Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
44 Cards in this Set
- Front
- Back
Content Analysis - How to formulate the research question
|
1. Decide unit of analysis
2. develop sampling plan 3. construct coding categories and recording sheet 4. coding and inter coder reliability 5. data collection and analysis |
|
Content Analysis- Limitations
|
- fallacy of misplaced concreteness
- unit of analysis - validity and reliability - missing data |
|
Reliability
|
extent which the same technique applied again to the same subject will give the same result
|
|
Aspects of Reliability
|
-Stability across time
- representative across populations - equivalence (questions measure same thing) |
|
How to improve reliability
|
- Conceptualization (appropriate definition
- Increase level of measurements (interval/ratio) -Multiple indicators (more questions) - pretest, pilot studies, and replication |
|
Validity
|
extent to which our measure reflects what we think or what we want to be measuring
|
|
Face Validity
|
measure that relates to what we are interested in finding out even if it doesn't encompass the concept
|
|
Criterion validity
|
predictive nature of a test
ex. success in college measure with SAT scores |
|
Construct Validity
|
measure logically related to another variable as you thought it would be
ex) researching happiness,, measure financial stability |
|
Content Validity
|
how much a measure covers a range of meaning
- was the full range of dimensions related to a concept covered? ex) measuring prejudice but only measure race |
|
Internal Validity
|
addresses the true cases of outcomes in your study
strong internal validity = reliable measures of IV |
|
History effect
|
specific events occurred between 1st and 2nd measure in addition to IV (external factors)
|
|
Maturation Effect
|
internal changes in the subject
|
|
Testing effect
|
effect of pretest on posttest, knowing the questions
|
|
Instrumentation effect
|
researcher changing test
|
|
Selection bias (effect)
|
bias of different individuals selected for each condition
|
|
Mortality effect
|
people drop out of study
|
|
regression to the mean (effect)
|
more tests = closer to the mean
|
|
External Validity
|
how well you can generalize your study to the population
|
|
Levels of measurement
|
Nominal- different categories
Ordinal - different categories, ranked (grades) Interval - different categories, ranked, meaningful distance between (temp) ration - different categories, ranked, meaningful distance, true zero |
|
Principles of good measurement
|
Mutually exclusive, exhaustive, unidimensionality
|
|
Types of Scales
|
Likert
Thurstone Bogardus Social Distance Semantic |
|
Population
|
theoretical group of cases of interest
|
|
Parameter
|
characteristics of cases true to population
|
|
Sample
|
Cases chosen for study, usually from sampling frame
|
|
Sampling frame
|
list as many cases as possible
|
|
Statistics
|
characteristics of cases true to the sample
should represent parameters |
|
Sampling Elements
|
Units of Analysis
|
|
Non-probability sample
|
Convenience
quota purposive snowball deviant case - sample extreme cases sequential - same until saturation theoretical - generating theories, not findings |
|
Probability sample
|
Simple Random sample
- random table # generator -sampling distribution - central limit theorem - confidence interval |
|
Experimental Design
|
1. start with causal hypothesis
2. modify one specific aspect of a situation that is closely connected to the cause 3. compare outcomes |
|
Experimental Design - Matching
|
assigning people to groups due to a variable that the researcher think is really important to take into consideration as it might be related to the DV
|
|
Experimental Design - Random Assignment
|
people assigned to different treatment groups to ensure the group attributes for the different treatments will be roughly equivalent and therefore any effect observed between groups can be linked to the treatment ad is not a result of the characteristics of the subjects
|
|
Experimental Design type
|
One shot case study - treatment then examination (no pretest, no control)
One group pretest posttest design Pretest/posttest control group design Posttest only control group design |
|
What is asked in a survey?
|
Behavior,
-attitudes/beliefs/opinions -characterists -expectations -self classification -knowledge |
|
Survey Design
|
1 - hypothesis, type of survey, write questions, response categories, layout
2 - how to record data, pilot test 3 - decide population, sampling frame, sample size, sample 4 - find respondents, interviews, record data 5 - Analyze Data 6 - Report Data |
|
Survey Questions to Avoid
|
-slang
-ambiguous -emotional language - double-barreled -leading -beyond knowledge -false premises - distant future intentions -double negatives -overlapping response categories |
|
Survey with Sensitive Topics
|
-create comfort and trust
-use enhanced phrasing -establish desensitizing context -anonymous methods |
|
Social Desirability Bias
|
tendency of respondents to answer in a manner viewed favorably by others
|
|
Contingency Questions
|
asked to gauge if respondent is qualified or experience enough to answer
|
|
Layout Design Issues
|
Format - matrix, horizontal vs vertical
Length Mail in response rate |
|
Types of Survey
|
-Mail - low response rates, quick, not personal
-Telephone - invasive, quick, direct, can re-explain question -face to face - personal, give more time, more trust, slow, expensive -Internet - need access, issue of true identity, quick, cheap, wide spread, no time or geography boundries |
|
Interview Bias
|
- errors by respondent
- interviewer error - intentional subversion - influence by interviewer's expectations - failure to probe - influence on the answers |
|
Coding
-manifest (from respondent) -latent (inferred) |
- 4 methods - code sheet, direct entry, optical scan, bar code
cleaning data - possible code cleaning, contingency cleaning |