To provide a break from adolescent psychology notes, as well as distracting me from deleting trackback spam (wich blogspirit seems to do nothing to stop), my weekly notes for Scope & Methods. Most of the post is concerned with the different methods used in political science.
Next Week: Big Week
Must Turn in Preliminary Research Design
Requirements:
Introduction
Literature Review
– where does research fall in the literature?
– not authorwise, but themewise
– (“there are three different ways …, but this way is better …”)
– (don’t say “Hibbing (2006) argues…”)
– (say “Genetic approaches … (Hibbing 2006))
– not enough space for long quites in preliminary research design
State the Hypothesis
State the Methods
Conclusion
3-5 pages
(will be available on blackboard)
Research Methods
Is the chosen research method appropriate for the research question? Does the method let you test the question?
The method and the operationalize help determine each other
Qualitative Research
– non-numerical research
– holistic, “deeper meanings,” contextual, subjective understanding of the actors
Methods of Qualitative Research (Qualitative)
1. Intensive Interviews
– Question Protocol
– Purposive Sample
– Snowball Sample — ask people being interviewed who else to interview)
– qs: how many? whom? how selected? length? re-interviews? protocol?
2. Focus Groups (Qualitative)
– group intensive interview, with a social element
– moderator has to corel discussion
– not private expression, which is both good and bad
– qs: how many? how many in each? who? how selected? question protocol?
Qualitative & Quantitative Research
3. Administrative Records & Documents
– speeches, media, court cases
– What are the data?
– What data are available
– How were data collected?
– Why were data collected?
– How good are the data in terms of fit? Biases?
– How available/accessible are data?
3.5 Content Analysis
– sub-method of AR&D used to understand text
– Manifest Content Analysis
— get a surface understanding
for example, count appearances of the word “crusade” or “crusades”
— quantitative
– Latent Content Analysis:
— get the underlying meaning of a story
— for example, is the tone of an article anti-Israeli
— quantitative or qualitative
4. Case Studies
– what cases are selected?
– why are these cases selected?
– what approach or approaches are you using?
– what research methods will you use (then answer questions from those)
– Controlled Comparisons
— Method of Difference (Most Similar Systems Design): pick similar systems to study the difference in the study variable, but can you actually find cases that are similar enough?
— Method of Agreement (Most Difference Systems Design): pick dissimilar systems to study the similarity in the study variable, but can you actually find cases that are dissimilar enough?
– Congruence Procedures
– Congruence Procedure 1: compare your country’s variable to the “typical” value – between
– Congruence Procedure 2: compare your country’s variable to itself – within, over time, etc
– Process Tracing – go back through casual chains to discovery why a string of events happened
Quantitative Research
– where numbers are assigned to data
– more precise, less contextual
– in political science, held in higher esteem because more “scientific”
– but, prof is skeptical of quantitative bias
– typically, larger “n” than in qualitative studies
5. Surveys
– losing external validity with rise of technology
– Types of Surveys
— Cross-Section Surveys: done at one time
— Panel Survey: interview same people several times
– What is the population?
– How will you select the sample?
– What is the sample size? (n should equal about 1000, almost regardless of population size)
– How will the survey be administered?
– How long will the survey be?
– Preliminary draft of survey
6. Pre-existing surveys (not discussed in class)
– What surveys are available
– How was the sample chosen?
– What is the sample size?
– What are some of the questions you’ll analyze
– How good is the fit with your research?
7. Longitudinal Studies (not discussed in class)
– What is the population?
– How will you select the sample?
– What is the sample size?
– How will the survey be administered?
– How long will the survey be?
– preliminary draft of survey
– How will you keep in contact with the respondents?
– How often will they be re-interviewed and over what period of time?
8. Experiments
– most scientific of all the methods
– yet, questionable external validity
– one of the few methods that allow easy testing of causality, not just correlations
– normal experiment design:
— Experimental Group: Pretest -> Stimulus -> Posttest
— Control Group: Pretest -> Posttest
– by randomdly assigning members, you should be randomly assigning error
– Who are the subjects and how will they be chosen?
– How many subjects will there be?
– How will the experiment work? What is the stimulus? Will there be a pretest and posttest? Will there be a control group? How will the stimulus be randomly assigned?
– Preliminary draft of pretest and posttest
– How will the pretest and posttest be administered?
– How will the stimulus be presented?
A Few Words on Mixed Methods
– more work
– “quantitative and qualitative methods can complement each other, cover the faults and accentuating the strengths”
