Experimental design

Question 1

Confounding Variable

Answer 1

Another variable that occurs
along the independent
variable
* Uncontrollable
* Cannot determine whether IV or confound responsible for the effect.
Confounding occurs when
- Causal relationship -cant have other possible effects
Internal validity-Ensure only the IV can be
the cause of changes to the DV
- NO other variable or compound should affect the result

Question 2

Basic experiment:

Answer 2

The independent variable is what we manipulate
Dependent is what we measure

Question 3

3 basic steps for planingexperiment

Answer 3

1. Operationally define independent variable (create at least two levels
   ex. no caffeine high level of caffeine
2. 3.

Question 4

Sampling strategies

Answer 4

• Simple Random
• Stratified Random: divide population into subgroups (then randomly sample each of these groups
• Cluster: randomly sample from a specific group (class)
• Convenience: based off trait or skill
• Purposive
• Quota: certain traits like different numbers or gender activities are represented
  important because women take more surveys more than men could be an imbalanced measure

Question 5

Assign into groups

Answer 5

Three Designs for Assigning Participants to
Conditions
* Independent groups
- Repeated measures: each participant goes through both groups
ex. no caffeine and caffeine
* Matched pairs
ex.

Question 6

(Independent groups design (between subjects design)

Answer 6

Each participant is assigned to only one level
to avoid carry over effect-where prior experience has influence on variable you are measuring

Question 7

Pretest postest design

Answer 7

• see if there are any changes before and after
• Pretest, ideal for smaller studies
  good for retention
  Disadvantage/Caution:
• Pretest can tip participants off to the reason why you’re doing the study
• Time consuming
• can reduce external validity

Question 8

Solutions to pre- test sensitization

Answer 8

• Mask study’s real purpose
  *Use deception
• Embed test with distractors
• Solomon Four-Group Design-control statistically whether taking pretest actually influenced results
  ex.
  1. group doesn’t get pretest, other does
  2. everyone tested on mood
  3. Two groups further divided into control and experimental groups

Question 9

Repeated Measures Design

Answer 9

Different participants
assigned to each
level of IV
* Repeatedly measured
on the dependent
variable after being
in each condition

Advantage?
- Saves time and $$$
- Need fewer participants
- Important is participants are scarce or if study is super $$$
- More sensitive to finding significant differences because they are the same people
- Less error variance ( different people trying to look for consistencies across outcome)

Disadvantage:
- Order effect: The sequence in which they experience the test changes outcome
*practice
*fatigue
* contrast effects (comparatively, murder is much worse than theft)

Question 10

Repeated Measures: Solutions for
Contrast Effects

Answer 10

• Counterbalancing: swap orders every participant of the study goes through every variable

Issue: more conditions = significantly more orders
* Spacing Time Intervals: switch up order of presented conditions
condition presented once in each row in each column

Question 11

Repeated Measures: Spacing Time
Intervals

Answer 11

• Rest period may counteract fatigue
  effect
• Longer participant commitment
• Mortality

Question 12

Matched Pairs Design

Answer 12

Match people on a
participant characteristic
* Obtain a measure of the matching variable from each individual, and then randomly assign participants
* Ensures groups are equivalent