All research projects have a research design. It is the first thing one asks about a project (or at least it should be one of the first things). There are three main components of a research design

  1. (G) groups : research ALWAYS compares something to something, even if completely implicitly (even an ethnography of group X presumes to tell us something about group X in contrast with all the not-X in the world, and all manner of comparisons typically arise during qualitative analysis)
  2. (O) observations : what are we observing, measuring, taking note of?
  3. (X) treatments : we almost always want to study the effect of something and we conceptualize this in the most general terms as "something done to some of the entities we are studying" (with the assumption that we'll compare them to others to whom this was not done). In some cases the "treatment" is possessing some demographic trait as when we compare men and women.
  4. (T) time : if we are going to be more than just descriptive we are usually interested in process and so our research has to unfold in time
In the most generic form of research design we assign subjects to random groups (R), observe (O) or measure before and after some treatment (X) at different points in time (Ti)
T1 T2
R O X O
R X O
R O O
R O

Problems

For each of the following, draw the diagram which represents the research design.

  1. To investigate the effectiveness of a new teaching technique we break the class up into two sections. Both are given a test at the start of the semester to measure how much sociology they know. The first section is taught using the conventional technique, the second section uses the new technique. At the end of the semester both classes are given a test to measure how much sociology they know.
  2. Another research group has a little bit more funding and so they replicate the above experiment but with a variation. They split the class up into four sections. Two get the new pedagogy and two the conventional. In addition, one of each does not take the pre-test. Sketch out the design and explain why this might be a good idea.
  3. You have been put in charge of a marketing research project - they are glad to have a social scientist on board. Your goal is to determine what effects different levels of advertising have on consumption behavior. Based on the results of the project you will recommend the amount of money to be budgeted for advertising different products next year. Your supervisor will require strong justification for your recommendations, so your research design has to be as sound as possible. Develop a research project to address this problem. Focus on the kind of research designs you would use, not how you would conduct the research.1
  4. Researchers want to know whether bicycle helmets reduce the chance of cyclists receiving bad head injuries in an accident. They cannot use standard experimentation and compare a control group of non-helmet wearers with helmet wearers, measuring the chances of head injury, as this would be unethical. Instead, researchers look at medical records, comparing the number of head injury sufferers wearing helmets against those without.2
  5. The method described in the previous problem is vulnerable to a selection bias problem. Explain.3
  6. Consider a study to compare two new methods for teaching. The researchers want to compare two methods, the current method and a modern method. They select two groups of children and match pairs of children across the two groups according to ability, using the results of their last reading comprehension test. They then put one of each pair in a new method class and the other in a current method class.4
  7. A student researcher is interested in texttiles. She has the good fortune to get an internship at an archaeological museum that has lots of texttile artifacts that she can examine. She is particularing interested in the diffusion of the "rasterized weft" form of weaving. She will have access to a large number of samples from two times and several places before and after contact with a well known migrating tribe and she has examples of when this tribe conquered the locals and examples of when the encounter did not seem to involve conquest. Sketch a possible research design.
  8. Some years ago I was involved in a study of the effectiveness of community anti-substance abuse coalitions called "Fighing Back." We had 13 sites around the country. Among other data collection methods, we used data from the Fatal Accident Reporting System (FARS). We gathered data for the five years before the Fighting Back program and for five years after the program. We also got the same data for 5 "comparison cities" in the same state as each of our sites. Draw the research design implicit in this description.
  9. A researcher wants to study the effects of a popular animated TV show on audience attitudes. Her first attempt is using a survey that includes questions about how often the respondent watches the show (if ever), how much s/he likes the show, and then an array of serious attitude scales that have been field tested many times. Sketch her research design and offer a critique.
  10. A researcher wants to understand what works and what does not work when companies try to establish a brand identity, in particular, an identity that contains strong elements of corporate social responsibility. She has plenty of resources; suggest a research design.

References

Bernard, H.R. 2006. Research Methods in Anthropology: Qualitative and Quantitative Approaches. CHAPTER 5
Trochim, W.M.K. 2006. Research Methods Knowledge Base. "Design"
http://www.experiment-resources.com/bobo-doll-experiment.html (note : poorly written)