Chap 12 Flashcards
Define nonequivalent control group design
A quasi-experimental study that has at least one treatment group and comparison group, but participants have not been randomly assigned to the two groups
(Groups are not perfectly equivalent)
(Ex: head start program children vs non head start children)
Repeated measures: quasi-experiments
Participants experience both levels of an independent variable
Define interrupted time-series design
A quasi experimental study that measures people repeatedly on a dependent variable before, during and after the “interruption” caused by some event
(Ex: vacation and job burn outs, tested before during and after vacation)
Define quasi-experiments
Similar to true experiments except experimenters don’t have full experimental control
Define nonequivalent groups interrupted time-series design
It combines two of the previous studies (nonequivalent control group and interrupted time-series)
One treatment group and comparison group (without treatment) and tested multiple times throughout study
What’s the main problem with quasi-experiments?
Internal validity: the researchers ability to draw causal conclusions from the results
We’re there alternate explanations for your findings?
Selection effects are relevant for which group design
Independent group designs
When does a selection threat to internal validity apply
When the groups at the various levels of an independent variable contain different types of participants
This makes it unclear whether it was the independent variable or the different type of participants in each group that led to a difference in the dependent variable between the groups
Define maturation threats
When and experiment or quasi-experimental design has a pretest and post test and a treatment group that shows an improvement over time but it is not clear whether the improvement was caused by the treatment or whether the group would’ve improved spontaneously or even without the treatment
(Comparison groups help rule these out)
Define history threat
Occurs when external historical events happen for everyone in a study at the same time as the treatment variable or by the common external event.
(Comparison groups help rule these out)
Define selection history threat
History threat that applies to only one group not the other
Historical event systematically affects the subjects only in the treatment groups or in the comparison group not both
Define regression to the mean
Occurs when an extreme score is caused by a combination of random factors that are unlikely to happen in the same combination again so the extreme score gets less extreme overtime
(Ex: Phillies lucky 22-1 win over the Reds, did not repeat itself in the next game)
Could be just luck, and luck changes over time
Regression effects are a threat to validity primarily when a group is selected because of its extremely high or low scores.
Define attrition (mortality) threat
Applies when people drop out of a study for some systematic reason.
Occurs mainly in designs with protest and post test
(To check for you should only count people who complete all parts of study)
Define testing threats
Can occur whenever researchers measure participants more than once.
A testing threat is a kind of order fact in which participants tend to change as a result of having been test before
Repeated testing my cause people to improve regardless of the treatment they received
It can also cause people to decline because of fatigue or boredom
Define instrumentation threats
Can be an internal validity threat when participants are tested or observed twice
A measuring instrument to change over repeated uses and this change to threaten internal validity
(Ex: study uses two versions of a test with different standards)
Comparison group can help rule this out
Define observer bias
When the experimenters expectations influence their interpretation of the results
To check: Was the study double blind?
Define experimental demand
And participants guess with the study is about and change their behavior in the expected direction
To check: you can think about whether the participants were able to detect the study’s goals and respond accordingly
Define placebo effect
When the participants improve but only because they believe they’re receiving an effective treatment
To check: comparison group that received inert, placebo, or treatment
Why do we do quasi-experiments
Opportunity: take advantage of real-world opportunities
External validity: enhances external validity. We don’t have to question real world setting because it did (not manipulated)
Ethics: not withholding treatment or benefits it happens on it’s own
Construct validity: how well study is manipulated or measured
Statistical validity: how large the groups differences were (the effect size) and whether the results were statistically significant
How are quasi-experiments and correlation all studies different and the same
Same: not randomly assigned, 2 measured variables
Different: in quasi-experiments researchers tend to do more meddling
Define small-N design
Obtain a lot of info from a few cases
Each participant is treated as a separate experiment.
Almost always repeated measures design in which researchers observe how the person or animal response to several systematic design conditions.
Individuals data are presented
Researchers decide whether result is replicated by repeating our experiment on a new participant
Define large-N designs
Participants are grouped. The data from an individual participant are not of interest to themselves; data from all participants in each group are combined and studied together.
Dara are represented as group averages
Researchers decide whether a result is replicated by doing a test of statistical significance
Define single-N design
Restricting the study to only one animal or person
Define experimental control
What the experimenter controls
Manipulation
The task given
Ex: strong manipulation task would be easy for someone and difficult for someone else
Define replication
Repeating the study
Define study special cases
Takes advantage of a unique population
Just as a quasi-experiment it can take advantage of natural accidents laws and historical events
Are the disadvantages of small-N studies
Few participants might not represent the human population very well
May not be ethical to test on more people in some cases
What are the three small-N designs
Stable-baseline design: researchers spent in several weeks recording baseline information, then taught new strategy/given treatment
The stable baseline meant that there was not a single extremely low point for much improvement would almost definitely occur
Define multiple-baseline design
First recorded a baseline of multiple behaviors
Overcorrection was applied to one behavior and then later extended to other behaviors
Define reversal designs
Researchers observe a problem behavior both with and without treatment
Reversal designed the practitioner takes the treatment away for a while the (reversal phase) to see whether the problem behavior returns
If the treatment was really working then the behavior should worsen again when the treatment is discontinued
