CH 8 & 9 - Between & Within Groups Designs

Question 1

Between Groups Design - Variance

Answer 1

2+ groups formed at random from pool of subjects and each group receives different treatment (value of IV)

Groups are compared

1 score per participant (independent measures)

Difference in DV between groups = systematic variance

Difference in DV withing group = non=systemativ variance

Question 2

Between Groups Design - Statistical Significance

Answer 2

Ratio of between group variance to within group variance

F = between-group variance (systematic) /within-group variance (error)

Large between-group variance is good
Large within-group variance is bad

Question 3

How to keep within-group variance low

Answer 3

Limit individual differences

• Standardizing procedures
• Holding a participant variable constant
• Increase sample size

Question 4

Ho to keep groups as equivalent as possible at the start of the research study

Answer 4

• created equally
• treated equally, except for IV
• composed of equivalent individuals

Question 5

Randomization

Answer 5

Participants are randomly assigned to groups before treatment/intervention/manipulation.

*This is the most powerful technique to equalize pre-existing differences by spreading them evenly

**Not the same as random sampling
Random sampling - ransom selection of participants from larger population
Random assignment - random assignment of participants to either group

Question 6

Types of Randomization - Random Assignment

Answer 6

Free Random Assignment - coin toss to ensure assignment is based on chance

2+ use a table of random numbers

Should lead to equality but no guarantee

Question 7

Types of Randomization - Matched-Group

Answer 7

Participants matched into pairs based on critical variables that could be confounds (age, intelligence …) Then within each pair, randomly assign 1 participant to either the treatment or the control group (and the other will be automatically assigned to the other group).

*Inherent in within-group design

Question 8

Types of Randomization - Randomized Block

Answer 8

Matched into groups larger than pairs, matched in blocks based on a range of values. Then within each subgroup, randomly assign participants to either the treatment or the control group.

Question 9

Advantages and Disadvantages of Between Group Design

Answer 9

Adv
- Very simple design
- No time-related/order effects

Dis
- Needs many participants
- Individual differences & environmental differences
- Groups but be equivalent before the manipulation

Question 10

Within Groups Design

Answer 10

1 treatment group, each subject is given every treatment so you get multiple scores for each participant.

Scores at different IV values are compared for the same subject

Each participant serves as their own control.

Question 11

2 Kinds of within-groups designs

Answer 11

• Concurrent measures
• Repeated measures

Question 12

Within-group designs - Concurrent measures

Answer 12

All levels of IV present at same time - choice paradigm

Simplest case scenario - subjects choose the value of the IV they prefer.

Ex: toy preference, present multiple toys and measure how much time is spent with each

Question 13

Within-group designs - Repeated measures

Answer 13

One measure after the other
Most common approach of within-subjects

This accomplishes:
1. equating groups (by using same participants)
2. reducing within-group variance (by controlling for individual differences)

Controlling for individual differences increases sensitivity and ability to detect treatment effect.

Error variance is reduced as participant is their own control (individual differences are eliminated)

Question 14

Within-group designs - Major problems

Answer 14

Time-Related Effects - maturation, history, instrumentation
Order Effects: fatigue, interference, boredom, carryover effects (effects one treatment may have on another)

Exposure to one manipulation may influence responses to later manipulations.

Solutions = Randomization and Counterbalancing

Question 15

Counterbalancing

Answer 15

Full Counterbalancing
All possible treatment orders are used equally – equal numbers of participants in each treatment condition.

Ex: 2 treatments (2 possible combinations) - AB BA

Partial Counterbalancing
Latin Square Design
Each treatment (A, B, C & D) occurs equally often in each position in the experiment.

Ex: left-hand clicker for positive, right-hand clicker for negative. Switch is for group 2.

Question 16

Reversability

Answer 16

Within-subjects designs are not adequate if the experimental conditions produce a lasting effect on the participant that cannot be reversed.

IVs that permanently alter the development or state of participants have irreversible carryover effects.

Question 17

Reversal Design

Answer 17

ABA design: verifies the presence/absence of carryover effects.

Condition A = measure behavior at baseline
Condition B = measure during intervention
Condition A = measure after intervention stopped

If the behaviour returns to baseline there is no carryover effect.

Question 18

Advantages and Disadvantages of Within Groups Design

Answer 18

Adv
- fewer participants needed
- good when participants are hard to find
- greater sensitivity to treatment effect due to elimination of variance caused by individual differences
- each participant acts as his own control
- very powerful design under suitable conditions.

Dis
- not suitable when there are order effects, such as carryover effects, fatigue, practice effects….
- participant attrition may be a problem.