4: factorial designs

Question 1

why add another factor (IV)?

Answer 1

• Found effect of Factor A on DV
• All EVs were controlled
• But what if we purposely changed one of EVs (Factor B)?
• Would the effect of Factor A on the DV generalize to a different context or would it change?

> If the effect of Factor A differs as the level of Factor B is changed, then the lines will not be parallel

Question 2

recall example explains why adding another variable is better

Answer 2

• Found effect of Factor A on DV
• e.g., longer recall intervals decrease
probability of correct recall

• All EVs were controlled
• e.g., participants in all conditions distracted during the recall interval
• Would the effect of Factor A on the DV generalize to a different context or would it change?
• e.g., what if some participants were not distracted during recall interval?

• Longer intervals decrease recall if distracted
-BUT-
• Longer intervals have little effect on recall
when not distracted

object-preattentive-attentive-preception

Question 3

Visual search: feature (pop-out) (recall example)

Answer 3

• Feature search: The target stimulus “pops out” because a single feature sets it apart from distractors
• According to FIT, preattentive stage where single feature is analyzed in parallel across visual field
• Prediction: “pop out” effect from parallel processing means that increasing set size will NOT have an effect on RT to find target

Question 4

visual search: conjunction (recall example)

Answer 4

• Conjunction search: The target stimulus is set apart from distractors by a combination (conjunction) of
features

• According to FIT, attentive stage where features are bound for each object in series (i.e., one-by-one)
• Prediction: effortful search that requires moving attention from object to object (serial search) means that increasing set size will increase mean RT to find target

Question 5

Visual search: Testing for intercation (recall example)

Answer 5

• The effect of set size (IV1) on reaction time (DV) differs depending on the type of search required (IV2): an interaction between IVs
  • Feature (pop-out) search: no effect of set size on RT
  • Conjunction search: increasing set size increases RT

-Graphical analysis: The lines showing the effect of set size on RT will have significantly different slopes between the two search types

Question 6

factorial designs

Answer 6

• Adding a second independent variable to an experiment results in a factorial design
  • Contains two factors (IVs), each with two or more levels
  •m and n are the number of
  levels of each factor

-An m x n (“m-by-n”) factorial design has m*n conditions
• Design can be between subjects, within-subjects, or a
combination (mixed design)

• Each condition yields a cell mean for a given dependent variable

Question 7

factorial designs: Main effects

Answer 7

-The main effect of each factor is the separate effect of
that factor while ignoring the other factors

• Represented by differences in the marginal means for each level of that factor
  • Average across levels of other factors
• Analogous to a separate experiment with only that IV
  • But values of other factor(s) set to multiple levels in each condition
  • Interpretation of main effects is complicated if there is an interaction

Question 8

Main effects of factor A (columns)

Answer 8

-Calculate the mean score for each level of Factor A (e.g., Search Type)
• Collapsing across levels of Factor B (e.g., Set Size)

-Compare resulting marginal means (column means) to determine main effect of Factor A (Search Type)

-Analogous to conducting a single factor experiment with only Factor A (Search Type) as the independent
variable
• But collapsing (averaging) across levels of all other factors

Question 9

Main effect of factor B (rows)

Answer 9

-Calculate the mean score for each level of Factor B (e.g., Set Size)
• Collapsing across levels of Factor A (e.g., Search Type)

-Compare resulting marginal means (row means) to
determine main effect of Factor B (Set Size)

• Analogous to conducting a single-factor experiment with only Factor B (Set Size) as the independent variable
  • But collapsing (averaging) across levels of all other factors

Question 10

factorial designs: Interactions

Answer 10

-An interaction between factors is when the effect of one factor changes based on the level of another factor

-Simple main effect is the effect of one factor at a given level of the other factor
• There is an interaction when simple main effects differ

-Interactions are best interpreted graphically

Question 11

simple main effect

Answer 11

• Simple main effect is the effect of one factor (e.g., Search Type) at a given level of the other factor (e.g., Set Size)
• Search Type x Set Size interaction: the effect of Search Type changes at different levels of Set Size
  •Simple main effect of Search Type differs among levels of Set Size

Question 12

high order factorial designs

Answer 12

-More than two factors (e.g., independent variables) are
included in a higher-order factorial design

-As factors are added to an experimental design, its
complexity increases (more conditions)
• Number of subjects required (between-subjects factors) or duration
of experiment (within-subjects factors) increases
• Number of possible main effects and interactions increases, complicating their interpretation
• e.g., A x B x C design has three main effects and four interactions, including a three-way interaction