Terms and concepts from lecture 4

Question 1

Random selection

Answer 1

Data collection in which every person in the population has a chance of being selected. Normally there is an equal chance of being selected. There are two main types of random sampling techniques: Simple random sampling and stratified random sampling

Question 2

Random assignment

Answer 2

All subjects in a sample have an equal chance of being in any of the conditions used in the study. The researcher relies on randomization to control for nuisance variables and controls against various threats to internal validity (selection bias etc). Random assignment should theoretically produce equivalent groups, however, random assignment does not always produce equivalent groups. The likelihood of equivalent groups will increase as the size of the sample increases. You should always investigate whether groups are equivalent after randomly assigning to conditions. Examine group differences with regard to age, gender, SES, and other potential covariates.

Question 3

Matching

Answer 3

Still utilizes random assignment, but will ensure group equivalence on relevant variables. This is used to obtain equivalent groups, especially when a characteristic is known to be correlated with the DV. Participants are matched based on presentation of certain characteristics, then random assignment is performed. Ensures the groups are not different on that characteristic prior to manipulating the IV.

Question 4

Quasi-experimental design

Answer 4

Study control groups are still used, but subjects are not randomly assigned to conditions. Characteristics of the population or sample do not permit random assignment. thus, the plausibility of threats to validity may not be reduced. Since groups have already been formed, they are also referred to as nonequivalent control group designs. This type of study has to control for confounding variables by selecting appropriate controls or through statistical techniques.

Question 5

Multiple treatment design

Answer 5

Some research questions require that all subjects perform under all of the experimental conditions available. In multiple tx designs, the same participants perform in all of the conditions of a given experiment. Then subject scores are compared in the aggregate within the various conditions to determine the extent of a given effect. While the evaluation of treatment is “within-subjects” separate groups are usually used so that the different tx may be balanced. Different versions differ based on the number of tx and the way they are presented.

Question 6

Counterbalanced design

Answer 6

To balance the order of treatment across subjects.

Question 7

Crossover design

Answer 7

Half the subjects get tx A first, then tx B, while the other half gets tx B first, then tx A.

Question 8

Order effect

Answer 8

The point in time (earlier or later in the sequence) in which tx occurrence may be responsible for the results.

Question 9

Sequence effect

Answer 9

The arrangement of tx contributes to their effects (tx before, unlike order effect)