Holmes University is a small private Midwestern university with undergraduate and graduate degrees. It is trying to increase engagement with its alumni by sending out postcards. It has designed three postcards:
- Career: a postcard featuring career development opportunities and services
- Sports: a postcard featuring sporting achievements at Holmes U.
- Research: a postcard featuring research successes at Holmes U.
You are consulting for Holmes U. on this project, and order a survey of 367 Holmes U. alumni, who are each shown the three postcards. They tell you which is most likely to increase their engagement with Holmes U.
Your task is to decide from the data whether you have evidence that any of the postcards are a better fit for any of the groups of alumni.
- Describe the Alumni messaging data, and create graphs to visualize the data.
- Are the postcards equally likely to be chosen as most effective? Test this using a one-way chi-square test.
- Hint: use the One Way Frequencies dialog in SAS Enterprise Guide, and include the frequency of responses. Your chi-square test statistic should be 17.3678
- Is the choice of postcard independent of gender? Test this using a chi-square test for independence.
- Hint: use the Table Analysis dialog in SAS Enterprise Guide.
- Is the choice of postcard independent of degree earned? Test this using a chi-square test for independence.
- Write up your findings in a case report (you can use the template from below).
- Be sure to recommend which postcard should be used for each of the four groups: Male undergrad, Female undergrad, Male grad, Female grad; or say whether you think two or three of the options are equally good for any of those groups.
- Given that it is cheaper to produce fewer designs, do you think Holmes U should produce all three designs?
- Submit your report like this template:
In your executive summary, describe briefly what you did in your homework this week and what you found. You should include a short summary of the task you were asked to complete, describe how you analyzed the data, and what you found. You should also include any conclusions/recommendations that you generate.
The purpose of an executive summary is to explain the main findings and conclusions/recommendations of your report/project. It should be able to be read completely apart from (and in place of, for readers who aren’t going to read) your full report.
Your executive summaries in this class should be about 100-200 words, depending on how much you need to summarize. Your executive summary should be the last thing you write; how could you summarize your report if you hadn’t written your report yet?
In this class, you can keep your introduction very short. What have you been asked to do? How are you going to go about performing the tasks you need to perform? One or two paragraphs is adequate.
Describe your data. What data do you have? Where does it come from? Show and discuss appropriate summary statistics and/or graphs/plots of your data. Think of this as “supporting evidence” for the analysis you perform in the next section. Try not to spend more than about a page on this.
This is the main section of your report. Describe how you analyze your data, including a short explanation of the statistical methods you use. You may include output from Excel or Enterprise Guide here; but any output needs to be discussed in the text. (If you want to include output that you will not discuss, you can place it in the appendix.)
Try not to make this too “narrative”, i.e. “I did this, then I did this…”. Instead, tell the story of the data: I performed statistical method X on the data in order to determine whether Y was true. The analysis shows evidence for/against Y because…, etc.
Finally, what did you discover? What did we learn from your analysis? How can your analysis inform the decisions that will be made?