Dataset that will support conducting a multiple regression
Choose a dataset that will support conducting a multiple regression
analysis. There should be at least one outcome variable and at least
three predictor variables. It is recommended that there be more
than you need incase you need to down select.
• Download the data, save the file, clean the data and prepare it for
analysis. Describe any methods used to handle missing data and
document any modifications as well as your understanding of how
these modifications may have affected your analyses. Note, the data
set must contain at least 4 variables (i.e., columns). Also, you must
use a distinctly different data set than in the previous two
assignments.
• Characterize the data you are using. This would include completeness,
types of data, distribution of the different variables, and descriptive
statistics. For each column of data used, provide the following
information:
•
o Data type (use terminology such as nominal, ordinal,
interval, ratio, continuous, etc.)
o Units of measure
o Descriptive statistics, including for sub-groups
o Plot the data for all columns with labeled axes, titles, and
a paragraph descriptionFor each variable state how you
interpret this information (i.e., does it make sense? If not,
what are you doing about it.) How did you figure this
out? What implications does the distribution have on how
you will analyze the data? Characterize the distribution
(e.g., symmetric, normal, binomial, skew, kurtosis). What
do these characterizations imply for using the data?
• Conduct linear regression analyses that compares two models, one
with two predictors and one with three predictors. Determine which
has the better fit
