# writing statistics

1This assignment is aimed at giving students an opportunity to select and analyze a real quantitative research study and explain how it may or may not be able to help improve an aspect of health care delivery. This will be a 3 – 5 page paper that meets the general writing guidelines which are outlined in the syllabus.

This assignment should be carried out as follows:

Step #1: Study selection – Select the article that most interests you.

1. A Randomized Trial of Intensive versus Standard Blood-Pressure Control

Step #2: Summary: Write a 5 – 6 page summary and analysis of the key aspects of the article you selected. In this summary you should cover the following items in the article:

• Describe the research hypothesis in the article
• Describe the type of quantitative research design selected and why this design was chosen by the author(s), i.e. Descriptive, Correlational, Causal-comparative/quasi-experimental research or Experimental research.
• The following link provides a quick review of the differences http://www.bcps.org/offices/lis/researchcourse/develop_quantitative.html
• Variables tested
• Research design
• Data Analysis
• Results
• Conclusions

2For this assignment, students will be given data from a quantitative analysis and will be asked to analyze it using RStuido, SPSS, STATA, SAS or any other software (your choice).

Data set:

Minnesota Healthcare Database.xlsx

Students will develop an analysis report, in five main sections, including introduction, research method (research questions/objective, data set, research method, and analysis), results, conclusion and health policy recommendations. This is 5-6 page individual project.

Here are the main steps for this assignment.

Step 1: Students require to submit the topic using topic selection discussion forum by end of week 4 and wait for instructor approval.

Step 2: Develop the research question and

Step 3: Run the analysis using RStudio and report the findings using the assignment instruction.

3Week 1, Exercise:

The attached dataset, provides some information about hospitals in 2011 and 2012, download the data and then complete the descriptive table. Please use the following format to report your findings.

Table 1. Descriptive statistics between hospitals in 2011 & 2012

 Hospital Characteristics 2011 2012 p-value N Mean St. Dev N Mean St. Dev 1. Hospital beds 2. Number of paid Employee 3. Number of non-paid Employee 4. Total hospital cost 5. Total hospital revenues 6. Available Medicare days 7. Available Medicaid days 8. Total Hospital Discharge 9. Medicare discharge 10. Medicaid discharge

(Note: Master RStudio script is available for this exercise, but you need to modify that for this analysis)

Week 2, Exercise:

Use the dataset from week1 exercise and then answer the following questions:

• Compare the following information between teaching and non-teaching hospitals.
• What are the main significant differences between teaching and non-teaching hospitals? (use ttest)
• Comparing hospital net-benefit which hospitals has better performance? To answer this question first compute the hospital net benefits with subtracting hospital costs and revenues and then use ttest to compare the significant differences between teaching and non-teaching hospitals.
• Use a box-plot and compare hospitals-cost and hospital-revenues between teaching and non-teaching hospitals.
• Write a short paragraph and describe your findings.

Table 2. Descriptive statistics between teaching and non-teaching hospitals, 2011 & 2012

 Hospital Characteristics Teaching Non-Teaching p-value N Mean St. Dev N Mean St. Dev 1. Hospital beds 2. Number of paid Employee 3. Number of non-paid Employee 4. Internes and Residents 5. System Membership 6. Total hospital cost 7. Total hospital revenues 8. Hospital net benefit 9. Available Medicare days 10. Available Medicaid days 11. Total Hospital Discharge 12. Medicare discharge 13. Medicaid discharge

(Note: Master RStudio script is available for this exercise, but you need to modify that for this analysis)

Week 3& 4, Exercise:

The dataset provides Herfindahl–Hirschman Index, and herfindahel index categories, please use the herf_cat variable and answer the following questions:

Note: “The Herfindahl–Hirschman Index is a commonly accepted measure of market concentration used by antitrust enforcement agencies and scholars in the field. The HHI is calculated by squaring the market share of each firm competing in the market and then summing the resulting numbers” (NASI, 2015; pp: 14-16). read more from here:

Use the dataset from week1 exercise and then answer the following questions:

• Compare the following information between hospitals located inhigh, moderate and low competitive markets? (table 1)
• What are the main significant differences between hospitals in different markets? (use Anova test)
• Use the density curves and compare hospitals cost and revenues between three markets.
• What is the impact of being in high-competitive market on hospital revenues and cost? Do you think being in high-competitive market has positive impact on net hospital benefits? What about the number of Medicare and Medicaid discharge? Do you think hospitals in higher completive market more likely to accept more Medicare and Medicaid patients? What are the impact of other variables? Please discuss your findings in 1-2 paragraphs.

(Note: to answer to the last question, please compute the ratio-Medicare-discharge and ratio-Medicaid-discharge first and then run 2 ttest)high vs. moderate and high vs. low competitive market), please support your findings with box-plot).

Table 3. Comparing hospital characteristics and market, 2011 and 2012

 High Competitive Market Moderate Competitive Market Low Competitive Market ANOVA/Chi-Sq (results) Hospital Characteristics N Mean STD N Mean STD N Mean STD Hospital beds Number of paid Employee Number of non-paid Employee Internes and Residents System Membership Total hospital cost Total hospital revenues Hospital net benefit Available Medicare days Available Medicaid days Total Hospital Discharge Medicare discharge Medicaid discharge 14. Herfindahel index

(Note: Master RStudio script is available for this exercise, but you need to modify that for this analysis)

Week -5& 6

For this week exercise, we need to explore the impact of hospital characteristics on net hospital benefit, so please follow these steps to make your dataset ready for the analysis.

Step 1: As described for week 2 exercise, compute the hospital net benefits with subtracting hospital costs and revenues.

Step 2: As described for week 3&4, compute the ratio-Medicare-discharge and ratio-Medicaid-discharge

Step 3: Use the bed-size categories for this regression

First complete the descriptive table

Table 4. Comparing hospital characteristics and market, 2011 and 2012

 2011 & 2012 Hospital Characteristics N Mean St. Dev Hospital beds Bed Category Bed total <=49 50<=Bed total <=150 151<=Bed total <=250 251<=Bed total <=350 351<=Bed total <=450 Bed total >=500 System Membership Hospital ownership Public For Profit Non-for profit Other Total hospital cost Total hospital revenues Hospital net benefit Medicare discharge ratio Medicaid discharge ratio

(Note: Master RStudio script is available for this exercise, but you need to modify that for this analysis)

Question 5. Regression

1st Model:

`Run a linear model and predict the difference between hospital beds (use the bed-tot) and hospital’s ownership on hospital net-benefit? Discuss your finding, do you think having higher beds has positive impact on the hospital net benefit? What about the ownership?`
` `

2nd Model:

Now, estimate the impact of being a member of a system on hospital net benefit? And discuss your finding (nor more than 2 lines)? Is it significant?

3nd Model:

Now, include the ratio of ratio-Medicare-discharge and ratio-Medicaid-discharge in your model? How do you evaluate the impact of having higher Medicare and Medicaid patients on hospital revenues?

Based on your finding please recommend 3 policies to improve hospital performance, please make sure to use the final model for your recommendation.