- QQ：99515681
- 郵箱：[email protected]
- 工作時間：8:00-23:00
- 微信：codehelp

STAT 628 Applied Regression Analysis

Rules for the Final Project

Please read these rules and confirm by email that you have read and understood them before you receive your exam.

1. The deadline is indicated on the syllabus. ABSOLUTELY NO EXTENSIONS. Late submission will be severely penalized.

2. The primary option for submission is via Blackboard Learn (either a computer document or scanned image). Please only submit one copy from one group (if you are working as a group).

3. Email or Blackboard problems will not be accepted as an excuse for late submissions. It is your responsibility to make sure that your email/Blackboard connection works properly and that I receive the submission on time.

4. You may use any publicly available material you want, including books, the internet, etc. (You are NOT allowed to submit questions to internet discussion groups, though!).

5. If you find a solution of a test problem in a book or online, cite it in your submission and do not copy it as-is, but make changes which demonstrate you understand what you are writing. Otherwise we may not accept it.

6. Make an effort to make your submission clear and readable. Severe readability issues may be penalized by grade.

7. Submit your code separately in a file (or integrated into the solution such as an Appendix section) with comments and explanations. Even if the final result is wrong, the code may allow us to find the bug and award partial credit.

Part II. Develop a prototype project using a cross sectional dataset on hospitals in California.

Project Dataset:

Please follow the following link and download one year of dataset (There are multiple years starting from year 2002 but any year is fine as long as it is pertinent to the research question that you are asking):

http://www.oshpd.ca.gov/HID/Hospital-Financial.asp

This page contains the link to the data in Excel 2007 and 2003 version, documentation for data descriptions, as well as the column layout file which may help you locate the columns of variable quickly.

Hint on using the dataset:

Because of the significant differences in the different type of hospitals, psychiatric hospitals and children’s hospitals are excluded from analysis.

Suggested Steps for the Project:

1.Convert the raw dataset to the proper format for the software that you choose to use. Hint: because of the large number of variables in this dataset, you can choose to limit the number of variables before you make the conversion.

2.Develop a research question. Hint: use the framework as we discussed in class. For example, you may be interested in knowing the relationship of A with B, A being some managerial practice or firm characteristics and B being some performance measure. Here are a few examples you may know already:

Production example: A (production line speed) B (scraps), control: none, moderator: none

Insurance Innovation: A (firm size) B (number of months elapsed for adopting innovation), control: firm type in the same-slope model, moderator: none

Insurance Innovation: A (firm size) B (number of months elapsed for adopting innovation), control: firm type in the different-slope model, moderator: firm type

What pair of relationship you would like to study (state your A and B)? What are the possible control variables? What may be the major moderators?

3.Explorative Data Analysis. Hint: always start with univariate analysis such as calculating the mean, standard deviation, number of observations, skewness, kurtosis, potential outliers, distributions. Then you can move on to bivariate analysis. Some techniques such as correlation matrix and scatter plot can be quite helpful.

4.Simple Linear Regression Analysis: This step is optional but highly suggested. Usually before any relationship is shown in the multiple linear regression setting, you need to show the relationship exists in the simple linear regression setting.

5.Multiple Linear Regression Analysis, including outlying observations diagnostics, influential observations diagnostics, multicolinearity checks, etc.

6.Robustness Checks: if you have any results that tell the same story with 1) alternative measures, 2) different sample periods, 3) inclusion or exclusion of certain observations, and so on.

7.Conclusions: This is where you state your conclusions.

8.Limitations and Discussions: This is where you discuss the limitations of the study and discuss what can be further studied.

Suggested structure (and order) for the report:

Title

Author

Executive Summary: This is where you summary your research questions and your findings in an easy-to-understand way.

Introduction: You can introduce the background of your study, why it is important to study this question.

Hypothesis Development: You can state your hypothesis and how you develop your hypothesis in this section.

Data and Methods: You can describe the data and your regression model in this section.

Results: Report your summary statistics, correlation matrix, and regression results in this section.

Conclusions: State your conclusions here.

Discussions: discuss any limitations or further studies here.

References: list all references here.

Appendix: list all code in this section.

Word Template and Example

Here is a Microsoft Word template that you can use for preparing your report:

https://www.informs.org/content/view/full/14237

On blackboard, you can also find an outstanding example.

版權所有：留學生編程輔導網 2018 All Rights Reserved 聯系方式：QQ:99515681 電子信箱：[email protected]

免責聲明：本站部分內容從網絡整理而來，只供參考！如有版權問題可聯系本站刪除。