Research Proposal Topics and General Feedback
Applied Research Methods
3/31/03
Many of your project ideas reflected some good thought and a chance to get into the literature at least a bit. Some were much less developed--even in terms of ideas. There are some pieces of advice that apply to all:
The literature. You must get into it. Many of you are doing proposals based on personal/career experience. Remember that noticing something is only the very first step in getting a research idea. You must go to the literature then to see who else has studied aspects of what sparked your interest. Does this mean someone has done the study yet? No of course not. It does mean that people have probably theorized and collected evidence relevant to aspects of your interest.
What literature? Your texts and the professional peer-reviewed journals that are out there. You must review relevant literature to the extent that your research question and method springs forth from what you have reviewed. The review provides a foundation.
The broader question. Remember that this is not a proposal for program evaluation, but for research. Review the difference and be sure to focus on your participants as a sample--select as representative a sample as the group to whom you would like to generalize.
Your question should be one of comparison--beyond description. You can look for a relation between two things (correlation: does y increase as x increase?) or may ask the related question (e.g., are different conditions or groups or groups by conditions are associated with different phenomena).
What causes what? Very few of you are suggesting studies that could be experimental in design. Unless you are, do not talk about causal relations. Even if you do be sure to consider limitations of your procedure.
Operationalization. This is so key for everyone. The finding is a reflection of the procedures used to get it. Make sure you are talking about "measurable" indices of the constructs in which you are interested.
Here are some of your ideas for research proposals with some notes about some of the issues raised. Read through them to consider how other people's issues might apply to your questions or project too.
Sheryl is interested in examining attitudes toward methadone maintenance treatment. Remember the comparison is key--MMT compared to what with whom? This could be studied in an experimental design (see Milewski thesis for example).
Janice is interested in the relation between income and suicide. There might be many other factors involved and the literature will guide you toward those that require your consideration. Anything that might covary--e.g., if there were a relation between access to health care and suicide, then you would need to take that into account because it is clearly also related to income.
MaryBeth is interested in examining HS educational intervention effects on decreasing the pregnancy rate and increasing knowledge about safe sex and use of safe sex practices. Remember to plan to assess what is assessable: perhaps you want to talk about self reported sexual behavior or attitudes toward sexual practices, but the pregnancy rate implies a long term outcome. Try to keep these first proposals simple so that they can build toward a program that will contribute eventually to the kinds of things you really care about. And, remember the comparison group/condition. How can you match them up?
Nicole S. is interested in the relation between the political party affiliation of the leadership and the incidence of battering against women. This is a complex time series analysis if you look at the presidency over history. If, on the other hand, you look at states as your sampling units, then you can compare states to each other with the battering rate as the score, comparing democratic to republican states in the same year (important because the rate changes over time). What else do you need to consider when comparing states?
Jada would like to look at the effects of mentoring. Issues of control and comparison groups will be important.
Sara is interested in finding the relation between middle school students' peer groups and popularity, and the implications of group structure for differences in popularity. Ways to operationalize constructs will need to be identified--how do we index popularity or group structure? The idea of nested groups is important here because popularity in the peer group and in the larger group may or may not be the same thing depending on the relation of the group to the whole. There may be a question of coherence here--is there less variability (of popularity) within peer groups than within random selections of students.
Jackie wants to look at the effectiveness of after school involvement as a deterrent to teen pregnancy. Again, are the groups comparable to begin with and what is the window and what is the comparison condition (i.e., what factors are being held constant and which vary)?
Manisha is interested in assertiveness training and its potential to decrease teen sexual activity among girls. Again, are the groups comparable to begin with and what is the window and what is the comparison condition?
Joyce wants to look at whether Family Group Conferences help families feel more empowered... than whom doing what? Again, the issue of comparisons comes up.
Barbara is interested in finding whether workplaces that have policies prohibiting discrimination on the basis of sexuality encourage self disclosure among gays and lesbians in their employment. Operationalization will be key here--self disclosure, policy, encourage. Also, what kind of workplaces are we talking about and what are the relevant factors to consider?
Erin is interested in finding empirical evidence that indicates that violence against women should be considered a "hate crime." Attitudes and reactions and judgments could be helpful. Could be an experimental scenario. Again, operationalization will be key.
Chris wants to look at the relation between homelessness and addiction, attitudes toward mental illness and the reintegration of mentally ill individuals into society. These are at least two proposals. Be careful about causality in the first. Could use experimental paradigm for the second (third?).
Nicole B. is interested in whether arrests or warnings produce differences in the likelihood that individuals repeat the offense that was the reason for the original complaint. Could take multiple juristictions as sampling units and compare rate of repeat between people who were warned and people who were arrested in a given time period.
Julie would like to look at how to minimize revictimization while encouraging victims of sexual abuse to disclose the abuse. Test one procedure against another? Actual experience of victims vs. simulations and ask people to rate. Could be experimental with right simulation conditions.
Palak is interested in the degree to which formal education predicts careers. Specify and operationalize. Some of this is obvious--what is the difference in occupational prestige of college grads vs. hs grads. Some may be more subtle--may want to look at different aspects of formal education or at different types of individuals (e.g., students with and without learning disabilities).