Flashcards in Chapter 20 - Research Methods Deck (16)
Types of Research Methods
Secondary Data Analysis
Is there a superior research method?
There is no one superior method. This will vary depending on what you want to study.
Availability of numerical data
focus of process (how, why) and giving voice to participants
data taken at one point in time
study done at more than one point in time
Limitations of Experiments
•Artificial environments – not a ‘real’ setting
•Validity – are you measuring the right thing?
•Hawthorne effect – when people know they are being watched and therefore act differently
•Not applicable to all things that sociologists want to study
•Replicating a Positivist Approach to Research (not conducive to rapport building; “subject” rather than participant)
Limitations of Surveys
•Designing “good” questions is difficult (mutually exclusive, exhaustive categories)
•Forgetfulness (memory fade and telescoping)
•Requires literacy – you have to be careful with what you’re saying; Questions can be interpreted in different ways.
•Data is limited to what is on paper
•Low response rates
Limitations of Observatory Studies
•Low reliability – many argue that participant lower reliability. If one researcher tries to replicate the study of a similar researcher, chances are they wont have similar findings.
•Generalizability – findings may not be generalizable to the group overall (ie. Society wide, culturally, globally)
•Hawthorne effect – subjects changing their behaviour when they know that they are being observed
•Ethnocentrism – the values of the researcher will alter and shape the findings
(If values run counter to the philosophy of the group being studied, biases may be formed.)
Limitations to Secondary Data Analysis
•Incomplete data – reliant on the quality control of the people who originally collected the data – may be incomplete, sloppy, have not asked the important questions, missing variables
•Accuracy of data – were the original collectors careful in their collection?
•Biases of original creators – biases of the people who collected the data embedded in the data itself; especially when there are biases from a different historical period that were more prevalent back then.
does not give voice to participants (viewed as ‘subjects’) because the researcher was assumed to hold all of the knowledge and expertise
identity not revealed to the public, the researcher is able to identify a given persons responses with that person.
identity is not revealed to the public, the researcher is not able to identify a given person's responses with that person. Critical for illegal behaviour.
Classic Unethical Studies
Zimbardo - Mock Prison - emotional and physical harm
Stanley Milgram - psychological experiment (ask participants to shock another person) - deception
Laud Humphreys - sociolofical participant observation study (tearoom trade) - deception
Tuskegee Syphilis Experiment - did not tell people they had syphilis and it was passed on - deception.
On a consent letter
•Purpose of the study
•Procedures involved in the research
•How you will provide confidentiality
•Participation and withdrawal
•Questions about the study