Bus. 500
Todd Easton
Inferential Statistics Paper
The assignment for this paper is to test a hypothesis using data. The hypothesis you test should probablybe about the relationship between two (or more) variables; the sort of hypothesis one can test using the chi-square distribution or linear regression. I am open to other sorts of hypotheses, but please clear your idea with me in advance if you wish to pursue one.
Ideally, the hypothesis you test will be interesting and important (though I know that may be a lot to ask). As you think about what to do, let interest and importance guide you. If you are choosing between two hypotheses, the first with an element of surprise and the second self-evident, pick the first. If you are choosing between two hypotheses, the first with an impact on the welfare of a business or human beings, and the second with little impact, pick the first.
When you select a hypothesis, think also about what you know. You can do a good job of developing a hypothesis and testing it only if you have some prior knowledge of the terrain you will travel in the paper. If you decide to work on a topic about which you have only a passing familiarity, plan on doing substantial reading before developing your hypothesis.
Finding data
For many students, the biggest challenge of writing this paper is finding appropriate data to test their hypothesis. If you do not know where to the look, you might begin by looking over the links on our course website: click on “Web sites providing access to data.” If you have no luck after searching, feel free to run your hypothesis past me; I might be able to help you find numbers.
In the end, suitable data may not be available. If that happens, be prepared to develop another hypothesis. The fact is, there are lots of great hypotheses that cannot be tested for lack of suitable data. If you need to minimize the time spent searching for data, consider testing a hypothesis with data from one of the data sets on the Survey Documentation & Analysis web site at U.C. Berkeley (http://csa.berkeley.edu:7502). Click on “SDA Archive” to see the data available for download. They include a huge data set containing attitudinal data on samples of Americans (GSS Cumulative Datafile 1972-2004 - Full Analysis). There is also a dataset created from a survey of California residents asking them questions about their jobs and related issues (California Workforce Survey 2001-2002). Because SDA’s access tools are so well designed, both of these data sets are unusually easy to use.
If you download data from the SDA site, or another similar site, the easiest-to-import format will probably be ASCII, comma delimited. If you save such a file to disk (it will probably have a .txt suffix), Excel has a Text Import Wizard that will automatically convert the data into an Excel file, separating each variable into its own column, provided you check the “Comma delimited” box in Step 2 of the Wizard.
Data analysis
Once you formulate your hypothesis and find the data to test it, enter the data into an Excel spreadsheet (if it’s not already in that form). Use Excel to do all the calculations you need to do to test your hypothesis (or else make alternative arrangements with me in advance). Also, create at least two Excel charts or tables to include, supporting the text of the paper.
An excellent paper:
1)Begins with an interesting, important research question.
2)Explains well the research question, its interest, and importance. If there is a relevant academic literature in business, economics, or some other field, the paper includes this literature.
3)Gains insight into the answer to the research question by correctly testing a plausible hypothesis.
4)Explains the data used to test the hypothesis and carefully documents its source, so that (if it is not proprietary) anyone could find it.
5)Explains well the hypothesis test, its result, and the conclusions one should draw from the test.
6)Fulfills guidelines 1.1 to 1.10 in the Pamplin School’s writing handbook, “Expectations for Student Writing”.
7)Carefully cites any outside sources drawn upon, using the format in Section 2.3.2 of “Expectations for Student Writing.”
8)Does a scrupulous job (if it uses outside sources)of restating the author’s ideas (see Section 2.1 of “Expectations for Student Writing”). If you cannot restate the ideas in a satisfactory way, then borrow your author’s words and use quotation marks to indicate the borrowing.
9)Is typed and a maximum of 1500 words long. In addition, you upload:
a) awell-organized Excel workbook tohttps://learning.up.educontaining your data, calculations, and your charts or tables and
b) a Word document to your paper.