Title:
1That Swimsuit Becomes You Sex Differences in
Self-Objectification, Restrained Eating, and Math
Performance
2- Objectification Theory
- women, more so than men, are portrayed as
though their bodies were capable of representing
them. - Duncan, 1990 Fromme Beam, 1974 Gardner, 1980
Goffman, 1979 Soley Kurzbard, 1986
Frederickson Roberts, 1997 Van Zoonen, 1994
Bartkey, 1990 Henley, 1977 - posits that these objectifying views are
internalized, and, in anticipation of external
judgement, women learn to judge themselves. This
is called - SELF-OBJECTIFICATION
3- Self-Objectification can be either
- TRAIT
- Personality trait, e.g. how chronically focused
on appearance are you? - STATE
- When a specific context causes you to be
self-conscious, in an evaluative way
4The Problem! Consequences of Self-Objectification
- this vigilant self-monitoring drains mental
energy and consumes attentional resources from
important activities - manifested in diminished mental performance
- Increased shame and anxiety
- Disordered eating
5Experiment 1 HypothesisSelf-Objectification
produces BODY SHAME, which in turn predicts
RESTRAINED EATING
N 72 (all women)
- Trait self-objectification measured to determine
individual baseline - State self-objectification manipulated by
randomly assigning participants to experimental
condition (either swimsuit or sweater) - BODY SHAME measured in disguised questionnaire
- RESTRAINED EATING measured in consumer report
taste test
6BODY SHAME
- Variables
- Explanatory
- Categorical trait self-objectification and
experimental condition - Response
- Quantitative body shame
7RESTRAINED EATING
- Given 2 cookies and asked to evaluate.
- Experimenters wanted to measure how much
participants ate to determine relationship
between self-objectification, shame and eating. - Results - three response categories
- True restraint (ate less than ½ of 1 cookie)
- Symbolic restraint (more than ½, but still less
than 1) - No restraint (more than 1)
8Before we can look at the relationship between
s-o, body shame, and eating
- Potential Confounding Variable?
- People who liked the cookie more would eat more?
- To avoid this, researchers needed to show that
there is no relationship between amount eaten and
how much they cookie was liked
9- Mean liking of cookie (µ) 7
- I 3 (restraint, symbolic restraint, no
restraint) - N 72
- Null Hypothesis
- µr µs µn
- Alternative Hypothesis
- Not all the µ are equal
- P(2, 69) 2.5, p .0895
10P(2, 69) 2.5, p .0895
- Can we reject the Null Hypothesis?
- No!
- So we accept µr µs µn
- Since there is no relationship between µ, we can
rule out mean liking as a confounding variable - Conclude that amount consumed is due to other
factors (body shame)
11Results
- Highest body shame level most often predicted
(57) symbolic restraint group
12Experiment 2 Hypothesis Self-Objectification
diminishes math performance
- Direct Response to Experiment 1
- Replicate findings
- Extend tests into domain of attention and mental
performance - Address bias (of not representing whole
population) by testing men
13Experiment 2 Hypothesis Self-Objectification
diminishes math performance
N 82 (40 men, 42 women)
- Trait self-objectification measured to determine
individual baseline - State self-objectification manipulated by
randomly assigning participants to experimental
condition (either swimsuit or sweater) - Body Shame measured in disguised questionnaire
- NEW test of math performance (GMAT)
14Results
- State Self-Objectification
- Analysis of Covariance (ANCOVA) used to determine
validity of relationship between experimental
condition and self-objectification. - F(1, 73) 8.15, p lt .01
- Is there a relationship?
- YES!
15Results
- Body Shame
- ANCOVA again (BMI covariate)
- Categorical explanatory variables
- Experimental condition
- Trait self-objectification
- sex
- Significant relationships established between
each three variables and body shame
16Significant relationships established between the
explanatory variables and body shame
- Self-objectification as explanatory variable
- F(1, 73) 4.50 and p lt .05
- Experimental Condition as explanatory variable
- F(1, 73) 6.58 and p lt .05
- Gender split
- For Men, trait self-objectification was
explanatory - F(1, 35) 7.19 and p lt .05
- For Women, most significant relationship found
between experimental condition and body shame - F(1, 37) 5.83 and p lt .05
17(No Transcript)
18Math Performance
- To test the hypothesis that self-objectification
would lead to performance decreases, they
analyzed math scores using ANCOVA
19Results
- No significant relationship between experimental
condition and math score emerged for men. - For relationship between experimental condition
and math score for women - F(1, 32) 3.94, p .056
- I 2 What are the 2 groups?
- Sweater or swimsuit
- Do we reject Null Hyp. and conclude that there is
a relationship? - YES!
- Women in the swimsuit condition performed
significantly worse on the math test than women
in the sweater.
20(No Transcript)
21- The men in Experiment 2 served as a comparison
group to help establish that consequences of
self-objectification are not part of human nature
more generally but rather are specific to women
22Consequences of Type I Error
- For relationship between experimental condition
and math score for women - F(1, 32) 3.94, p .056
- We rejected the Null Hypothesis and concluded
that there is a relationship - BUT if there is no relationship
- Incorrect research conclusion, misleading
23Consequences of Type II Error
- Cookie liking test
- P(2, 69) 2.5, p .0895
- Did not reject Null Hypothesis
- Concluded there was no relationship between
amount of cookie eaten and how much they liked
the cookie - If Type II Error had occurred, and the
researchers failed to reject the Null Hypothesis
even thought it was false, then - Cookie liking would have been a confounding
variable - Possible incorrect conclusions to entire study!
- Amount eaten would not have been due to body
shame, but to how much they liked the cookie!
24BIAS?
- Participants were undergraduates at the
University of Michigan and Duke University - Non-representative of the larger population?