Author Question: The difference between a randomized block design and the more modern variant of including the ... (Read 163 times)

futuristic · **on:** Jul 8, 2018

The difference between a randomized block design and the more modern variant of including the extraneous variable in a multivariate analysis is which of the following?

a. The potentially extraneous variable is treated as an ordinal variable in the randomized block design but can be considered as a ratio or interval variable in a multivariate analysis.
b. In a multivariate analysis, the extraneous variable is excluded from analysis.
c. In a randomized block design two to three times as many subjects were required for statistical analysis.
d. In a randomized block design, the sample is chosen so that it lacks the potentially extraneous variable.

Question 2

The following diagram represents what type of design? Group Subject selection Measurement Treatment Measurement Experimental R M T M Comparison NR M M

a. Posttest only design
b. Factorial design
c. Pretest-posttest control group design
d. Time series design

maaaaaaaaaa · **Reply #1 on:** Jul 8, 2018

Answer to Question 1

ANS: A
The randomized block design uses the two-group pretest-posttest or the two-group posttest pattern with the addition of a blocking variable. The blocking variable, if uncontrolled, is expected to confound the findings of the study. To prevent confounding of the findings, the subjects are rank ordered in relation to the blocking variable. This procedure ensures that the experimental group and the comparison group are equal in relation to the potentially confounding variable. The effect of blocking can also be accomplished statistically (through the use of analysis of covariance) without categorizing the confounding variable into discrete components.

Answer to Question 2

ANS: C
The pretest and posttest design with a comparison group is the most commonly used design in social science research. Subject selection is random (R) for only the experimental group, and non-random (NR) for the comparison group. Measurements for both groups (M) occur both before and after treatment. This quasi-experimental design is the first design discussed here that is generally interpretable. The uncontrolled threats to validity are primarily due to the absence of randomization and, in some studies, to the inability of the researcher to manipulate the treatment.