True Experimental Designs:

Between Subject

 

Pretest-posttest control group design

 

R G M T M

R G T M

E.g. Rosenthal & Jacobson (1966), Teacher expectations:

Pretest and posttest IQ. Treatment: deceiving teachers about student ability.

But pretest may sensitize.

Pretest needed to show exactly what change the treatment produces. But otherwise unneeded (if randomization worked).

(Construct validity problem.)

 

 

 

 

 

Combines the previous two designs.

Allows test for pretest sensitization.

Posttest-only control group design

 

R G T M

R G M

 

 

 

 

 

Solomon four-group design

R G M T M

R G M M

R G T M

R G M