Chapter 11 Flashcards
an experiment with all the features of an experiment but lacks control
quasi-experiment
Major ways that quasi-experiments are lacking
Indépendant variable is not manipulated
Dependant variable does not have control of some part of the measurement
Extraneous environmental variables
Participant characteristics
When to use a quasi-experiment
The effects of social programs or interventions when participants have something that cannot be manipulated like depression
Why not use a quasi-experiment
they are more vulnerable to interval validity threats
What does an “O” mean in a quasi-experiment design
observation or measurement period
What does an “X” mean in a quasi-experiment design
the treatment
The basic threats to internal validity
History, maturation, testing, instrumentation, regression to the mean, attrition
A treatment occurs and afterwards the dependent variable is measured once
One group post-test only design
Why use a one group post-test only design
used in knowledge retention studies or after a program
Limitations to a one group post-test only design
There is no way to know participants baseline
A dependant variable is measured once before and once after treatment
one group pretest-postest design
Why use a one group pretest-posttest design
Capable of assessing the amount of change from a treatment
Limitations on one group pretest-posttest design
All the threats to internal validation are possible alternative explanation
A dependent variable is repeatedly measured in intervals before and after a treatment
simple interrupted time-series design
why use a simple interrupted time-series design
helps rule out the threats of maturation, testing and regression
Stituations in which, at the start of a study, participants in the various conditions already differ on a characteristic
selection
The interaction of selection with another threat to internal validity
section interactions
Selection x history
participants in one group experience outside events that don’t effect the other
Selection x maturation
when one group’s behaviour is modified and the other’s is not
selection x testing
one group experiences testing effects
selection x instrumentation
there is an issue in the measurement tools in one group
selection x regression
one group is selected on the basis of extreme scores
selection x attrition
one group has a lot of people who drop out
participants in one condition are exposed to a treatment and then both groups take the post-test
posttest -only design with a non-equivalent control group
Issues with posttest -only design with a non-equivalent control group
you cannot compare the two groups based on change because there is no pretest
A pretest and posttest is conducted with one group getting the treatment in-between
pretest-posttest design with a non-equivalent control group
Advantages of a pretest-posttest design with a non-equivalent control group
provides information to judge effectiveness of treatment, how much the dependent variable changed, and informs about types of threats present
Limitations pretest-posttest design with a non-equivalent control group
Cannot control for selection x history
a series of pre- and postreatement scores are taken from two groups at certain times
Interrupted time-series design with a nonequivalent control group
Advantages Interrupted time-series design with a nonequivalent control group
Allows us to know if the treatment caused the change and compare the change with another group
Limitation Interrupted time-series design with a nonequivalent control group
Still lacks control over all threats of internal validity
One group receives a treatment and a nonequivalent group initially does not receive the treatment but is then exposed to it at a later point in time
switching replication design
Pretest postest design with switching replication advantages
Stronger support for the findings, provides more information about internal valitity threats
the initial treatment group no longer receives the treatment when the groups are switched.
switching replication with treatment removal
The evaluation of research methods to assess the need for change or effectiveness. Quasi experiments are used to do this
program evaluation
Benefits of program evaluation
help organizations determin if a new program is needed, if it is effective, if it needs improvement, and if scientific findings can be used in the real world
constraints of program evaluation
Usually asked for by stakeholders who expect certain results
Compontents of a program evaluation
needs assessment, program theory and design assessment, outcome evaluation and efficiency assessment
Determines whether there is a need for a social program
needs assessment
the rational for why a program has been designed in a certain way
program theory and design assessment
Determines whether a program is being implemented as intended
process evaluation
What to include in a process evaluation
- is the program being delivered as attended
- is the person qualified to deliver it
- is the members of the target population getting the program
- is the correct amount getting the program
Assesses a program’s effectivness
outcome evaluation
Occurs when knowledge, services or other experiences that are intended for one group are unintentionally received by another group
contamination
the program’s effectiveness and benefits in relation to costs
efficiency assessment
the process of implementing a program to new groups while maintaining effectiveness
program diffusion
Phases of a program diffusion
dissemination, adoption, implementation , sustainablity
