L33 Psych 301 Lecture Notes - Lecture 5: Interference Theory
7 Dec 2020
School
Department
Course
Professor
• Basic experimental design: manipulate one variable (IV), measure behavior on another
variable (DV), control influences of other variables➔ cause and effect is established
• Complex experimental design can mean more than one IV and/or more than two levels
of an IV
o How many levels should one use?
▪ Few levels is good logistically—experiment will take too long, analysis will
be difficult/time-consuming with too many levels
▪ Use more than two levels if:
• You have reason to predict a non-linear relationship
o Two levels can only find a linear relationship
• You are doing exploratory research and don’t really know what to
expect
• 3+ levels provides…
o Do tests reduce impairment due to proactive interference?
▪ Compare (2 levels)
• Testing after lists 1-4 vs. no testing after
lists 1-4
▪ Evidence for test’s protection against PI
• IF recall is better on list 5 when testing
occurred during earlier lists. Measured by:
o Correct recall of list 5
o Number of intrusions from lists 1-4
▪ But results could be due to re-exposure to material
after each list for Lists 1-4. Experiment 3 examines
this possibility by comparing 3 levels of the IV
• Testing after lists 1-4, no testing after lists
1-4, restudy after lists 1-4
• Results showed restudying does not reduce
PI
• 3 levels helped eliminate a plausible
alternative explanation
▪ What happens when we add another IV?
• Adding IVB, control vs. failure, to IVA, favorability of rating in
Better-than-Average-effect
o No matter how everyone did, they were told they did
really poorly on the IVB task
• Addition of the IVB ended up causing the BAE
Document Summary
If recall is better on list 5 when testing occurred during earlier lists. Measured by: correct recall of list 5, number of intrusions from lists 1-4, but results could be due to re-exposure to material after each list for lists 1-4. Experiment 3 examines this possibility by comparing 3 levels of the iv: testing after lists 1-4, no testing after lists. 1-4, restudy after lists 1-4: results showed restudying does not reduce. Pi: 3 levels helped eliminate a plausible alternative explanation, what happens when we add another iv, adding ivb, control vs. failure, to iva, favorability of rating in.
