Instructions
The Week 1 dataset (Excel) contains data from the 2017 CDC Behavioral Risk Factor Surveillance Survey (BRFSS). There is a row for each state. The dataset has the below variables in columns.
The group-level, or aggregate, unit of analysis is the state. The outcome, or dependent, variable is Fair or Poor Health. The remainder of the variables are the predictor, or independent, variables.
Your analysis task is to determine whether any of the predictor variables are associated with the prevalence of people in each state who report having fair or poor health. You will use tasks under the ‘Statistics’ and ‘Graphs’ sections in SAS Studio.
Your communication tasks are to create a table, at least one figure, and a professional abstract to communicate your results.
State: Name of state. This is the unit of analysis.
Fair or Poor Health: Crude prevalence of respondents in each state who reported having ‘Fair or Poor Health’. This is the dependent variable.
Depression: Crude prevalence of respondents in the state who responded yes to ‘Ever been told you have depression’.
Physical Activity: Crude prevalence of respondents in the state who ‘Participated in 150 minutes or more of Aerobic Physical Activity per week’.
Unable to Work: Crude prevalence of respondents in the state who said they were ‘Unable to Work.
No Health Care: Crude prevalence of respondents in the state who said they had ‘No health care coverage’.
Days Clear: Number of clear (sunny, cloud free) days per year in the state.
Expanded Medicaid: Whether the state had expanded its Medicaid program. 1=Yes; 2=No.
Complete the following.
Conduct an exploratory (descriptive) analysis of the data and create a professional table in word or excel to show the descriptive analysis results. See resources for making tables. Be sure to include the table number and title.
For the continuous variable, use the Summary Statistics task to calculate:
Mean and Range (maximum – minimum)
Standard deviation
Measures of normality
For the categorical variable, use the One-Way Frequencies task to calculate:
Count of observations in each category
Percent of observations in each category
Conduct correlation and t-test analyses to determine whether ‘Reported Fair or Poor Health’ is associated with any predictor variables using appropriate statistical tests in SAS. You will conduct Pearson’s correlation using the Correlation Analysis task and 2-sample t-test using the t-test task. See videos for instructions for completing these tests in SAS Studio. If needed, videos in the learning materials can help you choose the best statistical test.
Use the tasks in the ‘Graph’ section in SAS Studio to create at least one graph or figure to illustrate one of these relationships. Figures to consider include, but are not limited to, scatterplots, box plots, or histograms. You must use the ‘Appearance’ section in each task to add axis labels, legend, and title for all figures. See videos for instructions for using these tasks.
Draft a structured abstract, maximum 400 words, with the below sections. See resources for information about writing abstracts.
Background: The research question and hypothesis
Summary: briefly describe your methods
Results: summarize the important findings of the descriptive and analytic analyses with measures.
Discussion: Interpretation, limitations, and public health application of the results.
See the Course Resources for information about creating tables, formatting figures, and writing abstracts.
You will submit:
A word document with your abstract, figure, and table.
Your SAS output.