Term
|
Definition
a measure of the outcome variable before any treatment is administered |
|
|
Term
|
Definition
a measure of the outcome variable after the treatment is administered. |
|
|
Term
|
Definition
have a comparison group and have random assignment to T or C Have comparison group who does NOT get the treatment. have random assignment to T or C |
|
|
Term
measurement validity of true experiment |
|
Definition
This is not good in an experiment because it is simulated in a lab and is not in real life situations. |
|
|
Term
causality:true experiment |
|
Definition
Through control, we get association & time order,Through randomization, we get non-spuriousness |
|
|
Term
Threats to causality:true experiment (4) |
|
Definition
non-comparable groups, endogenous change, exogenous change, treatment misidentification |
|
|
Term
|
Definition
the T group is not actually equivalent to the C group |
|
|
Term
|
Definition
natural developments in the subjects that is not caused by the treatment. |
|
|
Term
|
Definition
other things that happen that are not caused by the treatment. |
|
|
Term
treatment misidentification |
|
Definition
get an extra treatment by accident |
|
|
Term
|
Definition
method of creating T&C groups for true experiment where you flip a coin, random number generator.com |
|
|
Term
|
Definition
method of creating T&C groups for true experiment where you Match pairs of subjects on things related to outcome, Randomly assign one to T, one to C |
|
|
Term
block or block group design |
|
Definition
method of creating T&C groups for true experiment where you put similar people in blocks instead of pairs; Randomly ½ of block to T, ½ to C |
|
|
Term
|
Definition
Different from true experiments in ONE WAY: Assignment to T vs. C is not random |
|
|
Term
methods for creating T&C groups for quasi experiments(3) |
|
Definition
non-equivalent control group(not randomly assigned to T&C), time series design(before and after), multiple time series(not randomly assigned to T&C) |
|
|
Term
|
Definition
usually deductive and explanatory, observational as opposed to experimental. |
|
|
Term
How do surveys deal with spurious variables(2) |
|
Definition
§ Experimental: through randomization and control § Observational: through statistical “control”, if possible. No control over what happens. |
|
|
Term
Threats to measurement validity in surveys (3) |
|
Definition
Item non-response Social desirability response Effect of question order and examples |
|
|
Term
When is survey data generalizable? |
|
Definition
IF you have a random sample of a population, you can generalize to that population |
|
|
Term
ways of surveying(modes of administration) |
|
Definition
Mail, Phone, In-person, email / web who might be excluded from the “complete list of people in the population” |
|
|
Term
where are dependent and independent variables usually located in an article? |
|
Definition
Dependent variable usually described in title only, independent variables listed on left side |
|
|
Term
qualitative research designs |
|
Definition
Learn about the social world by going deep into a social setting. Deep and often personal immersion in data. Trying to study: Detailed social processes like interactions & norms |
|
|
Term
8 ways qualitative methods are different from experiments & surveys |
|
Definition
1.the kind of data collected is different. 2.frequently (but not always) inductive. 3.more often exploratory – focusing on issues, phenomena, or processes not studied before. 4.focuses on the social context by considering the interrelations between people and the role of social, political, economic, and other institutions. 5.focus on meanings for people involved & on the ways that people understand their surroundings. 6.more studies idiographic causal stories 7.details of the research design sometimes change as the research progresses. “reflexive” = develop design as a reaction to discoveries and/or needs. 8.The subjective perspective of the researcher is usually acknowledged |
|
|
Term
In depth interviews. Cross sectional versus panel |
|
Definition
Cross-sectional = interview once, Panel = interview again multiple times |
|
|
Term
Intensive vs Structured questions in-depth interviews |
|
Definition
Intensive = tailor each interview to respondent, Structured = ask each respondent same Q's |
|
|
Term
measurement validity for in-depth interviews |
|
Definition
|
|