LISTSERV 15.5 - AEJMC Archives

Subscriber's Corner

Public Email Lists

AEJMC Archives

View:

Options:

Join or Leave AEJMC
Reply | Post New Message
Search Archives

Subject: AEJ 05 ChoiY MCS An Examination of third person effect with Q methodology: How does my ideal body image differ from the perceived ideal image of others?

From: Elliott Parker <[log in to unmask]>

Reply-To: AEJMC Conference Papers <[log in to unmask]>

Date: Mon, 6 Feb 2006 06:10:41 -0500

Content-Type: text/plain

Parts/Attachments:

text/plain (1990 lines)

This paper was presented at the Association for Education in Journalism and
Mass Communication in San Antonio, Texas August 2005.
If you have questions about this paper, please contact the author
directly. If you have questions about the archives, email
rakyat [ at ] eparker.org. For an explanation of the subject line,
send email to
[log in to unmask] with just the four words, "get help info aejmc," in the
body (drop the "").

(Feb 2006)
Thank you.
Elliott Parker
====================================================================

An Examination of third person effect with Q methodology:
How does my ideal body image differ from the perceived ideal image of
others?
by

Yun Jung Choi
Doctoral Student
&
Jong Hyuk Lee
Doctoral Student

S. I. Newhouse School of Public Communications
Syracuse University

*Student Paper (Leslie J, Moeller Award Competition)

Contact:
Yun Jung Choi
S. I. Newhouse School of Public Communications
Syracuse University
215 University Place
Syracuse, NY 13244-2100
[log in to unmask]

Manuscript submitted to the Mass Communication and Society Division
of the 2005 AEJMC, San Antonio, TX

An Examination of third person effect with Q methodology:
How does my ideal body image differ from the perceived ideal image of
others? Abstract
The third person effect was examined with the Q methodology.
Participants were asked to sort images of women to represent their
ideal image and their perception of other's ideal image. The third
person effect was observed in the study. People's their own ideal Q
sort loaded on one factor while their Q sorts representing their
perception of others' ideal image loaded on another factor. People's
own body image differs from their perception of others' ideal body
image that resembles media's portrayal of ultra thin models. For some
the third person effect was not observed, and one participant showed
the first person effect. Additionally, an intensive study, a study of
single subject was conducted to see how social distance mediates the
third person effect. The intensive study shows that the increasing
social distance reinforces the third person effects.

The basic concept of the third person effect is that people tend to
overestimate the mass media messages' influence on the attitudes and
behavior of others (Davison, 1983). People tend to think others are
more vulnerable to the mass media messages, such as news, television
commercials, pornographic images on the web, obscene lyrics, and TV
violence. Third person effect has been supported by both survey and
experimental studies (Gunther, 1991, 1992, 1995; Gunther & Thorson,
1992; Lasorsa, 1989; Perloff, 1989. McLeod, Eveland, & Nathanson, 1997).
This study tests the third person effect with the Q methodology.
Invented by a British Physicist-Psychologist William Stephenson
(1935), Q methodology provides a means to study people's
subjectivity. Q methodology identifies views of groups who share a
similar view toward an idea or a perspective by producing several
factors of opinions.
This study examines whether the third-person effect is observed when
people are asked to sort pictures of women according to their ideal
body image and their idea of others' ideal body image. If the two
images are the same, there is no third person effect and the Q
methodology would yield a same factor, but if the two ideal images
differ, the third person effect is present and several factors would
emerge. Two intervening variables, self-esteem and physique anxiety
scales were also asked to see how those two variables affect the
strength of the third person effect.
Additionally, an intensive study, a study of a single participant, is
conducted to explore the role of social distance in the third person
effects, which has been found in other survey studies (Brosius &
Engel, 1996; David & Johnson, 1998). One participant was asked to
sort Q items 14 times to represent her perception of the ideal body
images of others in a varying social distance from a close friend to
a person living in other continent. How those Q sorts are factored
together would illustrate how the social distance influences the
third person effect.
This study expands the literature of the third person effect by
testing the hypothesis with the Q methodology. A theory becomes
robust when it withstands tests of all methodologies. Q methodology
can capture more intervening factors of the third person effect that
are not easily observable in survey and experimental research. This
study is one of very first attempt to test the third person effect
with using image stimuli. The third person effect studies are
typically tested with survey and experiments methods where
participants are asked to rate the impact of media on themselves and
others on point scales from non-influence to considerable influence.
This study uses picture images of women as the Q sample. By asking
participants to sort images, people's subjectivity, their own
uniquely valid views and beliefs on body image can be captured more readily.

Theory
The third person effect
The third person effect posits that people think others are more
influenced by the media than themselves (David & Johnson, 1998). The
term, "third person" is derived from the expectation that a message
will not have a lot of influences on "me" (the first person), or
"you" (the second person), but on "them" (the third person). People
tend to underestimate the mass media effects on themselves but
overestimate the effects on others.
The third person effect hypothesis has been supported by various
methods with extensive topics. People are shown to estimate that
the news media have more impact on others' opinions of the 1996
presidential election than on their view (Salwen, 1998). In a study
of the O.J. Simpson trial, people thought that they were not
influenced by the media overage, but thought media exerted a great
influence on others' opinions of Simpson's guilt or innocent (Salwen
& Driscoll, 1997). The third person effect has been found with the
advertising messages. Gunther and Thorson (1992) found that people
think others are more influence by the liquor and bear commercials,
and household products than themselves. Other studies conducted with
entertainment messages suggests that people believe that others are
more influenced by watching the pornography than themselves (Gunther,
1995) and antisocial rap lyrics have more negative consequences on
others than themselves (McLeod, Eveland, & Nathanson, 1997).
Recently, scholars begin to acknowledge that the third person effects
have a social distance dimension. Studies found that as the social
distance between a respondent and a comparison group increases, the
third person gap increase. Studies have shown that people think the
general public is more vulnerable to the mass media messages than
their friends and acquaintance (Brosius & Engel, 1996), and a vague
person is more influenced by the media than a specific person, such
as a celebrity (Duck & Mullin, 1995).
Third person effects have been shown for the impact of mass media
images on ideal body images (David & Johnson, 1998). Female students
were asked to think about the effect of television programs such as
Baywatch or commercials with extremely attractive models on female
viewers' perception of ideal body image. Their evaluations of the
mass media's impact on themselves, their class mates, other women in
campus, and U.S. women in general were measured. A robust
third-person effect was observed. In addition, as the social distance
increased, stronger third person effects were observed.
Another study examined race as a variable in the third person effects
by asking white and black students to estimate the effects of
magazine ads with attractive models (David, Morrison, Johnson, &
Ross, 2002). The study found that both black and white students think
that others would identify more closely with same race models than
with different race models. Black respondents think that other black
women would be affected more when the race of the model is black than
when the model was white. Similarly, white students think that other
whites would be influenced more by white models than they would by
black models.
Mediating variables for the third person effect on body image have
been identified. David and Johnson (1998) found that self esteem and
physique anxiety have effects on the third person effect. Their
finding suggests that those with high self-esteem exhibit stronger
third person effect, whereas high physique anxiety is related to
lower third person effects when the topic in question concerns body images.

Women's body image
Ultra-thin body images dominate the mass media. A study found that
the yearly mean weight of Playboy models is significantly less than
the general public mean (Garner et al., 1980). Further more, since
1970s, Miss American pageant winners weigh significantly less than
the other pageant contestants (Garner et al., 1980). Women are also
expected to have slimmer figures than men in the society. In a
content analysis of 33 television shows, 68% of female characters are
rated thin compared to 17.5 % of men, whereas only 5% of the female
characters are rated heavy compare to 25% of men (Silverstein et al., 1986).
From the content analysis of women's image in television,
Silverstein and his colleagues conclude;
"present day women who look at the major mass media are exposed to a
standard of bodily attractiveness that is slimmer than that presented
for men and that is less curvaceous than that presented for women
since the 1930s. This standard may not be promoted only in the media
and it may not even originate in the media, but given the popularity
of television, movies and magazines� the media are likely to be among
the most influential promoters of such thin standards" (Silverstein
et al., 1986, p. 531).
Women in mass media are not only thin but also expose themselves a
lot. In 1991, about 31.9 percent of the 248 magazine ads showed a
body-revealing clothes or nudity (Kang, 1997). In the study, body
revealing clothes are defined as mini-skirts, tight skirts or evening
gowns that exposed cleavage, short-shorts, see-though clothes, halter
dress, or swim suits. Nudity was defined as unclothed models,
including models in translucent under wears and lingerie.
Soley and Reid (1988) compared nudity portrayed in magazines between
1984 with 1964. Six magazines, Esquire, Playboy, Redbook,
Cosmopolitan, Time and Newsweek were included in the analysis. The
results indicate that the ads in the 1984 magazines contained more
sexually explicit content, defined categorically as demure dress,
suggestive dress, partial dress or nude than the 1964 sample ads. In
1964, 5 % of the models were nude compared to 8% in 1984. In men's
magazines, of the 1984 sample, 43.7% of the ads depicted suggestively
clad, partially clad, or nude females.
These thin and over exposed body images of women in media are found
to influence the ideal body images of the media users. Myers and
Biocca (1992) found that exposure to ideal body images in the media
lead to changes in their own body image perceptions among college
women. They found that a distorted body image happens in two stages.
First, body image advertising makes young women feel thinner than
they normally do at first. However, when the upbeat feeling induced
by the advertisements disappears, women are faced with the cold
reality, a reality that are not congruent with their ideal. Harrison
and Cantor (1997) also found a correlation between the media exposure
and the drive for thinness which sometimes results in eating disorders.
In a focus group study, Goodman (2002) found that the influence of
thin images in the media on women tends to be a consequence of
long-term exposure that idealizes thinness, which shapes their
attitudes, beliefs, and behaviors to varying degree. Women idealize
thin bodies, consider media models as influences on their ideal body
image, and aspire to be like the ideal body image (Duke, 2000).
This aspiration to be like the models on mass media can bring
negative results on teenage girls and women. Studies show that the
mismatch between the ideal body image and the actual self can lead to
body image dissatisfaction (Richins, 1991), dissatisfaction with
one's attractiveness (Martin & Kennedy, 1993), and a drive for
thinness (Harrison & Cantor, 1997). In extreme cases, body-image
dissatisfaction or physique anxiety hurts an individual's self-esteem
or self-concepts, and even lead to behavioral consequences, such as
eating disorders (Thompson, 1995; Akan & Grillo, 1995).
Literature on ideal body image points out that the people's ideal
body image reflect the ideal body images portrayed in media.
Therefore, it can be assumed that people think others have ultra thin
and sexually seductive super model images as their ideal body images
as a consequence of a long exposure to the mass media.

Research Questions
According to the third person effect, people think that the media do
not have much impact on them, but others are influenced by the media
that others' ideal image resembles the ultra slim body types that are
suggested by the media, whereas my ideal body image is different from
the ideal body image reflected in the media. Thus people's own ideal
body image would be different from the perceived ideal images of the others.
The Q methodology groups people's ideas that share a similar view. It
can be assumed that when people are asked to sort Q samples twice
according to their own ideal body image, and again according to their
idea of others' ideal image, there would be discrepancies and more
than two Q factors would emerge.
Research questions are asked rather than hypotheses because the Q
method does not test hypotheses using inferential statistics, but
shows how many and what kinds of factors of views exist in human
subjectivity.

RQ1: How do people sort their ideal body images and other people's
ideal images?
How does people's their own ideal body image differ from their
perception of others' ideal body image?

Self-esteem and physique anxiety are found to be related to the third
person effects of the ideal body image (David & Johnson, 1998).

RQ2: How does self-esteem influence the third person effect of the
ideal body image?

RQ3: How does physique anxiety influence the third-person effect of
the ideal body image?

Finally, the relationship between the third person effect and the
social distance is examined.

RQ4: How is the third person effect mediated by the social distance?

Method
Q methodology
Q methodology is used in this study to test the third person effect.
Invented by British Physicist-psychologist William Stephenson (1953),
Q methodology is associated with factor analysis. Respondents sort a
deck of cards call Q items, and their sorts are factor analyzed to
draw out correlations among the respondents (Kerlinger, 1986). This
way attitude structures that exist within individuals or groups can
be discovered.
In Q studies participants are asked to rank order stimulus items (Q
sample) to some instructions, e.g., from "most appealing" to "most
unappealing" (Kinsey, 1993/1994). After participants sort Q samples
to represent their own viewpoints, the data are factor analyzed. In
usual factor analyses, participants' answers are factored together so
that certain perceptions or attitudes are grouped together to
represent a factor, but Q factor analysis involves factoring of
individuals who share similar views. People who have sorted the items
in a similar fashion will load together on a factor. Each factor
represents a point of view or shared opinions about the topic asked.
Each factor also represents broad patterns that can be found in the
larger population.

Q sample
Typically Q samples are sampled from a concourse, a volume of
discussion on any topic (from the Latin Concursus, meaning "a running
together"). The Q sample is important in Q methodology because it is
these items of opinions or ideas that are later grouped together in a
factor to represent a point of view. It is important to represent
comprehensive and diverse views that exist in the population in the Q sample.
For this study, 36 images of women were sampled to represent images
of women. Literature on women's body image in the media suggests that
women's ideal image is represented in two dimensions � thinness and
nudity (Kang, 1997; Soley & Reid, 1988; Garner et al., 1980). This
study uses a structured sampling method with a 4 (four levels of body
types) _ 3 (three levels of skin exposure) configuration. The 4
levels of body types are skinny, slim, normal, and heavy, and the 3
levels of skin exposure are a lot of exposure, normal exposure, and
no exposure. As Figure 1 shows, three images of women are sampled for
each 12 cell according to the sampling structure.
[Insert Figure 1 about here]
Image of skinny models and pictures of celebrity who are criticized
for being too skinny are included in the skinny category. Typical
slim models with slender figures are included in the slim category,
whereas images of women who are not necessarily considered slim by
the media standard but have an average weight are included in the
normal body type category. Those images of women in plus size clothes
catalogues are included in the heavy body type category.
In terms of the exposure, a lot of exposure is defined as women
clothed in bikinis or swim suits, normal exposure is defined as women
in short sleeves, tank top shirts, or summer dresses that have some
exposure of skin. No exposure is defined as women in formal suit or
winter clothes that do not have skin exposure.
Images were gathered from the internet. Celebrity pictures, images
from clothes catalogues, and magazine pictures such as Sports
Illustrated swimsuit edition are included in the Q sample. In order
to reduce the bias caused by the attractiveness of the facial
features, model's faces were filtered with haze using Photoshop. The
sampled images are shown in Figure 1.

Participants
The participants, so called P sample, in this study are 23 people,
which is a typical number of participants in Q studies. Participants
composed of 10 males and 13 females, and their age ranged from 23 to
65. Eighteen of them are graduate students at a northeastern university.

Q sorting procedure
Participants were asked to sort the Q samples twice according to
their own ideal body image and their idea of other people's ideal
body image. At first, participants are asked "what is your view of
the ideal body image? Please rank the picture from the most appealing
to the most unappealing." After they sorted the images, participants
were given another question, "How do you think most other people
would sort these images to represent their ideal image?"
The Q sort statements are conventionally arrayed in a forced,
quasi-normal distribution. In this study, a forced distribution with
nine-point scale was used (Figure 2). The images that they think
represents their ideal image were given +4., whereas the images that
are most unappealing to them are given the �4 score.
[Insert Figure 2 about here]
After participants sorted the images, self-esteem and physique
anxiety of the participants were measured using the Self-esteem scale
(Rosenberg, 1965) and the Social Physique Anxiety Scale (Hart, Leary,
& Rejeski, 1989). The self-esteem scale has 10 statements, which are
rated on a 4 point scale from 1 (strongly agree) to 4 (strongly
disagree). The social physique anxiety scale also has 12 statements,
which are rated on a 5-point scale ranging from 1 (not at all) to 5
(extremely). Additionally, respondents' age, race, occupation, and
media uses were asked (see Appendix A).

Analysis
Q factor Analysis was conducted with a statistical software, SPSS
(Statistical Package for the Social Sciences). The principal
component method and Varimax rotation were used to factor
individuals. Each 23 participants produced two Q sorts, therefore a
total of 46 Q sorts were factor analyzed together. The SPSS provided
6 factors with the eigen-value greater than 1.0. However, not many
people are loaded on the 4th, 5th, and 6th factor, and most
significant loadings in those factors were already captured by the
first three factors, therefore only three factors were chosen as the
final Q factors for the study.

Results
Three factors, Factor A, B, and C were produced as the results.
Eigen-vales for the three factors, A, B, and C, are 13.30, 6.54 and
6.42, and the explained variance are 28.29, 14.21, and 9.97. The
correlation coefficients between the three factors are all less than .04.
As the factor matrix table, Table 1 shows, 11 people loaded on Factor
A for their perception of others' ideal body image, and the same 11
people loaded on Factor B for their ideal body image, although some
people are confounded on other factors. Four people loaded on Factor
A for both their own ideal body image and others' ideal body image.
Three people loaded on Factor A for others' ideal image whereas their
own ideal body image did not load on any factor. Two people loaded on
Factor C for their ideal image and on Factor A for others' ideal
image. Two people loaded on Factor C for both their own ideal body
image and other's ideal image. One person did not load on any factor.
The factor scores of the each Q sample images are shown in Tale 2.
[Insert Table 1 about here]
[Insert Table 2 about here]

The mean score for the self-esteem was 3.21, out of the highest
possible score being 4.00, with the standard deviation of .41. The
mean score for the physique anxiety was 3.37 with the standard
deviation of .60. The highest possible score for the physique
anxiety, the score 5.00 represents the lowest physique anxiety.

Factor A: Models in bikinis
Factor A is the sports illustrated swim suite factor. As shown in
table 3, the two top rated images are from the sports illustrated
swimsuit edition that belongs to the slim body type with a lot of
exposure category. All six images of slim and skinny body types in
swim suits are ranked high in this factor. The slim body type is
rated higher than the skinny body type. Images of women in the no
exposure category were also not rated high in this factor.
[Insert Table 3 about here]
The heavy body type is rated as the most unappealing in Factor A. All
nine images of heavy body type women regardless of the exposure
levels are rated low in this factor. Interestingly heavy women with
moderate or no exposure are rated higher than the heavy women with a
lot of exposure. In sum, Factor A represents the media's typical
portrayal of ideal women's body image. Skinny and slim body images
with a lot of exposure, which are often found in women's magazines
and TV ads, are rated high in this factor.
Factor A is also the others' ideal body image factor. 20 people out
of 23 considered Factor A as their perceptions of others' ideal body
image. Those who loaded in this factor as others' ideal body image
are likely to believe that the media shape other people's perceptions
of ideal body image.

Factor B: A girl next door
Two images in the normal body type with normal exposure are rated the
highest in this factor (Table 4). The second most appealing body
images are the mix of slim body types in bikinis, a normal body type
with no exposure, and a skinny body type with no exposure. The normal
body type of women with normal exposure is rated the highest while
images of women with a lot of exposure are not considered very
appealing in this factor. The two most unappealing images in Factor B
are the two skinny women in bikinis. In this factor, three images of
heavy women in swimsuits were rated more appealing (� 2) than the
skinny body type with a lot exposure and moderate exposure (-3).
[Insert Table 4 about here]
Factor B represents most people's their own ideal body image. Eleven
people loaded in this factor as their ideal body image. All 11 people
loaded in this factor also loaded on Factor A for their idea of
others' ideal image. People loaded in this factor as their own ideal
body image have ideal body images that are more realistic than the
ultra thin and slim figures proposed by the media, yet still desirable.

Factor C: Traditional beauty
Two images of women with the normal body type in swimsuits received
the highest score in this factor (Table 5). Three women of the normal
body type with some exposure and no exposure received the second
highest score in this factor. In this factor, women in the normal
body type category are considered the most appealing. Slim women
received the next highest scores followed by the heavy and the skinny
body type. The skinny body type is rated the least appealing in this
factor regardless of the exposure levels. Three images of heavy women
in swimsuits received -2, 0, and -1 scores, which can be interpreted
as moderately appealing. This indicates that Factor C is generous to
the heavy body type.
Factor C can be labeled as the elderly factor because two eldest
participants, aged 65 and 62, are loaded together in this category.
The two participants loaded on Factor C for both their own ideal body
image and others' ideal body image.
[Insert Table 5 about here]

Intensive Study
Previous studies show that the third person effect gap widens as the
social distance of the others increases (Brosius & Engel, 1996; David
& Johnson, 1998). In order to test the role of social distance in the
third person effect, an intensive study was conducted one month
later. An intensive study is a study of a single subject. A single
person is asked to sort the Q sample under different conditions. It
"reflects interest in 'intrasubjectivity,' that is, in an in-depth
examination of one person who sorts the Q-sample under many different
conditions of instruction" (McKeown & Thomas, 1988, p. 37).
One participant was chosen to participate in this intensive study
because her Q sorts on ideal body image and her perception of other's
ideal body imaged loaded on Factor A and Factor B respectively
without any confounding which indicates a typical third person
effect. Also, this participant was not exposed to the purpose of the
study after one month.
The social distance was considered in two dimensions in this study �
geographical distance and gender difference. This participant
recently moved from South Korea where she worked as a PR
practitioner, therefore, her home town, Pusan was considered as the
starting point in calculating the social distance. To vary the social
distance, her perceptions of the ideal images of a family member, a
close friend, a colleague at work, a person living in Pusan (her
hometown), a person living in Seoul (the capital of the country where
she used to live), a person living in an Asian country, and a person
in the other continent were asked. This study is about media's
portrayal of ideal women's body image and its influence on viewers',
therefore the other dimension of social distance is considered as the
gender. The set of seven questions was asked twice for both cases
where the hypothetical person being a female and a male. Therefore
the participant was asked to sort the Q sample 14 times.
In addition to 14 Q sorts obtained from the intensive study, two Q
sorts reflecting 'my own ideal image,' and 'my perception of others'
ideal image' obtained from the previous study were included in the
analysis. Therefore a total of 16 Q sorts were factor analyzed. A two
factor solutions emerged with a verimax rotation. Eigen-values for
Factor A-I and B-I are 7.0 and 6.9 respectively.
Table 6 shows the factor loading. A male colleague, a man living in
an Asian country, and a man in the other continent were factored
together in Factor A-I. On Factor B-I, myself, two family members, a
female member of the family (sister), a male member of the family
(brother), and a man living in Pusan, her hometown are factored
together. A total of 9 Q sorts were confounded on the two factors.
Although it is statistically non-significant, among the confounded Q
sorts, socially distanced people, others, a man living in Seoul,
women living in an Asian country and other continent, and a male
friend showed higher factor loading coefficients for Factor A-I than
B-I. The participants' perception of ideal images of women who have
lesser social distance from the participant, a female friend, a
female colleague, a woman living in Pusan, her hometown, and a woman
living in Seoul showed higher factor coefficients for Factor B-I than
for Factor A-I.
[Insert Table 6 about here]
The overall pattern of the intensive study showed an evidence of the
social distance effect in the third person effect theory. Those close
to her socially and psychologically are thought to be sharing the
same ideal body image with her own ideal image, while those who have
the some distance with her, a male colleague, men living other
countries loaded on a factor that shares a different ideal body image
from hers. The pattern is clear. The more socially distance a person
is, the stronger the third person effect is shown so that the
perceived ideal body of those socially distanced people resembled the
ideal body image suggested by the media.

Factor A-I: Women in swimsuit
This factor is very similar to Factor A shown in the previous study.
The two top ranked ideal body images are the same (Table 7). One
difference from the previous Factor A is that women in swimsuits
regardless of body types received high rating. Even heavy body type
women are moderately rated when they were in swim suits (- 2). The
Ultra thin body type is considered appealing this factor while the
normal body and the slim body types with normal exposure receive
medium ratings. Heavy women, when they were not in swimsuits, scored
the lowest..
[Insert Table 7 here]
Factor B-I: Women in full suits
The factor B-I represents my ideal body image and is very similar to
Factor B in the previous study (Table 8). The normal and slim body
types with normal to no exposure are rated high in this factor. One
difference from the previous Factor B is that women in no exposure
category are rated high in this factor. Women with no exposure were
well liked while women in swimsuit are not rated high in this factor.
The highest rating for bikini is 1, and the two top rated bikini
models for Factor A received moderate ratings (0). The thin body type
is rated low but not as low as the heavy type.
[Insert Table 8 here]

Discussion
This study examined the third person effect hypotheses with the Q
methodology. Unlike other methods that use referential statistics, Q
methodology does not tell whether a hypothesis is supported or not,
but reveals several factors that exist in people's opinion. Hall
(1997) suggested that members of the same culture must share sets of
concepts, images, and ideas which enable them to think and feel about
the world, and thus to interpret the world in roughly similar ways.
Several shared sets of concepts are observed in this study through Q sorting.
First, the third person effect is observed. A lot of people's their
own ideal Q sorts are loaded on Factor B while their Q sorts
representing their perception of others' ideal body image loaded on
Factor A. People's own ideal body image differs from their
perceptions of other people's ideal image. The fact that factor array
for Factor A resembles the ideal body image portrayed in the mass
media suggests that people think other people are more influenced by
the media messages than themselves. Most people chose body images
different from the media's ideal body image as their ideal body image.
The result also shows that people's own ideal images are diverse
while their opinion of others' ideal body image is uniform. Two
factors, Factor B and Factor C represent people's ideal body image,
whereas 20 people out of 23 people loaded on Factor A for others'
ideal body image. Regardless of what their own ideal body image is,
people's opinions of other people's ideal body image are loaded on Factor A.
It is important to note that several people did not show the third
person effect. Four people loaded on Factor A for both their ideal
body image and their idea of others' ideal body image. Two people
loaded on Factor C think their own ideal image and others' ideal
image are similar. This finding indicates that the third person
effect is not applicable to all people. It can be inferred from the
result that there are some people who think they are influenced by
the media as much as other people.
Some people showed the first person effect, which hypothesizes that
people think they are more influence by the media than others. As
Table 1 indicates, one person's (ID 12) Q sort for his ideal body
image loaded on Factor A with higher coefficients (.70) than his Q
sort for others' ideal image (.53). His opinion of other people's
ideal body image is also confounded on Factor A and Factor B. This
man clearly rates those ideal image of women suggested by the media
as his ideal body image while considers more ordinary realistic
images of women as others' ideal image. It is interesting to note
that this single person is a male.
Factor C indicates that the elderly have their own distinct idea of
ideal image that is different from young people's ideal body image.
The fact that their own ideal body is not different from their
perceptions of others' ideal body indicates that those elderly people
are less sensitive to media messages. They are the only two people
whose Q sorts for others' ideal image are not loaded on Factor A.
This indicates that there are some who are indifferent to the media.
Both their own ideal body image, and their thoughts on others' ideal
body image are different from the standard ideal body image suggested
in the media.
The idea that increase social distance reinforces the third person
effect has been confirmed through the intensive study. The fact that
the participant herself, members of her family, and a person living
in her home town loaded together in a factor, and socially and
physiologically distanced people, a male colleague and men living in
other countries are factored together on the other factor shows that
the social distance variable mediates the third person effect.
Lastly, the roles of self-esteem and physique anxiety in the third
person effect are not identified in this study. No clear pattern is
found. The two variables were found to influence the strength of the
third person effect when the media's influence on people's ideal body
weight and behavioral intentions, such as when eating disorders are
analyzed (David & Johnson, 1998). This study is more focused on
women's sexual images rather than weight. Perhaps different context
made the effect of the two measures unobservable.
There are some limitations in the study. The dresses women wore in
the Q sample might have influenced the Q sorting process. Some people
prefer some styles of clothing while others prefer other styles of
dresses. The fashion styles in the Q sample should be controlled in
future studies.

References

Akan, G. E., & Grillo, C. M. (1995). Sociocultural influences on
eating attitudes and behaviors, body image, and psychological
functioning: A comparison of African-
American, Asian-American, and Caucasian college women. International
Journal of Eating disorders, 18(2), 181-187.

Brosius, H. B., & Engel, D. (1996). The cause of third-person
effects: Unrealistic optimism, impersonal impact, or generalized
negative attitudes towards media influence? International Journal of
Public Opinion Research, 8, 142-162.

David, P., & Johnson, M. A. (1998). The role of self in third-person
effects about body image. Journal of Communication, 48(4), 37-58.

David, P., Morrison, G., Johnson, M. A., & Ross, F. (2002). Body
image, race, and fashion models: Social distance and social
identification in third person effects. Communication Research, 29, 270-294.

Davison, W. P. (1983). The third person effect in communication.
Public Opinion Quarterly, 47, 1-15.

Duck, J. M., & Mullin, B. (1995). The perceived impact of the mass
media: Reconsidering the third person effect. European Journal of
Social Psychology, 25, 77-93.

Duke, L. L. (2000). Black in a Blond World: Race and girls'
interpretations of the feminine ideal in teem magazines. Journalism &
Mass Communication Quarterly, 77, 367-95.

Garner, D. M., Garfinkel, P. E., Schwartz, D., 7 Thompson, M. (1980).
Cultural expectations of thinness women. Psychological Reports, 47, 483-491.

Goodman, J. R. (2002). Flabless is fabulous: How Latina and Anglo
women read and incorporated the excessively think body ideal into
everyday experience. Journalism & Mass Communication Quarterly, 79
(3), 712-727.

Gunther, A .C. (1992). Biased press or biased public: Attitudes
toward media coverage of social groups. Public Opinion Quarterly, 5, 147-167.

Gunther, A .C. (1995). Overating the X-rating: the third person
perception and support for censorship of pornography. Journal of
Communication, 45(1), 27-38.

Gunther, A. C. & Thorson, E. (1992). Perceived persuasive effects of
product commercials and public service announcements: Third-person
effects in new domains. Communication Research, 19, 574-596.

Gunther, A. C. (1991).What we think others think: Causes and
consequences in the third person effect. Communication Research, 18, 355-372.

Hall, S. (1997). Representation: Cultural representations and
signifying practices. Thousand Oaks, CA: Sage Publications.

Harrison, K., & Cantor, J. (1997). The relationship between media
consumption and eating disorders. Journal of Communication, 47(1), 40-67.

Hart, E. A., Leary, M. R., & Rejeski, W. J. (1989). The measurement
of social physique anxiety. Journal of Sports Psychology, 11, 94-104.

Kang, M. (1997). He portrayal of women's images in magazine advertisements:
Goffman's gender analysis revisited. Sex Roles. 37, 979-996.

Kerlinger, F. N. (1986). Foundations of behavioral research (3rd
ed.). New York: Holt, Rinehart and Winston.

Kinsey, D. F. (1993/ 1994). Humor communicability. Operant
Subjectivity, 17 (1/2), 49-61.

Lasorsa, D. L. (1989). Real and perceived effects of "Amerika."
Journalism Quarterly, 66, 373-378, 529.

Martin, M. C., & Kennedy, P. F. (1993). Advertising and social
comparison: Consequences for female preadolescents and adolescents.
Psychology and Marketing, 10, 513-530.

McKeown, B. F., & Thomas, D. B. (1988). Q methodology. Newbury Park, CA: Sage.

McLeod, D. M., Eveland, W. P., & Nathanson, A. I. (1997). Support for
censorship of violent and misogynic rap lyrics: An analysis of the
third-person effect. Communication Research, 24, 153-174.

Meyers, P. N., & Biocca, F. A. (1992). The elastic body image: The
effect of television advertising and programming on body image
distortion in young women. Journal of communication, 42(3), 108- 133.

Perloff, R. M. (1989). Ego-involvement and the third person effect of
televised news coverage. Communication Research, 16, 236-262.

Richins, M (1991). Social comparison and the idealized images of
advertising. Journal of Consumer Research, 18(1), 7-83.

Rosenberg, M. (1965). Society and the adolescent self-image.
Princeton, NJ: Princeton University Press.

Salwen, M. B. (1998). Perceptions of media influence and support for
censorship: The third-person effect in the 1996 presidential
election. Communication Research, 25, 259-285.

Salwen, M. B., & Driscoll, P. D. (1997). Consequences of third-person
perception in support of press restrictions in the O. J. Simpson
trial. Journal of Communication, 47(2), 60-75.

Silverstein, B., Perdue, L., Peterson, B., & Eileen, K. (1986). The
role of the mass media in promoting a thin standard of bodily
attractiveness for women. Sex Roles, 4, 519-532.

Soley, L., & Reid, L. N. (1988). Taking it off: Are models in
magazines ads wearing less? Journalism and Quarterly, 65(4), 960-966.

Stephenson, W. (1935). Technique of factor analysis. Nature, 136, 297.

Stephenson, W. (1953). The study of behavior: Q technique and its
methodology. Chicago: University of Chicago Press.

Thompson, J. K. (1995). Assessment of body image. In D. B. Allison
(Ed.), Handbook of assessment methods for eating behaviors and
weight-related problems: Measures, theory, and research (pp.
119-144). Thousand oaks, CA: Sage.

Figure 1. Q sample structure and samples

Four body types
Skinny
Slim
Normal
Heavy
Three levels of exposure
A lot of exposure