Ch8: Between-subjects designs

Question 1

To how many levels of the IV are subjects in a between-subjects design exposed to?

Answer 1

ONE level of the IV

Question 2

Forms of between-subjects designs

Answer 2

TWO GROUPS:
-treatment and no treatment
-two different levels
-two different categories
MULTIPLE GROUPS:
-multiple levels of IV (including 0)
-multiple categories

Question 3

What are the advantages and disadvantages of between-subjects designs with 2 groups?

Answer 3

Pro:
-able to maximize differences between conditions (extreme ends of a continuum)
Con:
-not as much info can be obtained relative to multiple groups

Question 4

Advantage/disadvantage of between-subjects designs with multiple groups?

Answer 4

Pro:
-greater understanding of relationship between variables due to more levels of the IV along continuum
Con:
-more difficult to find an effect

Question 5

How would you determine significance in a 2 group between-subjects design with interval/ratio data?

Answer 5

independent sample t test.

Significance means groups 1 and 2 differ

Question 6

How would you determine significance in a multiple-group between-subjects design with interval/ratio data?

Answer 6

ANOVA test.
Significance means there is a difference between groups, but it doesn’t tell you where the difference is.
A follow-up t test is then required

Question 7

How would you determine significance in a between-subjects design with nominal/ordinal data?

Answer 7

Chi-squared test for proportions

Question 8

Advantages of between-subjects designs

Answer 8

• sometimes you have no choice
• some research designs are logistically easier when using between-subjects design
• some variables can only be measures once
• individual score is independent from the other scores
• No contrast effects
• no fatigue effects
• no order effects (and other time-related effects)

Question 9

Cons of between-subject designs

Answer 9

• individual differences between participants can become confounding variables. Greater risk of confounding variables
• require more participants
• individual differences can produce HIGH VARIABILITY in scores, making an effect harder to detect
• Threats to internal validity: selection of subjects, selection x maturation interaction, and mortality

Question 10

Why is mortality a threat to internal validity in between-subject designs, but not in within-subject designs?

Answer 10

Because in a between-subject design, losing participants causes uneven conditions. In a within subject design, if a participant is lost, there aren’t people being lost in one condition more than another

Question 11

Do you want MORE or LESS overlap for an effect higher in significance to be found?

Answer 11

LESS overlap for significant effect.

High variance within a treatment makes it harder to see effect

Question 12

Define “contrast effects”

Answer 12

magnification or diminishment of perception as a result of previous exposure

Question 13

threat to internal validity which involves bias in assigning participants to groups is called:

Answer 13

selection of subjects (mrS smith)

Question 14

Different experimental groups would have grown apart regardless of IV. This threat to internal validity is known as:

Answer 14

selection x maturation interaction (mrs Smith)

Question 15

Controlling for confounding variables:

Answer 15

• random assignment
• matching
• holding constant

Question 16

principle behind the “holding constant” strategy for controlling for confounding variables

Answer 16

keeping specific extraneous variables at a specific level

Question 17

When creating groups for a between-subject design, the researcher must make sure that the groups are:

Answer 17

• created equally: process for selection is the same in both
• treated equally:individuals should receive same experience
• composed of equivalent individuals

Question 18

Matching

Answer 18

involves assigning individuals to groups so that a specific variable is balanced, or matched, across groups. The intent is to create groups that are equivalent with respect to the variable matched

Question 19

Control group issues and their definitions

Answer 19

1. Diffusion:
   - participants exchange info among eachother
   - this may impact behaviour of the control group
2. Compensatory Equalization:
   - control group demands that they receive treatment
3. Compensatory Rivalry:
   - control group tries to outperform treatment group
4. Demoralization:
   - control group gives up/they know they aren’t receiving treatment
   - sometimes hard to tell when the observed difference in behaviour is due to treatment, or due to demoralization