Shared Flashcard Set

Details

Methods Exam
2
53
Other
Not Applicable
11/30/2005
Click here to study/print these flashcards.

Create your own flash cards! Sign up here.

Additional Other Flashcards

 

 

Cards

Term
Procedure Reliability & Validity
 Definition
procedures should be similar every time data is collected
Term
 Instrumentation
 Definition
 Change in instrument between groups or in pre/post tests
Term
 Treatment
 Definition
Change in the treatment (i.e. from a Coke to gum)
Term
 Environment
 Definition
the environment is different between the two measurements
Term
o History Threat
 Definition
an event external to the study that influences people’s behavior within the study.
Term
 To prevent the history threat:
 Definition
• Keep people isolated
• Keep pre and post test close in time
Term
 To see if the history threat has occurred:
 Definition
• Compare control group’s posttest scores to the experimental group’s posttest scores
Term
o Testing Threat
 Definition
an earlier measurement in a study may influence subsequent measurements
Term
 To prevent the testing threat:
 Definition
• Disguise purpose by asking about additional things
• After-only design
• Increase time between pre and post test
Term
 To see if the testing threat has occurred:
 Definition
• Solomon Group Design – compare Group 1’s posttest to Group 3’s posttest
Term
o Hawthorne Effect
 Definition
awareness of a researcher’s intent can influence behavior (confounding effect)
Term
 To prevent the Hawthorne Effect:
 Definition
• Be unobtrusive
• Hide your intent
Term
o Selection
 Definition
selection of the sample may influence the validity of the conclusions
Term
 Two problem with Selection
 Definition
• If no random assignment to groups, the groups may vary beforehand
• May not be generalizable to the population
Term
 To prevent the Selection threat:
 Definition
• Use random selection
• Randomize the conditions
Term
o Regression
 Definition
when groups have been selected on the basis of their extreme scores (Ceiling or Basement effect)
Term
 To prevent the Regression threat:
 Definition
• Don’t select by extreme scores
Term
o Mortality
 Definition
Attrition/people quitting; the loss of research participants from the beginning to the end
Term
 To prevent Mortality threat:
 Definition
• Rewards are based on the completion of the study
• Decrease time
• When you collect data (?)
Term
o Maturation
 Definition
people are continually growing and changing, and these changes may influence the results of the study
Term
 To prevent Maturation threat:
 Definition
• Keep it short
Term
o Diffusion of Treatment
 Definition
when experimental subjects pass on the IV to the control group
Term
 To prevent Diffusion of Treatment:
 Definition
• Isolate the participants
Term
o Personal Attributes Effect
 Definition
When a particular characteristic of the researcher influences people’s behavior
Term
o Unintended Expectancy Effect
 Definition
Inadvertently letting participants know the behaviors that the researchers desire
Term
o Observational Bias:
 Definition
 Observer Drift: fatigue
 Observer Bias: knowledge of hypothesis biases results
 Halo Effect: earlier judgments affect later judgments
Term
EXTERNAL VALIDITY
 Definition
concerns the generalizability of the findings from a study (whether the conclusions from the study can be applied to other people/texts, places, or times
Term
Simple Random Sample
 Definition
assign each person/text a consecutive number and then select from these numbers in such a way that each number has an equal chance of being chosen
Term
Systematic Sample
 Definition
chooses every nth person from a complete list after starting at a random point - the inverval used to choose the nth person is the sampling rate
Term
Stratified Sample
 Definition
categorizes a population with respect to a characteristic a researcher considers to be important (stratification variable) and then samples randomly from each category
Term
Cluster Sampling
 Definition
randomly selecting units or cases (i.e. branch offices) of the target population
Term
Convenience
 Definition
Respondents are selected nonrandomly on the basis of availability
Term
Volunteer sample
 Definition
respondents choose to participate in the study
Term
purposive sample
 Definition
respondents are nonrandomly selected on the basis of a particular characteristic
Term
Quota Sample
 Definition
respondents are selected nonrandomly on the basis of their known proportion in a population
Term
Network/Snowball Sample
 Definition
respondents are asked to refer the researcher to other respondents
Term
ECOLOGICAL VALIDITY
 Definition
research that describes what actually occurs in real-life circumstances
- High: Ethnographic/Naturalistic
- Low: Lab Studies
Term
REPLICATION
 Definition
conducting research that repeats or duplicates a previous study
- Exact: Only has different participants
- Partial: Similar; alters only one or two methodologies
- Conceptual: Testing the concept, but in a different way; usually very different (i.e. survey one time, experiment the second time)
Term
Control
 Definition
attempt to systematically rule out variables that are possible causes of the effect other than the variable hypothesized
Term
 Attribute Variable
 Observed/Natural Variable
 Definition
Gender
HIV + participants
Term
How do you rule out initial differences between conditions?
 Definition
 Random Assignment: equivalent groups – this is different from random selection
 Pre-tests: quasi-equivalent groups
Term
- Pre-Experimental Designs
 Definition
o Post-test Only Design (BAD!)
o Pretest – Posttest Design
o Posttest Only Control Group Design with NON-random assignment
Term
- Quasi-Experimental Designs
 Definition
o Pretest-Posttest Control Group Design (NON-Random)
o Time-Series Design
o Multi-Time Series Design (NON-Random)
Term
- Full/True Experimental Designs
 Definition
o Pretest-Posttest Control Group Design w/Random Assignment
o Pretest-Posttest Comparison Group Design
o Solomon Four Group Design
o After-Only Control Group Design
Term
o Pretest – Posttest Design
 Definition
 O1→X→O2
 Doesn’t control for: history, maturation, testing, instrument decay, or regression
Term
o Posttest Only Control Group Design with NON-random assignment
 Definition
 X→O1
→O2
 Doesn’t control for biased selection or mortality
 Does control for history
Term
o Pretest-Posttest Control Group Design (NON-Random)
 Definition
 O1→X→O3
O2→X→O4
 Controls for all but regression to mean
 Example: to identify maturation look at O2 and O4 – if there is a change than maturation has probably occurred (or history)
Term
o Time-Series Design
 Definition
 O1→O2→O3→X→O4→O5→O6
 Controls for all but History
Term
o Multi-Time Series Design (NON-Random)
 Definition
 O1→O3→X→O5→O7
O2→O4→X→O6→O8
Term
o Pretest-Posttest Control Group Design w/Random Assignment
 Definition
 R O1→X→O3
R O2→X→O4
 Rules out Initial Differences, but may not control for testing effects.
Term
o Pretest-Posttest Comparison Group Design
 Definition
Full Experiment
 R O1→X1→O3
R O2→X2→O4
Term
o Solomon Four Group Design
 Definition
Full Experiment
 R O1→X→ O3
R O2 → O4
R X→O5
R O6
Term
o After-Only Control Group Design
 Definition
Full Experiment
 R X→ O1
R O2
 Campbell likes this design the most (Table on page 182)
Supporting users have an ad free experience!