Final Flashcards
Consequence of error on research participants
waste their time, could be uncomfortable procedures, may not implement proper safeguards
Consequence of error on science
erroneous results of research can lead other researchers down false trails; difficult to purge false results
Consequence of error on the public
they take results as true
3 institutional factors leading to scientific malpractice
1) professional schools reward only supported hypotheses and label others as failures
2) research is expensive and grants are contingent on success
3) RAs are not invested and may manipulate data to protect job
intend to be neutral advice giver, reviewing all the scientific evidence on both sides of an issue and pointing out strengths and weaknesses of the evidence on each side
societal mentor
advocates one side of an issue and uses science as a tool for gaining implementation of social policies
social activist
role is to provide specialized knowledge in order to help the jury or judge reach a decision – debate about whether you can take one side when sworn to tell the whole truth
expert witness
the probability of not making a Type II error
statistical power
variance that results from the IV
treatment variance
all variance in the DV that is not caused by the IV
error variance
_____ tests operate by dividing the variance of the DV scores into two pieces (i.e., variance caused by the IV and variance caused by everything else)
parametric
parametric tests are the ratio of ___ variance to ___ variance
treatment; error
in parametric tests, the ___ the ratio, the less likely results are due to error
larger
Variables that are related to/can influence the DV in a study but are not a focus of the research (e.g., gender differences)
extraneous variables
How do you treat extraneous variables so they are not part of error variance?
treat them as IVs or assign into groups
How do you interpret results of multiple group experiments?
one-way ANOVA and post hoc tests
the effect that one IV has independently of the other IV
main effect
occurs when 2 or more IV combine to produce an effect over and above their main effects
interaction effect
designs with multiple independent variables
factorial designs
Experiments that attempt to achieve a balance between control and naturalism in research by studying people’s natural behavioral responses to manipulated independent variables in natural settings
field experiment
2 problems with field experiments
Lower control over IV, DV, extraneous variables, lower control over assignment of participants to IVs
A-B-A design
o A: baseline data
o B: adding reward
o A: remove reward
What is multiple baseline design across settings?
o Baseline
o Contingent reinforcement
o Noncontingent reinforcement
o Contingent reinforcement
Allows a researcher to make inferences by systematically and objectively identifying specific content or characteristics of messages, high on naturalism, can examine data using quantitative or qualitative techniques
content analysis
what the text says in coding schemes
manifest
the interpretation of the underlying meaning in coding schemes
latent
4 rules for coding scheme
1) All terms must be clearly defined
2) Must be a category for each behavior
3) Could have an “other” category, but make sure it is infrequently used
4) A behavior must fit into one and only one category
In content analyses, this focuses on what people, in principle, could be exposed to because it is available to people who want it (e.g., all programs shown on NBC from 7am to 10pm)
availability-based
In content analyses, this focuses on what people are exposed to (e.g., the most popular television shows on NBC from 7am to 10pm)
exposure-based
3 elements of qualitative research
1) Emphasizes collecting richly detailed information about people’s experiences and meaning of those experiences
2) Focus on ways in which experiences are similar and different
3) Rather than starting with a hypothesis, this research starts with a question
in an interview, all respondents answer all questions in specified order
structured interview
Interviews that have the same core questions, but may change order and use probing questions
semi-structured interview
Interviews that use artifact such as photo to elicit dialogue
unstructured interview
the application of social and behavioral research methods to determine the effectiveness of a program or intervention
evaluation research
4 questions of evaluation research
To what extent did the program achieve its goals?
What aspects of the program contributed to its success?
How can the effects of a program be improved?
How valid is the theory underlying a program?
Goal impact: when during the course of a program an effect should occur
timing
Goal impact: how much or an impact a program should have, or effect size of the program
magnitude
Goal impact: how long the effects should last after the clients leave a program
durability
4 criteria evaluating impact of program
Degree of change (effect size)
Importance of change
Costs of the program
Acceptability of the program
3 sources of null results in evaluation research
True null results: program failure
Program implementation failure
Program evaluation failure
What should be done with missing data?
Need to have the sample be the same for all analyses, or explain when you have a smaller sample for some analyses
Recommended to use 5% missing data rule and not all from the same subscale
Need to add a value to data that will not bias the data as a whole
