Two-Component Model of Recognition Memory: Early Thoughts and Two-Process Theory - Prof. M, Exams of Psychology

Download Two-Component Model of Recognition Memory: Early Thoughts and Two-Process Theory - Prof. M and more Exams Psychology in PDF only on Docsity!

Two components of recognition

memory

PSY 400, Human Memory

March 8, 2005

Housekeeping

• Should have done Phonological Similarity by today.
• Report is due Tues after Spr Break
• Exam is Thurs after Spr Break.

Important things from last time

• Recognition memory nomenclature.
• Signal detection theory.
• Distinction between discriminability and bias.

Overview of Today’s Material

• Brief signal detection review
• Two-process theory
  1. Early thought: Atkinson and Juola (1973)
  2. Episodic/semantic distinction
  3. Recollection and familiarity
• Techniques for measuring recollection and familiarity
  1. Remember/know
  2. Process dissociation procedure

Important nomenclature

Response Probe Yes No Old hit miss New false alarm correct rejection

Confidence levels in yes/no

recognition memory

• Yes/no recognition asks you to rate your confidence on a 2-point scale.
• Rather than responding “yes” or “no”, you give a more graded response.

Sure No Sure Yes Sure Yes Sure Yes Sure Yes

Bias affects hit and fa rate

Discriminability

• How far apart the distributions are, in units of the standard deviation, called d′.
• d′^ is a measure of discriminability

Estimating d′^ from data

Given a hit rate H and a false alarm rate F A, you can estimate d′^ as

d′^ = z(H) − z(F A),

where z(x) is the z-transform.

Reciever operating characteristic

curves

• ROC curves plot P (hit) as a function of P (fa)
• Allow a way to assess memory at different criteria.
• No discriminability means no memory means P (hit) = P (fa)
• If SDT applies, you should see a special type of curve.

Imagine multiple criteria on our old

and new distributions

ROC curves

• Several different criteria for “yes”
• ROC is hits vs false alarms... -... for several different criteria.

(^00) 0.2 0.4 (^) P(fa)0.6 0.8 1

1

P(hit)

One possibility

page 27, Kahana book.

What’s a zROC curve

Simply plots z-transformed HR as a function of z-transformed FAR.