Saturday, December 5, 2009

Statistics podcasts on Youtube

This fellow, Keith Bower (at has some useful podcasts about statisitical tests which we began discussing when we went to the computer lab just before Thanksgiving. Here's one for the P-value: "If the P-value is low, the null hypothesis must go." See also here:

The correlation coefficient, or Pearson's r: The measure of the strength of the linear relationship, or association, between two variables, and ranges from -1.0 to +1.0.

On confidence intervals:

On the mean and standard deviation:

Friday, November 27, 2009

Research Methods on Youtube!

As an alternative to reading your ever-so-exciting Singleton & Straits course textbook, or listening to me blather on in class, you might try searching around on Youtube for some interesting finds. I typed in "research methods" and came upon the following:

(1) Summary of empirical research:

(2) Theories and hypotheses:

(3) Intro to quantitative research methods: This guy is a little dry, but solid. See how what he tells you compares with what I've taught you.

(4) Get a research method (quantitative vs. qualitative): You'll recognize what this is a spoof of.

Wednesday, November 18, 2009

Talk the Talk, Walk the Walk

I talk with research methods students about my own research in the hopes that it will illustrate key points from the readings, make research seem like something real and practical that can both answer interesting questions as well as potentially lead to gainful employment.

Current research projects in development include a study with public health researcher colleagues about the impact of arrest (e.g., single, multiple) on future health outcomes such as depression, smoking and other substance use. This involves the use of existing, longitudinal data from the National Longitudinal Survey of Youth (NLSY97). Other studies involve original data collected via surveys and focus groups (for instance). I've also begun to explore possibilities for administering surveys online using a randomly selected sample available from the Study Response Project (

Friday, August 21, 2009

"You want me to do what??" What research methods do undergraduate criminal justice majors really need to know?

August 21, 2009: The idea for blogging - on this topic in particular - started with a solicitation from a publishing company to write a textbook. After some pondering, textbook writing didn't seem like such a great idea. (For one thing, it takes time away from other things that matter more.) However, starting a blog on the topic, and other subjects pertaining to research methods and statistics, specifically written with my undergrad students in mind as the main audience, seemed like a more promising idea. So here goes.

I teach Research Methods in Criminal Justice at Kean University in NJ. As Kean CJ majors know, this is part I of a two-part, year long research component that CJ majors must go through. I've also taught the part II, Senior Seminar.

I currently teach out of the 2005 Singleton & Straits, Approaches to Social Research, 4th Edition text, although there are other good texts as well. (Books by Maxfield & Babbie and Hagan come to mind.) The Research Methods course covers the basics of many methods and their building blocks (operationalizing variables, measures' validity & reliability, sampling, sample size, etc.). The idea is to prepare students for what's coming in part II/Seminar. As a colleague says, "In Methods, we teach you how to build a house. In Senior Seminar, you go build the house."

Do the students actually use everything they learn? Will they, as undergraduates, conduct a program evaluation? No. Even so, I devote three class lectures and an in-class exercise to experimental design, and outcome and process evaluation. Students must also read three chapters on these topics. Even though they (likely) won't complete a program evaluation as an undergrad, evaluation is something to keep in mind since, among other reasons (a) they might conduct an evaluation one day, should they go on to graduate school, and (b) there's gainful employment in evaluation research. Many funding streams coming out of the Office of Justice Programs (OJP) require an evaluation component.

So what research methods ARE most relevant to undergraduate CJ majors? "Relevant" in the sense that they'll actually propose and carry out a selected method over the course of 16 weeks in Seminar as part of an original research study. In my humble opinion, they are the following:

(1) a survey of their fellow students, randomly selected from courses available that semester; target sample size of 150+/-

(2) a survey and three focus groups of a smaller sample, randomly selected or not

(3) A content analysis of randomly selected newspaper articles using a search engine such as Westlaw. Seminar projects students did last semester for which they did a content analysis included the effect of media violence on children's behavior, and the influence of a partner on a police officer's behavior specifically as relates to a shooting in the community.

(4) Analyses of existing data downloadable from archives such as I've personally worked with the National Longitudinal Survey of Youth (NLSY) and the National Survey of Families and Households (NSFH), but there are many, many more.

(5) Building a database of existing data culled from different sources, such as the UCR and the Census. Seminar projects that students undertook for which they built and analyzed a database included the influence of social disorganization variables (e.g.,. poverty, residential mobility/% of renters as opposed to home owners) on community-level arrests for domestic violence.

(6) In-depth interviewing of a small sample of students about a topic of interest (e.g., whether Megan's law notifications to community residents makes the residents feel safer about living near sex offenders)