LEVELS OF PROCESSING IN FACIAL MEMORY Sarah Babcock, Rose Ann Calvieri, Lauren Cudney, Vedran Dzebic,Silvia Eleftheriou, and Jeff Mazurkewich,
Post on 23-Dec-2015
218 Views
Preview:
Transcript
LEVELS OF PROCESSING IN FACIAL MEMORY
Sarah Babcock, Rose Ann Calvieri, Lauren Cudney, Vedran Dzebic ,Silvia Eleftheriou, and Jeff
Mazurkewich,
TOPIC DEVELOPMENT Progression of ideas for possible studies:
Initial thought: Combine spatial memory with decision making task
Replaced spatial memory with word memory
Word memory is well studied, we wanted to approach memory from a different angle
Memory involving different levels of processing regarding objects
Decided to examine faces instead of objects
RATIONALE FOR THE STUDY
We encounter faces constantly, making facial memory a crucial skill for social interactions
Examining techniques of facial memory can potentially improve ease of everyday social interactions
LITERATURE: CRAIK & TULVING (1975)
Experiment Processing words at different levels
Subjects presented with words & asked questions to various levels Shallow = same font Medium = counting letters in word Deep = synonyms
Results Better memory recall for words processed at a
deeper level
LITERATURE: BRUCE & VALENTINE (1985)
Experiment Priming task: Subjects were shown the name or
photo of a celebrity Recognition task: Subjects were then shown a
series of photos of celebrities and asked to name them
The subject had seen either the name, same photo or a different photo of the same celebrity in the priming task
Results Better memory for faces which subject’s had seen
pictures of in the priming task (same or different)
Experiment Source memory for faces (internal or external) Questions were asked about faces in a
presentation of facial images The subject either generated the answer or the
answer was provided for them (accompanied the face)
Results Subjects were better at source memory for faces
which asked them to generate answers (internal), than for faces accompanied by the answers (external)
LITERATURE: GEGHMAN & MULTHAUP (2004)
PURPOSE OF THE EXPERIMENT
Purpose: To determine whether different questions can
elicit deeper levels of processing
To establish if deeper levels increase subsequent memory on a facial recognition task
LEVELS OF PROCESSING
Questions: 1) What is this person’s most attractive feature? 2) What job do you think this person has? 3) How old is this person? 4) What is this person’s gender?
Examine whether any of these questions will result in deeper level of processing, measured by accuracy of facial recognition
HYPOTHESIS
If different levels of facial processing can be achieved, deeper level processing will lead to better recognition of faces
Hypothesized to elicit shallower processing: Questions about gender and age
Hypothesized deeper processing: Questions about attractiveness and occupation
Kirkland, Reynolds and Pezdek (1992)
METHODS
Subjects
30 subjects 24 in experimental group 6 in control group All Mac undergrads Age range 18-24, Mean 20
METHODS
Stimulus/Materials
Study Task 32 faces
Experimental group Each Face paired with one question
Control group No questions presented
METHODS
Stimulus/Materials
Recognition Task 60 faces (28 novel)
Have you seen this face in the previous presentation? Yes/No responses
All subjects given same task
METHODS
Cover Story
Study Task – Subjects were told: There will be questions about the faces They need to answer as quickly as possible the questions We are looking at how much you can tell about a person
by their appearance
Recognition Task: Subjects were naïve of recognition task to follow the
study task
LEVELS OF PROCESSING
Questions: 1) What is this person’s most attractive feature? 2) What job do you think this person has? 3) How old is this person? 4) What is this person’s gender?
Examine whether any of these questions will result in deeper level of processing,
measured by accuracy of facial recognition
SLIDESHOW EXAMPLE: STUDY TASK
ATTRACTIVENESS?
AGE?
GENDER?
JOB?
SLIDESHOW EXAMPLE: RECOGNITION TASK
DATA COLLECTION
Study Task Recorded subject’s responses to questions
Recognition Task Recorded if the subject answered yes or no
STUDY TASK DATA SHEET
Subject #
Question Orders
Picture # Group NumberS’s response
1 2 3 4
1 a d b c
2 b c a d
3 c b d a
4 d a c b
5 a d b c
6 b c a d
RECOGNITION TASK DATA SHEET
Subject #
Recognition Task Ss Response Sheet
Recognition task slide #'s
Picture Number 1 2 3 4 S’s response
S’s Correct Responses
2 54 0 0 0 0
4 21 a d b c
6 27 c b d a
8 45 0 0 0 0
10 43 0 0 0 0
12 14 b c a d
14 41 0 0 0 0
16 28 d a c b
18 24 d a c b
RESULTS
Group Results (Experimental & Control)
Independent t-test
Descriptives
One-way ANOVA
Post-hoc (Bonferroni)
GROUP STATISTICS
Group Statistics
Group N Mean Std. Deviation Std. Error Mean
Sstotal 124 49.4167 5.19964 1.06137
26 40.6667 6.37704 2.60342
INDEPENDENT T-TEST
Independent Samples Test
Levene's Test for
Equality of Variances t-test for Equality of Means
F Sig. t dfSig.
(2-tailed)Mean
DifferenceStd. Error Difference
95% Confidence Interval of the
Difference
Lower Upper
Sstotal
Equal variances assumed
.033 .858 3.531 28 .001 8.75000 2.47783 3.67440 13.82560
Equal variances not assumed
3.112 6.760 .018 8.75000 2.81146 2.05373 15.44627
DESCRIPTIVES
Descriptives
Data
N Mean Std. Deviation Std. Error
95% Confidence Interval for Mean
Minimum MaximumLower
BoundUpper
Bound
124 6.7917 1.02062 .20833 6.3607 7.2226 5.00 8.00
224 6.0833 1.44212 .29437 5.4744 6.6923 3.00 8.00
324 5.2500 1.89393 .38660 4.4503 6.0497 1.00 8.00
424 5.7500 1.32698 .27087 5.1897 6.3103 3.00 8.00
Total96 5.9688 1.53865 .15704 5.6570 6.2805 1.00 8.00
ONE WAY ANOVA
ANOVA
Data
Sum of Squares df
Mean Square F Sig.
Between Groups
30.115 3 10.038 4.741 .004
Within Groups194.792 92 2.117
Total224.906 95
POST HOCMultiple Comparisons
DataBonferroni
(I) Questions
(J) Questions
Mean Difference (I-J) Std. Error Sig.
95% Confidence Interval
Lower Bound Upper Bound
1 2 .70833 .42005 .571 -.4243 1.8410
3 1.54167* .42005 .002 .4090 2.6743
4 1.04167 .42005 .090 -.0910 2.1743
2 1 -.70833 .42005 .571 -1.8410 .4243
3 .83333 .42005 .301 -.2993 1.9660
4 .33333 .42005 1.000 -.7993 1.4660
3 1 -1.54167* .42005 .002 -2.6743 -.4090
2 -.83333 .42005 .301 -1.9660 .2993
4 -.50000 .42005 1.000 -1.6327 .6327
4 1 -1.04167 .42005 .090 -2.1743 .0910
2 -.33333 .42005 1.000 -1.4660 .7993
3 .50000 .42005 1.000 -.6327 1.6327
*. The mean difference is significant at the 0.05 level.
LIMITATIONS OF CURRENT STUDY
1) Response time:
5 second time limit Subjects provided arbitrary responses May have focused more on the question than
actual face Further studies: allow slightly more time for
subjects’ responses
LIMITATIONS OF CURRENT STUDY
2) Occupation Question:
Was predicted to elicit deeper processing Answer didn’t require face processing Future studies: questions relying more on facial
features (I.e ethnicity, cosmetic surgery)
LIMITATIONS OF CURRENT STUDY
3) Facial Images:
Atypical facial images compared to participants and people within the participants’ environment
Non significant results maybe due to quality of faces
Faces presented in grey-scale : require deeper processing
Future studies: Use of coloured images is more realistic
Use more updated faces similar to those within participants’ social environment
LIMITATIONS OF CURRENT STUDY
4) Number of Images:
Study task: 32 faces Recognition task: 60 faces Future studies: More faces in study and recognition
tasks Increase power
LIMITATIONS OF CURRENT STUDY
5) Control Group:
Smaller than experimental group Future studies:
Larger number of subjects in control group size = increase power
Within-subjects control group: Have some pictures without any questions in
the slideshow
IMPLICATIONS
If significant:
Faces and words are processed similarly Improve peoples ability to remember new
acquaintances Eyewitness testimony
CONNECTIONS TO PREVIOUS RESEARCH
Verbalization and conceptualization of faces lead to better facial recognition Itoh, 2005; Bruce & Valentine, 1985
Improved memory when face is paired with question, and when the answers are generated Geghman & Multhaup, 2004
Levels of processing may have had an effect on facial recognition Craik, & Tulving, 1975
Word memory vs. Facial memory
Mechanisms by which words are processed may not be the same mechanisms employed in facial recognition
Similar processing may be involved
CONNECTIONS TO PREVIOUS RESEARCH
FUTURE RESEARCH
Intentional vs. Incidental Previous works show there is no difference in
memory if the learning is incidental or intentional Craik, & Tulving, 1975
Examine if intentional or incidental learning has an effect on facial recognition
FUTURE RESEARCH
Facial images Examine the recognition of faces more typically
seen in the subject’s environment
Investigating recognition of faces of varying ethnicities
Neuroimaging: fMRI Examine areas of activation between different
questions Compare word processing to facial processing
top related