January 21, 2013
What is a research design?
• Aframework for the collection and analysis of data
- It is a direction, although sometimes vague
• It must meet certain criteria in order to produce useful results
• Research ethics must be considered when choosing a design
Choice of design
• The type of design you choose depends on the kind of explanation you want
Nomothetic explanations (tend to be quantitative in nature) involve
attributions of cause and effect, expressed in terms of general laws and
• They may be developed through particular research subjects.
• Want to be able to extrapolate to the larger population.
• E.g.: The prevalence of suicide in a particular social group is a function of the
level of integration individuals typically have in the group. Want to be able to
generalize to all youth
• 3 causes explained in textbook; correlation, time order, non spurious relationships
Ideographic explanations, typically found in qualitative research, involve a
rich description of a person, or group
• Deep look and perception and feelings of research subjects
• Tends to be localized, it only is meant to apply to particular groups and/or
persons. Don’t want to extrapolate
• Good when studying subgroups, such as gangs, in society
• E.G: Jade became addicted to crack because she never got over her parents
divorce, she felt she was never really accepted by her friends, and had a classmate
who offered her crack, plus other details of hoe Jane interpreted her life. Results
would only apply to Jade
Criteria for evaluating social research
• Causality in social research is directly related to the variables (characteristic or
attributes of data that vary or change) – gender, age are physiological variables
• Variables can also be subjective (interest in subject, beliefs)
• Reliability- the results remain the same each time a particular measurement
technique is used on the same subject. (Assuming that what is being measured has
• The results aren’t influenced by the research, location, timing, etc. • E.g.- New curriculum made- want it to work in all provinces, not just one.
Multiple researchers would test this; find out reasons why results may be
Replicability- The results remain the same when others repeat all or part of a
- The procedures used to conduct the research are sound
- Easier to replicate quantitative research
Validity: There is integrity to the conclusions
Measurement validity/ construct validity: involves the question: “Are you
measuring what you want to measure?”
• E.g.: is the number of murders recorded in annual police statistics a valid
indicator of a murder rate? – Have to ask yourself how they operationalize
murders- are they violent, what age group, all of Canada or just Ontario? How do
they define murder? Does it just relate to males?
Internal Validity: concerned with the issue of whether causation has been
established by a particular study.
• E.g.: did the study establish that personal income level in Canada really is
influenced by ones level of education? More importantly, could income be
influenced by something else? – Are variables related?
• Relates to the idea of independent and dependent variables, have you accounted
for all the independent variables?
Two key terms used when discussing causation:
• Independent variable/ Explanatory Variable: the proposed cause. E.g.: level of
education. Occurs first
• Dependent variable/ Response variable: The proposed effect. E.g.: level of
income. Occurs second as a result of the specified independent variable that
External validity: two primary concerns:
• 1. Are the findings applicable to situations outside the research environment?
- Naturalistic studied tend to satisfy this criteria- is the research design too
artificial? Too much manipulation- cannot apply to natural world. Experiments
done in manipulated setting, downside to experiments
- 2. Can the findings be generalized beyond the people or cases studied?
- Quantitative use representative sampling satisfy this
• Most of the points discussed so far apply mainly to quantitative research
• Lincoln and Guba (1985) maintain that qualitative work should be judges by
• They propose trustworthiness as a criterion for qualitative work
• Trustworthiness includes:
- credibility (internal validity)
- transferability (external validity) - dependability (reliability)
- confirmability (replicability)
Research designs: experiments
• True experiments are common in psychology and organizational studies but rare
in fields like sociology or political science
• Not always able to manipulate variables because they usually are long term issues
that cannot be set up as experiments in social science, such as poverty
• Can also be ethical concerns that can preclude performing experiments
• Even where applicable, experimental models do not get at the perceptions and
feelings of research subjects
1. Field experiment: conducted in real life surroundings. Interactions between
individuals in a park for example
2. Laboratory Experiments: take place in artificial environments, controls the
environment. Lab rats for example. Easier to randomly assign research subjects,
can assure randomization in a laboratory, allows us to have enhanced internal
validity, causation is established.
- Easier to replicate, because everything is controlled
Key concepts relevant to experiments:
• Experimental or treatment group: receives a treatment or manipulation of some
• Control group: does not get treatment or manipulation
• MUST HAVE Random assignment: participants are placed in the experimental
or control group using a random method
• Pre test: measurement of the dependent variable (outcome variable) be