OT 580 Research (Experimental Design)

An experiment...

1. directly attempts to influence a particular variable

2. participants are randomly assigned to a group

3. a control or comparison must be incorporated within the design

in an experiment you want to (do what to independent variables?)

maximize variability (want it to be strong/making a difference)

in an experiment you want to control variability due to....

extraneous variables

how to control variability due to extraneous variables?

1. random assignment of subjects to treatment groups

2. holding extraneous variable constant

ex. only males or females, certain IQ level

3. match the subjects on an extraneous variable

ex. Tx= 200,000 & 50,000

control = 200,000 & 50,000

4. build extraneous variable into the study (blocking)

5. uses subjects as their own controls (same people for experimental/control group)

6. statistical control of extraneous variables

how to minimize random error?

1. make sure subjects don't become fatigued

2. experimental setting free from distractions and fluctuations

3. PLAN (control situation as much as possible, participants facing the same thing)

4. all measuring devices are reliable

what are quasi-experimental designs

similar to experimental designs but lack random assignment (often pre-existing groups are used)

name the different types of experimental designs

1. randomized pre-test and post-test

2. randomized soloman four group design

3. designs using matching

4. factorial designs

5. repeated measure designs

factorial designs

designs with more than one independent variable (example- high/med/low dosage effects on males & females)