Lecture 17 Psyco 350, A1 Fall, 2006

1
Lecture 17 Psyco 350, A1Fall, 2006

• N. R. Brown

2
Outline

• Discrepant Partner Reports and the MSP
• USA random sample -- Web-based Survey
• Discarding the Self-Inclinators
• Age Effects
• Autobiographical Memory
• Introduction
• Retention Factors

3
Rated Truthfulness Bias
4
Self Incrimination
5
Self Incrimination
6
Self Incrimination Males
7
Self Incrimination Females
8
Estimated SPs TessW vs AlbertaT

• SP SI gt nonSI
• ? outliers tend to be SIs

9
MEDIAN SP Estimates w/ w/out SIs -- TessW




10
Strategy Usage ALL Rs vs Non-SI RsTessW
Non-SIs Only
All Respondents
SM very SMall number AP rough
APproximation EN ENumeration TA TAlly RA
RAte OT OTther NR No Response
11
Self-Incrimination Summary

• SI fairly common
• SIs tend to
• give high SP estimates
• favor approximation
• be more embarrassed

12
When SIs removed Summary

• NonSIs
• still produce discrepant partner reports
• strategy selection related to sex of respondent
• Implication
• Bad faith responding plays a role in the
  discrepancy but is not the sole cause.

13
Aging the Partner Discrepancy
14
Age Trends Mean SP
15
Age Trends Meanlog10(SP)
16
Age Trends Median(SP)



17
Age Trends Strategy Use Rough Approximation

18
Age Trends Attitudes

19
Summary

• Discrepancy strategy differences ROBUST
• country, mode, strategy instruction, SI status.
• Self-Incrimination fairly common related to
• estimate size, strategy use, embarrassment
• Age
• related to estimate size, strategy usage, SI
  status, embarrassment
• presence of partner discrepancy

20
Question Why is the discrepancy related to age?

• Cognitive
• Increased reference frame ? strategy shift
• Forgetting
• for high SP ?s shift from enumeration to
  approximation
• for low SP ?s shift from enumeration to just
  know
• Sampling
• ? increased use of CSW w/ age
• Older ? w/ younger ?

21
QuestionWhy do men favor rough approximation?

• Memory
• Motivation
• Distributional

22
Questions What are the origins implications of
self-incrimination?

• SIs flaming trolls?
• Do web-conventions encourage/support extreme
  response patterns?
• Will TessT (telephone version of TessW) produce
  the same pattern of outliers as TessW?
• What about the liars who lie about lying?
• Can truthful-but-inaccurate responses be
  improved?

23
My Current View

• SP Medians stable
• ? ? 5
• ? ? 3
• Median ?s explicable
• for ?
• prostitution, very-young partners, same-sex
  partners
• for ?
• response competition, partner ambiguity
• Implication medians can be used to study
• between group ?
• ? over time

24
My Current Views

• SP Mean ? robust, but numerically unstable.
• Reason
• means driven by outliers.
• between-study differences
• of outliers
• extremity of responses
• Who are the outliers?
• outlying liars
• regular Johns
• Don Juans/celebrities

25
My Current Views

• General Implication
• It may not be possible to solve lifetime SP
  discrepancy.
• Practical Implications
• use shorter time frames longitudinal designs
• screen retest self incriminators
• focus on outliers

26
Recall the first specific personal memory that
comes to mind

• box
• bread
• knife

27

• Rate importance of each, using 1-to-9 scale
• 1 utterly trivial
• 9 earth shattering
• Rate emotional intensity of each event
• 1 ho hum
• 9 !!!!!!!!
• Estimate of date occurrence (month year) for
  the event specified by each of the retrieved
  memories.

28
Expectation A mixed bag

• important/unimportant
• affect laden/affect-less
• recent/less-recent/old
• Questions
• What determines what we remember about our lives?
• How can we find out?

29
Autobiographical Memory

• Contents
• self
• facts
• events
• Issues
• function
• retention factors
• organization
• encoding retrieval processes

30
Autobiographical Memory Methods

• Cue-word Method
• cue word ? event memoryi
• event memoryi ? rate date
• Problems
• verifying event
• dating accuracy
• subjectivity of ratings

31
Autobiographical Memory Methods

• Diary Studies
• Diary Phase Participants record (and rate)
  events soon after they happen.
• Test Phase recall, cued-recall, recognition,
  dating rating.
• Problems
• restrictions on participants events
• generalizing from diarist to non-diarist
• generalizing from recorded events to non-recorded
  events.