Assignment: Comparing Means FromMore than Two Groups
This assignment will require that you extract data from the course database using MySQL workbench and evaluate the data for analysis using RStudio. In the space below each problem, include the MySQL scripts or R scripts and the respective output that was generated to solve the problem.
Yet again you are dealing with a pesky coworker that has suggested healthcare quality is poor in certain parts of the country. You work for a hospital located in midwestern region of the U.S. Your coworker previously worked for a hospital in the west and claimed that they provide much better care! You asked the coworker to prove it! The coworker showed you data suggesting that the readmission rates for acute myocardial infarctions in the west are the lowest in the country. You weren't convinced so you decided to do your own analysis. Hence, you decided to do the following:
1. Use the course database to extract the following attributes of data from the respective fields. Also, limit the results to just reveal data where the measure_nameis equal to the readmission data for acute myocardial infarction. Also, limit the data so that the number of discharges and number of readmissions is greater than zero. For your answer to this problem, show the MySQL script you used for your query.
|
Table Name
|
Attribute name
|
|
hospital_general_information
|
hospital_name
|
|
hospital_general_information
|
state
|
|
readmission_reduction
|
measure_name
|
|
readmission_reduction
|
number_of_discharges
|
|
readmission_reduction
|
number_of_readmissions
|
|
readmission_reduction
|
excess_readmission_ratio
|
|
regions
|
region_name
|
HINT: Remember to join the hospital_general_information, readmission_reductionand regionstables! Also, don't forget the WHERE clause! You should have 1,759 rows returned!
2. Export the results of the MySQL query to a CSV file and save the file as ami.csv. Import this spreadsheet to RStudio and attach the data. Next, create a variable based on the data you have available that offers a measure of the AMI readmission rate for each hospital.
3. Calculate the mean readmission rate for AMI for each region of the U.S.
4. Explain the distribution of the readmission rate for AMI. Demonstrate the method that you used to determine the distribution of the data.
5. Specify a null and alternative hypothesis to determine if there is a difference in the average AMI readmission ratebetween each region of the U.S.
6. Use a statistical test to determine if there is a difference in the average AMI readmission rate between each region of the U.S.
a. Include the R scripts and R output from your statistical test.
b. Interpret the p-value and specify if you should accept (fail to reject) or reject the null hypothesis.
c. If necessary, perform a Tukey post hoc test to compare the pairwise comparison. If not necessary, explain why.
7. Create a boxplot tocompare the average AMI readmission rate between each region of the U.S.
8. In layman terms, what would you tell your coworker regarding their idea that the west offers much better care then the midwest? Explain your answer.
Perform a statistical test to determine if there is a significant difference in the average number of discharges between hospitals located in the west versus those located in the Midwest. Include the following in your answer:
a. Null and alternative hypothesis.
b. The R scripts and results of you statistical test.
c. An interpretation of your findings relative to the hypotheses.
d. A plot showing the results.
e. Explain if this would impact the message you would deliver to your coworker.