposal for an ANOVA/ANCOVA analysis based on variables related to your dissertation project.
Include the following information:
Identify the following:
At most two categorical IVs that can include IVs with manipulation and/or subject/group variables.
One continuous DV.
One or more continuous CVs (if ANCOVA is used) from your variable set that would fit an ANOVA or ANCOVA design/analysis.
Create new variables if you don't already have exactly what you would need to do this type of analysis.
State a research question and hypothesis(es) concerning expected group differences (main effects and/or interactions) that would be compatible with the selected IVs and DV.
Elaborate on the design, clarifying the IVs, DV, and the CV (if used).
Describe and justify your proposed methods, along with any particular considerations, for each of the following:
Determining sufficient sample size: Specify considerations for the selected alpha level, desired power, and the desired effect size
Ensuring representativeness when selecting a sample
Identifying the procedure for assigning (or otherwise determining) the conditions/treatment groups/subject groups in which participants are studied in the experiment (randomization of error)
Screening data and evaluating assumptions
Interpreting the results of the statistical analyses to evaluate the stated hypothesis(es)
Discuss any possible concerns or limitations in the methods.
Read your partner's descriptions of the research questions and designs for using a factorial ANOVA or factorial ANCOVA analysis. Post your response addressing the following questions:
What are the strengths and limitations of your partner's design?
How does the design promote sample representativeness and randomization of error?
What are your thoughts about the proposed sample sizes?