notes

A few things I noticed that  needed clarification, from working with people in the lab yesterday. 

1) You need to use the standardized residuals to pick the most important cells (farthest away from zero), then read across that row to compare the percent in that most important cell to the percents in other cells in that row.  This should result in you interpreting at least two rows of each table, with each row containing at least one of the most important cells.  While it is fairly easy to pick the highest and lowest residuals, choosing which cells are most important is not exact -- typically you would not want to discuss a row that has "Don't know", "refused" or "unsure" unless it is the second (or, preferably, the third) row you discuss in that table.  The reason is that "don't know," "refused" and "unsure" responses just don't tell us very much. 

2)  All 5 tables must be listed on the blogsite, and each must be .05 or below for its P value. You must state the P value in the interpretation and say what it means; something like this: "The P value for this table is .000.  This is below the .05 cutoff.  This P value means there is a statistically significant difference between men and women on the question of Jazz music."  This is not required, but if you want, if the P value is .000 you may add the words "very strong", so it reads "very strong statistically significant relationship".

3) You want to think about the kind of paper you want to write -- what topics do you find interesting?

4) You want to make sure to clearly summarize the main findings in the table without using percents or numbers.  You may do this in several sentences.  Make sure it is totally clear.

5) You want to think about being able to tell a story.  This story is just an outline of the main finding for each table, threaded together into a thesis statement.  You don't need to write an introduction or conclusion that outlines that story for this draft of the paper that you are submitting now, but you will need to do that for a later draft that you will submit later.  For this draft you just need to get 5 pages of basic summary done for 5 tables.   These 5 tables will define the topics of your paper for the rest of the semester.  For the qualitative paper that is due later in the semester, we will pull quotes from the open-ended questions to match the tables you chose.  So if you chose a table covering race and rap music for the quantitative paper you submit now, you will be finding quotes on music from different race groups when you submit the qualitative paper later.

6) You want to be thinking about the articles you've read in the class and choose tables that are aligned with those articles.  You may reuse your quiz and other course writings as a rough draft for your literature review.  The lit review will be the last aspect of the paper which is the draft of the paper that is due at the very end of the semester. 

7) Also, be aware that in about two weeks I am going require that you orally present for 5 minutes, covering our research on homelessness or hunger.  This is not necessarily tied to what you submit for the paper, although I encourage you to think about how such a connection could be made.  You may include a table on homelessness as one of your 5 required tables for the paper.  You may include a table on hunger or homelessness as an extra credit table (up to two extra tables -- 7 total - will be accepted).  Extra credit tables may be integrated into the paper, or may be presented as an appendix. You are not guaranteed to be allowed to present on the table you chose for your paper.  There will be a signup sheet posted this week, which will have a limited number of slots to signup to present each significant table.

8) The paper will be submitted to course documents via upload by the end of the day Sunday Nov 9.  You want to present everything in a single word document.  For each table, you should present the table, then the written interpretation.   Each table has two parts, the crosstab table which has the percents and standardized residuals, and the chi-squared table which lists the Pearson chi squared P value (which must be .05 or below to be significant).

9) When interpreting the standardized residual, you want to give the actual number, and define the cell by column and row, then say "more than expected" if its positive residual or "less than expected" for a negative residual.  See the example I posted below.

Let me know if you have any questions.