Research Note Assignment
What could be a more practical use of statistics than to understand what the opposite sex looks for in a date? In this assignment, I want you to use a dataset of 550 speed-dating college students to investigate it. The dataset is on the web at and is in Excel format. You can use SPSS at any university computer to download, import, and analyze it.
Part 1: Are men from Mars and women from Venus?
The data set asks men and women to rate the following characteristics in terms of how important they are in looking for a date:
Attractiveness
Sincerity
Intelligence
Humor
Ambition
Shared Interests
Use your knowledge of SPSS to test whether men and women value those six qualities differently. This will involve doing difference of means tests on each variable, comparing men and women (the gender variable), and reporting your results.
Part 2: Which sex understands the other better?
The dataset asks students two sets of questions: 1) what do you look for in a date, and 2) what do you think the opposite sex looks for in a potential date? So we can see what men look for in a date, and simultaneously what women think men look for. Are they even close? What about men, do they have a clue what women want?
The stereotype is that women think about this kind of thing a lot, while men spend their time wondering when BCS voters are going to quit wasting our time with Ohio State in the championship game. I want you to use this dataset to find out if the stereotype is right. Who has a better picture of what the other sex wants?
This is going to be a bit harder than part 1. You can devise your own test, but if you’re not sure, here’s how I would do it. You’ll want to find out the average score for men on each of the “looks for” variables (like attractiveness, humor, shared interests, etc.). Then find the average score for women on the corresponding “date looks for” variables.” This will give you six observations for “how wrong women are about men.” Do the opposite for “how wrong men are about women.” Create a new dataset in SPSS, and put these numbers in, along with a gender variable (so you’ll have 12 observations, 6 with a value of “1” for gender, and 6 with a value of “0” for gender). Then ask SPSS to run a difference of means test about the difference. The small sample size (6) makes t-tests difficult, so discuss what you find even if the difference isn’t statistically significant.
This assignment is due on the regularly scheduled final date.
Gender and Dating
George Ehrhardt
5 April, 2008
[Your cover page should have your title, name, and date]
Executive Summary [this should also be on the cover page]
This paper examines [insert your main question]. I use [insert the name of your statistical technique here] on a set of [briefly describe your dataset]. These results show [briefly state your main result].
Introduction
Your introduction must contain the following elements.
1) The “hook.” Explain why we should bother reading your paper.
2) Explain your main question.
3) Briefly present your results.
I would put the first two in one paragraph, and the third in the second paragraph. Don’t try for literary merit—write simply and clearly.
Previous Studies[use sub-headings to divide your paper into sections]
In this section, you should describe what other people have written or said about the topic. This helps the reader put your research in a larger context. For example, if you were writing a scholarly paper, you would find articles on similar topics in academic journals, or if you were writing a report on crime statistics, you would want to mention reports by the FBI or other crime agencies.
In this paper, find three sources about gender and dating. You don’t have to be too serious here—articles in Cosmo would work fine, or a book like Men are from Mars, or something like that. They do have to be in print, though—you can use the online version of a magazine or newspaper, but you must be able to record the issue date and page in your bibliography (listing a URL isn’t good enough). When you’ve read your sources, spend 2 paragraphs describing what those sources say about your research. One paragraph should be about the Part 1 question, and the other paragraph on the Part 2 question.
As you describe what your sources say, make sure you cite them. You can do it in several ways. For example, you might say “Writing for Cosmo on 1 April, 2007, Jane Clemons says that …” or you can use footnotes. When you get a job, look around and see how your co-workers do it, but for right now, either is fine with me.
Methodology
This is where you tell the reader what you did in your research. When you do statistical research, this should have two parts. First you tell them about the data, then tell them about your analytical procedure.
First, you should describe each variable in your analysis. The simplest way to do this is to add an indented/bulleted list of all variables, like the following:
Gender: each respondent is identified as either male (1) or female (0).
Next variable:…
The Excel spreadsheet provides a brief description of each variable, but cutting and pasting those won’t be enough. If you have questions or want more information, I’ve put the original dataset codebook up on the ps3115 web page too. It describes some variables that I’ve deleted from the Excel spreadsheet, so you’ll have to look carefully.
Once you’ve told your reader about your variables, you should spend a paragraph telling them what you’re going to do with those variables. In this case, you are going to do difference of means tests, comparing men’s priorities to women’s. Give a one-sentence explanation of what a “difference of means” test is, then tell the reader what variables you are going to do it on.
You should also point out what you expect to find, based on the prior research you described in the previous section. For example, what does Cosmo say about whether men value a partner’s sincerity? If it says men don’t value it, then you should say here that you expect women to score higher on this test than men do.
Results
This is the most important part of your paper. You should go through your tests, one by one, and tell the reader what you found. The best way to do this is to present a table of your results, then explain what the table says (never let the numbers try to speak for themselves—they don’t). In part 1, you have 12 separate results to explain, though you might combine the two sexes’ results for each criteria into a single paragraph or bullet point. In Part 2, you have one main point to interpret.
Remember, when you interpret a difference of means tests, you should say two things: 1) a description of the results for non-statistical readers, 2) a formal statement of statistical significance. For example (use your own words in your paper), “Women place more importance on attractiveness than men do when evaluating potential dates. This result is statistically significant at the .03 level.”
Discussion
This wraps up your paper, and is where you remind the reader what you found, how it matches with previous research by other people, and what is important about your research that your reader should remember.
1) Remind the reader what you’ve found. Don’t list all your findings. Give a one sentence overview, then pick some good examples that support your main finding, and provide those. Notice that this makes four times now that you’ve told the reader about your findings: once in the Executive Summary, once in the Intro, once in the Results, and once here. Your reader won’t read the whole paper, so you have to make sure that even if he only reads a little bit, he’ll still get your main point.
2) Previous research. Tell the reader if your research matches what other people have found (refer back to the methodology section, where you described what they would expect from your data). If your results disagree with previous research, try and explain why.
3) What is important about your research? Why should we care what you think or discover?
Here at school all you usually think about is “you have to read this because you made me write this,” but soon you are going to leave school. Research into what employers want consistently find communication skills to be at the top. Your boss doesn’t have to read what you write—he can ignore you if he wants, and having your boss ignore you is a short cut to a dead-end job. Tell your reader why you are better than that dead-end job by making him care about your research.
You must use page numbers.1