[ Home ]                                                SUNY-Oneonta Department of Psychology, Steven J. Gilbert

Choosing the appropriate statistical test for your experiment

Design and Example Statistical Test

1 IV with 2 levels.  Between Groups Design

Example:

IV = Color of paper (Blue vs. White)

DV = Test Score

                

Subjects are given a test on either blue or white paper.

 

Color of Paper

BLUE

WHITE

S1

S2

.

.

.

S20

S21

S22

.

.

.

S40

                                            

 

Independent Groups t-test

 

Assignment #4

1 IV with 2 levels.  Repeated Measures (Within Subjects) Design

Example:

IV = Brand of Cola (Coke vs. Pepsi)

DV = Taste rating

 

Each subjects drinks both colas and rates each.

                

Cola

Coke

Pepsi

S1

S2

.

.

.

S20

S1

S2

.

.

.

S20

                                            

Note:  Counterbalancing is necessary in this design

 

Paired Samples (Within Subjects)  t-test

 

Assignment #5

1 IV with 3 or more  levels.

Between Groups Design

Example:

IV = Color of paper (Blue vs. White vs. Red)

DV = Test Score

                

Subjects are given a test on either blue, white, or red paper.

 

Color of Paper

BLUE

WHITE

RED

S1

S2

.

.

.

S20

S21

S22

.

.

.

S40

S41

S42

.

.

.

S60

                                            

 

Independent Samples (Between Groups) ANOVA

 

Assignment #6

1 IV with 3 or more  levels.

Repeated Measures (Within Subjects) Design

Example:

IV = Brand of Cola (Coke vs. Pepsi vs. Generic)

DV = Taste rating

                

Each subjects drinks each of the three colas and rates each one.

 

Cola

Coke

Pepsi

Generic

S1

S2

.

.

.

S20

S1

S2

.

.

.

S20

S1

S2

.

.

.

S20

                                            

Note: Counterbalancing is necessary in this design

 

Repeated Measures (Within Subjects)

One-Way ANOVA

 

Assignment #7

2 IVs

Between Groups Factorial Design

Example:

IV#1 = Pratfall (No Spill vs. Spill on Desk vs. Spill on Shirt)

IV#2 = Attractiveness of Candidate (High vs. Low)

DV = Impression of Candidate

                

Each subject sees one video in which a candidate for a teaching position who is either high or low in attractiveness, sips a cup of coffee and either does not spill it, spills it on the desk, or spills it on his shirt.

 

   

Pratfall

 

 

No

 Spill

Spill on

 Desk

Spill on

 Shirt

Attractiveness

 of Candidate

 

High

S1

S2

.

.

.

S20

S21

S22

.

.

.

S40

S41

S42

.

.

.

S60

Low

S61

S62

.

.

.

S80

S81

S82

.

.

.

S100

S101

S102

.

.

.

S120

                                            

 

2-Way Between Groups ANOVA

Assignment #11

2 IVs

Mixed Factorial Design

Example:

IV#1 (Repeated Measures) =  Illusion (Necker vs. Goblet/Faces

vs. Double/Helix)

IV#2 (Between Groups) = Sex of Subject (Male vs. Female)

DV = Number of perceived reversals/minute

 

Subjects are categorized according to sex.  Each subject views three bistable illusions, and records the number of reversals seen in each.

                

   

Bi-Stable Illusion

 

 

Necker Cube

Goblet/

Faces

Double/

Helix

Sex of Subject

 

Male

S1

S2

.

.

.

S20

S1

S2

.

.

.

S20

S1

S2

.

.

.

S20

Female

S21

S22

.

.

.

S40

S21

S22

.

.

.

S40

S21

S22

.

.

.

S40

                                            

Note:  Counterbalancing is necessary in this design

 

2-Way Mixed ANOVA

Assignment #12

2 IVs

Fully Repeated Measures Design

Example:

IV#1 (Repeated Measures) =  Illusion (Necker vs. Goblet/Faces

vs. Double/Helix)

IV#2 (Repeated Measures) = Instructions (Try to see vs. Suppress seeing)

DV = Number of perceived reversals/minute

 

Each subject views three bistable illusions two times (for a total of six viewings): once under instructions to see as many reversals as possible, and once under instructions to suppress seeing reversals.

                

   

Bi-Stable Illusion

 

 

Necker Cube

Goblet/

Faces

Double/

Helix

Instructions

 

Try to see

S1

S2

.

.

.

S20

S1

S2

.

.

.

S20

S1

S2

.

.

.

S20

Suppress seeing

S1

S2

.

.

.

S20

S1

S2

.

.

.

S20

S1

S2

.

.

.

S20

                                            

Note:  Counterbalancing is necessary in this design

 

2-Way Fully Repeated Measures ANOVA

See Dr. Gilbert

       [ Home ]