Research Methods Flashcards
Pros & Cons and examples
LAB experiments:
two adv & two disadv (at least)
+control over variables
+casual relationships
+objective data
- lacks ecological validity
- humans cannot be “matched” perfectly
- hawthorne effect
LAB experiments: explain why variables can’t be controlled
Payne & Payne discuss Bandura’s study, arguing against the validity of “anger ratings”– to what extent can human beings be matched & “controlled” as variables?
Participant Observation: three adv & disadv
- helps understand in depth
- overcomes knowledge boundaries of researchers
- avoids hawthorne effect / response bias
- depends on skills of researcher
- subjective interpretations
- may not be applicable to all sociological problems
- usually time consuming
- ETHICS?
Participant Obs: exemplify bad things
Cicurel and Whyte: “the outcomes of P.O depend on the skills of the researcher”
Hammersley: “there can be multiple, contradictory, and yet completely valid interpretations of the same phenomena”
Skeggs: took around 12 years to study caring course women in England. v expensive & time consuming
Participant Obs: exemplify pros
Liebow: “no bias in terms of what is relevant” otherwise the “questions themselves may have assumptions of the researcher”
Whyte: looked into italian americans, found that “cultural differences” mean participant obs. is necessary otherwise bias when designing studies
Jack Young: drug users (who felt “oppositional” to society) study. part. obs. especially useful in discovering the “self concept” of the subjects,
Sample: two major types?
Probability & Non-probability Sampling
Sampling: probability sampling types
Non restrictive: >> a. Simple random sampling Restrictive>>> a. Systematic Sampling b. Stratified random c. Cluster Sampling
Sampling: Non-prob sampling types
a. Snowballing
b. Quota Sampling
c. Convenience Smapling
Example of Snowball sample
Laurie Taylor used it to obtain introductions to various criminals while researching criminal behaviour & deviance
Questionnaires + and - (two each)
+ lots of people
+ cheap
+objective
- grammar and language interpretation
- operationalization problems
- researcher imposition of what’s important
- respondents lie
- subjects are not directly involved = no validity
Questionnaires going to many ppl ex
Gordon Et al reached 1.5k households using questionnaires on poverty
The British Crime Sruvey reaches 49k households on average
People LIE example
La Pierre: restuarants, sent a questionnaire asking if they discriminate and they said no,
then went with 2 chinese people and got rejected out the restaurant
Questionnaire Interpretation problems
Language and grammar & connotations: Deutscher
Blauner problems with operationalizing “worker satisfaction” in a questionnaire
“the instruments the researcher creates are the information they try to reveal” - Philpson
Official Statistics: Marxist Views
Levitas– Looked at statistics and found that basis used to calcuate unemployment was changed, public expenditure figures were manipulated by recording “income” as expenditure,
Nichols – the interpretation of “class” is inaccurate and misleading as it’s about $$ not about relationship to capital/means of production
Miles & Irvine– it supports ruling class, not WHOLLY fabricated lies, but can be manipulated when neccessary
Official Statistics: Phenomenologists
“Justice is negotiable” - Cicourel: historical views of phenomena change and official is subject to stereotype.
Atkinson & coroners & “common sense”
DP Farington and “defining crime”, did not include theft of property worth less than 5p
