|
|
BHI Research Methods MEBI 537, Fall, '08 | |||||||||||||||||
|
Final project description: The basic goal is to demonstrate that you can perform learning objective #2, by describing the pros and cons of two different research methods that might shed light on a general research question (or on two closely-related research questions). This project can also be thought of as the design of a research plan of study, such as should be part of a grant proposal. Of course, our focus here is on the methods used to carry out the research. Step zero : Pick a research question. The primary requirement here is that the question be of genuine interest to you personally. The question should also be fairly broad -- see step two and four. Step one : Select, and create a detailed plan (with a timeline) for answering your question. This is study design: you must choose an appropriate research methodology and tailor it to your question and environment. The first step in this plan is to refine your broad research goal into more acheivable, specific aims or study goals. Your plan should be feasible -- you may not propose that you will carry out interviews of 300 subjects over a one-month period. You may assume that you have skills you don't actually have, as long as they are feasible. E..g., you may assume you are (or can hire) an experienced Java programmer, and develop a prototype system ready for testing withing 6 months. Describe this plan #1 in a document; hand in for grading. Due week 5. Step two : Look at another student's research plan. Think about their plan, including possible problems and weaknesses. Then, imagine that this study actually gets carried out: What will happen? Make something up that exploits some weaknesses or strengths of the proposed study. Write up the "results" in a document; hand in for grading, and also give to the student who designed the plan. It may be best to write up this document in two parts: (a) A set of results, which may expose some of the strengths and weaknesses of the design and (b) a section that describes your overall opinion of the design. In the latter section, you may mention other weaknesses or strengths that you weren't able to bring out in your results. Due week 6. Step three: Shift gears. Think about your broad research question in a different manner, and pick an entirely different research methodology that might help answer the question. You may need to modify your question slightly to make this work; and you will certainly need to modify the more specific study goals. My preference would be that your two research methods cut across two of the three traditions we discuss in the course (computer science systems evaluation, qualitative methods, and hypothesis driven quantitative methods), but this is not a requirement. As in step two, design a detailed plan using this second research method. Describe plan #2 in a document; hand in for grading. Due week 8. Step four: Repeat step two but with plan #2. Due week 9. Step five: Summarize your experience in a synthetic final report. You should include your perceived pros and cons of the two different methods, and how these strengths and weaknesses played out with the "results" that were given to you be another student. Try to connect the two sets of research methods. For starters, you should connect and explain the two versions or views of your research question. Another nice way to synthesize your experience would be to connect all parts into one storyline. For example, perhaps some negative results from plan #1 led you to consider the research question (or a varient) with a new research methodology. Then, you may be able to use results from part two to either confirm or reject possible research flaws / weaknesses in your plan one study. Due week 10. Deliverables specifications: Deliverable #1, due Nov 2: For plan #1, your limit is 5 pages of text & figures. Furthermore, you must describe the problem you are addressing, your research question, and provide background and motivation in only one page. (If this sounds difficult, realize that this is also the requirement for grant proposals: they must have a one-page summary.) Thus, the bulk of your document must detail the research methods you will use to approach your research question. You may include references (although I do not expect/require a lengthy review of the literature), and it may be extra, beyond the 5 pages. Deliverable #2, due Nov 9: For your critique & fictitious results, your limit is 2 pages. You may focus mostly on the results, but you must at least have a "discussion" section that explains the results and talks about the implications they might have for the study design. Optionally, you may have have seperate sections labeled "results" and then "critique", where the latter points out problems and raises issues with the study design. Deliverable #3, due Nov 23: For plan #2 your limit is again 5 pages. In some cases, the first page may be very similar to page one of deliverable #1, but in others, the research question and motivation will be slightly different. As stated in the description, the broad problem and background should be the same. Deliverable #4, due Dec 2: This deliverable has the same specifications as #2 (but for plan #2, of course). Deliverable #5, due Dec 11, 5pm: For your final report, you must do some level of integration across your two plans (and sets of results, perhaps). Your report should present the study methods from both plans (perhaps abbreviated a bit), and I expect some discussion of how the two studies and methods relate to each other. You are free to incorporate results from your peers, if that helps your story, and you may also change (to some degree) the details of your two research methods, if in hindsight you feel you made some poor design decisions during the first round. (This is also your chance for rebuttal, if you think you've received some unfair critique from your peers!). The page limit for the final report is 10 pages. Grading rubric: Your overall final project grade will be determined as follows:
Recall that your final project grade makes up 55% of your overall course grade. See grading page. |
||||||||||||||||||
|
Last Updated: |
Contact the instructor at: gennari@u.washington.edu
|