You may use your own data, or use annotated data sets that will be provided, to
analyze. You may include any data tables you deem relevant (APA format is not
necessary for the tables), but the tables must be discussed in the text (more specific is
better than more general). At a minimum, you must include the regression model output
and any tables/charts for checking assumptions and diagnostics. While the data tables
do not need to be in APA format, statistical discussion in the text does need to be in
APA format (F’s and t’s). For class/practice purposes, please provide interpretations
even for nonsignificant findings.
In addition to discussion of your analysis and the results, you must include which
variables (IVs/predictors and DV/criterion) you selected, and why. You must have, at a
minimum, 60 valid records (rows of data) in your model. You must have at least 2
categorical and 3 interval level independent variables. One of the categorical variables
must have 3 or more categories. A literature review is not required, or requested. Make
sure your DV has an adequate range of values. i.e., is continuous with at least 15
possible values.
Some questions you will want to address . . .
– Anything of note in frequencies or descriptives?
– Did you have to change any data values? And if so, why?
– Is you model statistically significant? Report the F statistic. What is the R-squared?
– What is the interpretation of ALL the unstandardized coefficients, whether
statistically significant or not, including the constant?
– Which IV has the greatest influence on the DV? What is the interpretation of that
influence?
– What is the prediction equation?
– Have the assumptions – homoscedasticity, normality of residuals, linearity – been
met?
– Is multicollinearity present?
– Are there outliers, high leverage, or high influence points? You must include tables
of at least the top 10 figures after you sort the data.
– What are your overall conclusions, concerns, or observations about the model?
