HI‘HI‘HHWIHIWH

 

This is to certify that the
thesis entitled

VISUAL POLITICS:
CAMERA ANGLE AND FEMALE LEADERSHIP

presented by
Chou Chea

has been accepted towards fulﬁllment
of the requirements for the

MA. degree in Communication

\eikﬂﬁ

Uajor Professor’béignature

32-03-07

 

 

Date

MSU is an aﬁ‘irmative-action, equal-opportunity employer

_.___.__————._

 

 

UBRARY'
Mi”“~‘gqn State
University

 

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:/ClRC/DaleDue.indd-p.1

 

VISUAL POLITICS:
CAMERA ANGLE AND FEMALE LEADERSHIP

By

Chou Chea

A THESIS

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

MASTER OF ARTS
Department of Communication

2007

ABSTRACT

VISUAL POLITICS:
CAMERA ANGLE AND FEMALE LEADERSHIP

By
Chou Chea

This thesis examines how variations of camera angle (medium, high, low)
might affect judgments of female leadership competence in the political arena. It has
been believed that a low angle perspective helps accentuate the power of the
photographed person, whereas a high angle perspective projects the weakness of the
photographed person and a medium angle underlines the parity between the
photographed person and the viewers. These speculations have long been tested and
practiced in ﬁlm and photography industry, yet almost no studies to date have
investigated this potential, particularly as it applies to female political leaders. This
thesis tests if the camera angle effects remain when judgments of a female rather than
a male leader are introduced. One female graduate was chosen and photographed
from the three different camera angles. She was identiﬁed as a political candidate
running for a mayor ofﬁce in Portland, Oregon. One-hundred-and-ﬁﬁy-six students
rated her on a leadership competence scale. It was found that the assimilated female
leader was more positively judged when photographed ﬁom a slightly low camera
angle than from either a medium or a high angle perspective, and the implications of

the ﬁndings are discussed.

DEDICATION

To my parents, my grandparents, and my best friend, S. C. M.

ACKNOWLEDGEMENT

I would like to express my deepest gratitude to Dr. John L. Sherry, my
respected advisor, who had taught me a lot during my one year and a half at MSU and
who had guided and helped me throughout my thesis process despites his hectic
schedule. I would also like to thank Prof. Ronald Tamborini and Prof. Stephen Lacy
for all their thoughtful comments in regard to this thesis. Prof. Steven Lacy had been
particularly very helpful in ﬁnding me a cameraman. I would also like to thank my
god-aunt Ming Shally, god-uncle Pu Sovann, the coordinator of sponsored student
program Mary Gebbia, and the IIE-Fulbright program Officer .10an Moore, all of
whom have helped made my stay and study in the US a very delightful and
memorable experience, surely despite the cold weather I ﬁrst experienced. I would
also extend my gratitude and love to all my dearest friends here and also those back in
my country who have always been there listening to my concerns and offered me
great advice. Most of all, I would like to thank my most beloved parents and
grandparents; my memory of their life stories has always given me strength to go

through whatever obstacles I face.

iv

TABLE OF CONTENTS

LIST OF TABLES ................................................................... vii
LIST OF FIGURES .................................................................. viii
CHAPTER 1
INTRODUCTION .......................................................................................... 1
CHAPTER 2
LITERATURE REVIEW ............................................................ 3
Section 1
Tests ofCamera Angle Effects . ......3
Section 2

Visual Bias in Politics:

The Practice of Camera Angle Effects in

Political Joumalrsm9
Section 3

Female Political LeadershrplS

CHAPTER 3
METHODOLOGY ......................................................................................... 20
Overview ...................................................................... 20
Participants ................................................................... 20
Manipulations: Confederate and Camera Angles ....................... 22
Measures ..................................................................... 24
Assessment of Perceived Leadership Competence. . . . . . . . .....24
Additional Variables ............................................... 25
Subject Attributes ................................................... 25
Procedure ........................................................................ 26
CHAPTER 4
RESULTS ....................................................................................................... 28
Perceived Leadership Competence .................................................... 28
Camera Angle and Perceived
Leadership Competence ........................................... 28
Biological Sex and Perceived
Leadership Competence ........................................... 3O
Biological Sex*Camera Angle and
Perceived Leadership Competence .............................. 31
Perceived Attractiveness .................................................................... 32
Perceived Attractiveness and
Perceived Leadership Competence ............................... 32
Camera Angle and Perceived
Leadership Competence ............................................ 32

CHAPTER 5

DISCUSSION ................................................................................................. 34

CHAPTER 6

CONCLUSION .............................................................................................. 39

APPENDICES ................................................................................................ 40
Informed Consent Form ..................................................................... 41
Measurement. ..................................................................................... 42
Debrief Form ...................................................................................... 45
Photographs ........................................................................................ 46

BIBLIOGRAPHY .......................................................................................... 47

vi

LIST OF TABLES

Table 1
Summary of Camera Angle Tests ................................................................... 7

Table 2
Biological Sex‘Camera Angle Crosstabulation ............................................. 20

Table 3
The Trend of Participants’ Political News Consumption .............................. 21

Table 4
Comparison of Mean and Standard Deviation
of the Five Mediums ....................................................................................... 21

Table 5
Mean, Standard Deviation, and Corrected Correlation
with Total Scale Score of Competence-Scale Items ...................................... 25

Table 6
Camera Angle Effects on Perceived
Leadership Competence ................................................................................. 29

Table 7
The Effects of the Raters’ Biological Sex on
Perceived Leadership Competence ................................................................ 30

Table 8
The Effects of the Raters’ Biological Sex‘Camera Angle on
Perceived Leadership Competence ................................................................ 31

Table 9
Camera Angle Effects on Perceived Attractiveness ....................................... 33

vii

LIST OF FIGURES

Figure 1
Camera Angle Effects on Perceived
Leadership Competence ................................................................................. 29

Figure 2
The Effects of the Raters’ Biological Sex on
Perceived Attractiveness ................................................................................ 30

Figure 3
Biological Sex*Camera Angle Effects on
Perceived Attractiveness ................................................................................ 31

Figure 4
Correlation Check of Perceived Attractiveness and
Perceived Leadership Competence ................................................................ 32

Figure 5
Camera Angle Effects on Perceived Attractiveness ....................................... 33

viii

CHAPTER 1
INTRODUCTION

People form impressions about other people’s abilities, intelligence, integrity,
and honesty (Peterson, 2005). These impression formations of another person’s traits,
when it comes to election decisions, are one of the few central components among
party affiliation and political philosophy, which are used to guide vote choice. Bishin,
Stevens, and Wilson (2006) found that these character evaluations are statistically and
substantively signiﬁcant predictors of the vote. One common way people form
impressions about a political candidate is through the information content about those
traits of the candidate which they receive from the media (Peterson, 2005) and which
will be ﬁltered by their cognitions. Research in political communication has mainly
focused on the inﬂuence of verbal media text (either written or presented) on voters’
perception of a candidate’s trait and vote choice by analyzing campaign issues and
campaign debates in both press and electronic media (Peterson, 2005). However, until
recently much less research has focused on the more subtle power of photographs
despite the fact that people’s visual perception has been characterized to be passive
yet automatic and inﬂuential because it is a natural capacity based in both biology and
culture (Scott, 1994).

The visual image of the media can frame an empirical image through its
combination of elements captured in the image, the lighting, and the camera angle
(Etrnan, 1993). The combinations suggest whether the viewer should adopt a
sympathetic, respectful, disdainful, or some other attitudes toward the subject. Theory
of camera angles emerged as a ﬁlm theory and applies speciﬁcally to camera angles
as the impact. It posits that pictures taken from a low angle would be seen to be
powerful and dominant, pictures taken from a high angle would be seen to be weak,

and pictures taken from a medium angle would represent parity between the viewer

and the portrayed objects or people (Huss & Silverstein, 1968). This theoretical
assumption of media visual inﬂuence on viewer perception is widely accepted;
however, the nature of that inﬂuence remains speculative and can only be inferred
from limited empirical studies (Mandell & Shaw, 1973; Kappas, Hess, Barr, & Kleck,
1994; Tiemens, 1970; and McCain, Chilberg, & Wakshlag, 1977). Moreover, among
these studies, only male confederates were used in the experiments despite (1) the
possibility that perception toward a person’s attributes might be moderated by his/her
expected gender roles and (2) the current trend that more women are becoming the
public ﬁgures and more photographs of female leaders are taken.

The main goal of the present study then is to test the effects of different
camera angles on viewers’ perception of female leadership competence in politics and
to ﬁnd out whether the sex of the raters will inﬂuence their ratings of the competence.
In order to accomplish this, literature on theory of camera angle effects, visual bias in
politics, and female leadership issues will ﬁrst be discussed to develop well-formed

hypotheses to be tested in the proposed experimental study.

CHAPTER 2
LITERATURE REVIEW

Section 1
Tests of Camera Angle Effects

Since the earliest days of ﬁlm, the use of vertical camera angle has been part
of a director’s repertoire (Huss & Silverstein, 1968). Edmons (1982) traces the
conventional meaning of low angle shots back to childhood experiences of power and
status. He reasoned that the ﬁrst status ﬁgures in a child’s social development are
parents. The child’s physical relationship to these status ﬁgures is oﬁen portrayed in
children’s drawings: adult bodies are represented disproportionately with small heads,
huge bodies, and long legs. Edmons went on to hypothesize that when the camera
simulates this perspective, the audience associates with these childhood memories and
attributes power, dominance, and status to those objects that are presented through
low angle camera shots. On the other hand, high angle shots give power to the
viewers and attributes weakness to portrayed objects or people, and eye-level shots
represents parity between the audience and portrayed objects or people. Quite a few
experimental studies have tested the camera angle hypotheses, but the results are
somewhat varied in terms of the effects produced by the camera angles (Mandell &
Shaw, 1973; Kappas, Hess, Barr, & Kleck, 1994; Tiemens, 1970; and McCain,
Chilberg, & Wakshlag, 1977).

Mandel] and Shaw (1973) conducted an experiment with 143 undergraduate
students to ﬁnd out whether television images could subconsciously inﬂuence
judgments about a person presented in a short newscast by using subtly different
camera angles and slight variations in degree of bodily activity. They created a
simulated newscast, which included a story about the appointment of a chairman to a

state committee (photographed from three different camera angles, high, medium, and

low angle, and across two conditions, slightly active or no movement), among ﬁve
national and three local stories and a weather report. Subjects rated the model in low
angle condition (12 degrees below eye level with no concern to movement condition)
to be more potent (p < .05) and more active (p < .025) than those in medium angle
condition and high angle condition (12 degrees above eye level) respectively. The
model’s movement does not affect ratings of his potency at all (n.s.), and no
signiﬁcant difference was found on the evaluation of the model’s credibility and
trustworthiness across different camera angles.

Kappas, Hess, Barr, and Kleck (1994) used video records of real faces across
three camera angles to study the inﬂuence of camera angles on the attribution of
emotional expressions and attitudes of the photographed person. They ﬁlmed two
male and two female actors with three video cameras (eye-level, 40 degree from
below, and 40 degree ﬁ'om above) while displaying a neutral, a happy, and a
disgusted facial expression. The cameras were positioned at the same distance from
the face of the actors. Sixteen undergraduates rated the photos. They found that facial
expressions ﬁlmed from below were perceived as more positive and less negative than
expressions ﬁlmed from above, but no signiﬁcant difference was found on the
perceived attitudes (i.e., dominance and submissiveness). Sex of the raters had no
effect on the rating, and the data supports the notion that viewing angle introduces a
perceptual bias in the attribution of emotions to a stimulus person when his role is not
identiﬁed, but not the notion that vertical viewing angle would have an inﬂuence on
the perceived attitudes of the person.

Tiemens (1970) conducted an experiment with 99 undergraduate students to
ﬁnd out how communicator credibility (communicative ability, knowledge, authority,

and convincing ability) might be affected by camera angles. He ﬁlmed three male

speakers, each from three different camera angles, while presenting three different
newscasts. He kept constant the background, the position the speakers sat, the length
of the speech, and only let the camera angle vary. He found no signiﬁcant support to
the principle that low camera angles (approximately 18 degrees below eye level) and
high camera angles (approximately 18 degrees above eye level) inﬂuence the
perceived credibility of the male confederate newscasters, compared to the medium
camera angle. However, the results are questionable due to the use of three different
newscasts, the contents of which varied.

Inﬂuenced by and unsatisﬁed with Tiemens’ insigniﬁcant ﬁndings of camera
angle effects on perceived credibility, McCain, Chilberg, and Wakshlag (1977)
conducted two experimental studies to reexamine the effects of high and low camera
angle on a televised speaker’s source credibility, together with his interpersonal
attraction. In the ﬁrst study, four confederates (two males and two females) were
photographed while presenting the same persuasive message on whether or not the
college’s grading system should be revised. The photographs were medium shots
from the waist to the head of the same size across three camera conditions (eye-level,
23 degrees above, and 23 degrees below) and across ﬁve distances between the
camera and the speaker (24 feet, 20 feet, 16 feet, 12 feet, and 8 feet). Three hundred
and sixty undergraduate students viewed only one picture in each condition. Unlike
Tiemens (1970), McCain and his colleague found that higher angles consistently
produced higher credibility ratings (competence, composure, sociability, and
dynamism) of the assimilated speakers than medium and low angle shots (p<.05).
These ﬁndings were somewhat contrasting to the original camera angle theory, which

attributes all the positive judgments to low camera angle shots.

To further examine their ﬁrst ﬁndings, McCain and his colleagues conducted
the second study. In this study with 144 undergraduate students, they photographed
two male newscasters across three different camera angles (eye-level, 38 degrees
above, 38 degrees below the eye level) and across two image size conditions (close-up
and medium). The result was consistent with that of the ﬁrst study. They found that
the newscaster in high angle treatment was rated with signiﬁcantly higher credibility
(perceived character and sociability) and, in addition to that, higher task attraction
than in the low angle treatment (p<.05). However, perceived physical attraction was
not affected by the camera angle variations. Both studies’ ﬁndings indicated that
higher camera angle perspectives produce more positive evaluations of source

credibility than lower camera angle perspectives.

Table 1
Summary of the Camera Angle Tests

 

 

 

Camera Angle Confederate Dependent Result
Identiﬁcation Variable (p)
Medium High Low
Mandel] Low and high angle Chairman Potency sig.
& Shaw deﬁned as 12 degrees Activeness sig.
below or above the Credibility n.s.
eye-level plane Trustworthiness n.s.
Kappas, Low and high angle Confederates Facial Expression sig.
et a]. deﬁned as 40 degrees not identiﬁed Dominance n.s.
below or above the Submissiveness n.s.

eye-level plane

Tiemens Low and high angle Newscasters Credibility n.s.
deﬁned as 18 degrees
below or above the
eye-level plane

McCain, Low and high angle Presenters Credibility _s_i_g,
et al. (a) deﬁned as 23 degrees

below or above the

eye—level plane

McCain, Low and high angle Newscasters Credibility s_ig=
et al. (b) deﬁned as 38 degrees Attractiveness n.s.
below or above the
eye-level plane

 

Note: sig. represents the signiﬁcant ﬁndings supporting the camera angle theory; ns. means no
signiﬁcant support is found; gig, represents the signiﬁcant ﬁndings contrasting the camera angle
theory.

The summaries of various studies above show mixed results found on the
assumptions of camera angle effects with some (i.e., McCain, et al., 1977) even
opposite the prediction of the theory. However, the practice of varying camera angles
to achieve particular effects in photojournalism has never been distracted by these
unsatisfactory ﬁndings at all. In his effort to test whether camera angles effects
methods were commonly practiced by professional photographers, Kepplinger (1982)
surveyed 316 cameramen (one camera woman included) in Germany to ﬁnd their

Opinions on how camera angle creates a positive or negative display of politicians.

 

WEST!

a C
«n/v

\. J.

Not surprisingly, over two thirds of the cameramen (78%) said it was deﬁnitely
possible that a cameraman could make any person appear in a positive or a negative
way; less than one third (22%) said it might be possible; and only one said it was
hardly possible. Two thirds of all these cameramen shot a politician at eye-level, one
fourth from slightly above or slightly below eye-level, and none would choose a full
top or bottom view. Interestingly, all these cameramen conﬁrmed that a camera
position at eye level or slightly below eye level is preferred to achieve positive camera
effects on the photographed subjects (i.e., rest, sympathy, ease, power, liveliness, and
skill), whereas a full top or bottom view is mainly used to produce negative effects
(i.e., excitement, antipathy, tenseness, weakness, emptiness, and clumsiness).
Kepplinger’s ﬁnding seems to conﬁrm that the effects of camera angles, as stated in
the camera angle effects theory, are more indicative of the beliefs and habits of
cameramen and ﬁlm directors rather than effects with (strong empirical evidence.
However, providing the fact that the effects of camera angles on human
perceptions are very subtle and difficult to track and measure, and that other factors in
the design of those experiments (i.e., the ways the low and the high angles were
deﬁned, the types and the conceptualizations of the dependent variables used in the
studies, and particularly the ways the confederates were identiﬁed) may also play a
role, one cannot immediately dismiss the camera angle effects hypotheses at all. After
all, the camera angle effects hypotheses might at least work when people make
judgment about a person in leadership position as found in the Mandell and Shaw
(1973) study that when the confederate was identiﬁed as a chairman, variations of
camera angles signiﬁcantly affected the ratings of the confederate’s perceived
potency, which is traditionally a masculine-trait characteristic. Moreover, Kepplinger

(1982) mainly surveyed the cameramen who shot the pictures of politicians, which

were culturally considered as male occupations. Therefore, one assumption could be
made. That is, when used on politicians, variations of camera angle may affect the
perceived masculine traits of the photographed person. To understand if the above
assumption correlates with the general notion of photography in political journalism,
we will set out to see the practice of camera angles in photojournalism by examining
the existing authentic media properties in the realm of the political arena. The
expectation is to ﬁnd evidence of visual biases, particularly the camera angle biases,
in politicians’ photographs.
Section 2
Visual Bias in Politics:
The Practice of Camera Angle Effects in Political Journalism

Historically, the strongest predictors of the voting decision are party affiliation
(e.g., Campbell, Converse, Miller, & Stokes, 1960) and voters’ political philosophy
(Rosenstone & Hansen, 1993), suggesting that campaigns and the accompanying
media coverage has minimal effects on voters. However, this conventional academic
wisdom has recently faded and many scholars now come to believe that campaigns
and media coverage fundamentally shape voters’ decision in forming both perception
of the salient social issues and candidates’ traits (lyengar & Simon, 2000).

Empirical support for the inﬂuence of candidate trait perception on electoral
success is not uncommon. Bishin, Stevens, and Wilson (2006) examined the impact
voters’ evaluations of the candidates’ character had on their vote choice during the
2000 presidential election and found that character evaluations were statistically and
substantively signiﬁcant predictors of the vote. As the rise of candidate-centered
campaigns portends an increase in the role of media in portraying the candidate, the

study of voters’ perceptions of a candidate’s traits is becoming more important than

ever.

Kepplinger (1982) identiﬁed two ways that voters use media to form
perceptions concerning a candidate’s trait: through the verbal coverage of the political
party’s candidates and through visual coverage of the political party’s candidates. He
proposed that while the ﬁrst may be unbiased, journalists tend to deliver, either
deliberately or unconsciously, biased optical commentary through the application of
camera techniques, such as shot angles. In his content analysis of the election
campaign visual coverage in Germany in 1976 on two nationwide television channels,
Kepplinger found that the great majority of shots of presidential candidates were at
eye level, slightly below, or slightly above eye level. Despite the fact that he could not
ﬁnd any signiﬁcant trend in journalist statements’ bias toward Federal Chancellor
Schmidt, two signiﬁcant differences were found in visual analysis: while Federal
Chancellor Schmidt was shown in only 31 shots ﬁ'om the full bottom view or full top
view, the Challenger Kohl was presented in 55 shots from these views; and the
number of negative audience reaction shots (verbal and non-verbal) was signiﬁcantly
higher for Kohl (99) than for Schmidt (7). These kinds of visual bias found in
journalistic photographs are not uncommon in other studies (Greenwood, 2005;
Moriarty & Popovich, 1991; Tiemens, 1978; Grabe, 1996; and Lee, Ryan, Wanta, &
Chang, 2004).

Greenwood (2005) content analyzed photographs of presidents from the best
pictures of the year to ﬁnd out at which camera angle (medium, low, or high) and
which camera-to-subject distance (medium, close-up, or far-distance) presidents were
more likely to be photographed. He retrieved 194 photographs from an electronic
archive of the Pictures of the Year International competition (POY) at the University
of Missouri between 1943 and 2002. The results show that there were 67% eye-level

photographs, 23% low angle, and 10% high angle. The distribution of camera-to-

10

subject distance is not as dramatic as that of camera angle: 39% medium distance,
34% far-distance, and 27% close-up. These ﬁndings show that medium angle
photographs were highly-preferred when photographs of male presidential ﬁgures
were taken.

In a separate study on still photographs, Moriarty and Popovich (1991)
content-analyzed photographs and illustrations of the 1988 US presidential and vice
presidential candidates printed in three national weekly newsmagazines, US News and
World Report, Time, and Newsweek, between labor day kickoff and a week after the
general election to ﬁnd out if both candidates and parties were given equivalent
amounts of space and position or if one candidate received more emphasis. The study
also tried to elicit if subtle visual communication techniques being used were to
communicate negative or positive images. They found that, from a party standpoint,
Republican candidates accounted for signiﬁcantly higher number of visuals than
Democrats (chi-square=6.82, df=1, p<.01). Among the four candidates (Bush,
Dukakis, Quayle, and Bentsen), Bush’s photographs were signiﬁcantly larger than
either Dukakis’ or Quayle’s photos (F=4.413, sig.=.005). Bush was presented more
favorably in terms of facial expression (t=3.87, df=225, p<.01), low camera angle
(t=2.16, df=227, p<.05), and larger photograph size (t=3, df=204, p<.01). These
ﬁndings indicate that Bush’s visuals were treated more favorably than Dukakis’
visuals. Generally, Moriarty and Popovich found the biased displaying of presidential
candidates through photograph manipulations to be consciously or subconsciously
present in the work of the cameramen.

Not surprisingly, the above bias in photographing a presidential candidate is
not limited to still photography. Tiemens (1978) did the content analysis on the visual

content (528 shots) of three televised presidential debates between Gerald Ford and

11

Jimmy Carter to determine differences in how each candidate was portrayed by the
television medium. He found that the frequency and duration of appearance of the
candidates were roughly equal for each debate and equal camera movement of zoom
for each candidate. However, the number of tight shots (larger image size) for Carter
(25) was signiﬁcantly greater than that for Ford (8). Shots of Carter achieved the best
possible compositional balance 70% of the time whereas shots of Ford achieved the
best balance only 27% of the time. Camera angle shots in which either candidate
appeared alone were extremely difﬁcult to judge, and so were considered to be eye-
level. However, when both candidates appeared together, camera angle shots could be
determined. The results were that in the ﬁrst and the third debates, there were two
angle shots (low camera angle) that favored Ford, and in the second debate alone two
angle shots (low camera angle) favored Carter though Carter was shorter than Ford.
There were also more positive reaction shots (e.g., audience nodding and smiling) for
Carter than for Ford, and Carter showed more frequent eye contact.

Grabe (1996) conducted a similar study to Tiemens (1978), but in a different
social context ﬁ'om the American one. He did a content analysis on a 792-minute
televised material about the election campaign in the South African broadcasting
corporation’s coverage of the 1987 and 1989 elections to test whether all the political
parties in South Africa received fair coverage in terms of visual technicality. He found
out that the governing party received strikingly more than twice the air time coverage
of any other political parties and have beneﬁted through positively biased visual
portrayals of its candidates (i.e., low angle shot, positive cutaway shot, link of
candidates to positive person, close up shot, large audience view, positive
background, and long time-length of talk). Positive visual distortion was signiﬁcantly

associated with those images of the national party (p<.001) whereas negative visual

12

distortion (i.e., high angle or extreme low angle, negative cutaway shots, negative
environment, linkage of candidates to negative person, zoom out shot, empty audience
chairs and short time-length of talk) was signiﬁcantly associated with the other four
parties’ candidates. The ﬁndings projects the visual biases practiced in journalism, the
gap of which was so much more pronounced in South Aﬁ'ica than in the United
States.

In another similar study, Lee, Ryan, Wanta, and Chang (2004) conducted
content analysis to compare photographs of presidential candidates in the United
States (199 from The New York Times) and Taiwan (200 from the China Times)
within the last two month campaign period in 1996 to ﬁnd out if candidates in the US
and Taiwanese elections received a similar amount of photographic coverage and
whether the presidential candidates were portrayed differently in newspaper images in
the US and Taiwanese newspapers. He found that in the US, Dole’s and Clinton‘s
pictures were not signiﬁcantly different in terms of photograph coverage and
photograph size (measured in inches). In Taiwan, however, the ruling party candidate
Lee signiﬁcantly beneﬁted from larger photograph sizes than the other three parties’
candidates (F = 10.6, df=3, p<.01), but the difference between the number of
photographs was not signiﬁcant. Interestingly, the challenger in the US received more
coverage in terms of total space of newspaper photographs (chi-square=97.68, df=1,
p<.01) and in Taiwan the ruling party (chi-square=.89, df=3, p<.01). Taiwanese
photographs were more likely to be looking down at the candidates than the US
counterparts (11.1% versus 4.5%). Similar to Grabe’s ﬁndings, Lee and his colleagues
found that visual biases in journalistic practices were not restricted only to the United

States, but were even more pronounced in other countries such as South Africa and

Taiwan.

13

 

0

\z

The research summaries above portray a wide range of visual bias (e.g., visual
coverage, image size, camera angle, etc.) which is either consciously or
subconsciously practiced in media to convey either positive or negative traits of each
politician to viewers. Consistent with camera angle effects theory, varying camera
angles to achieve particular impacts on the viewers is not unusual in the real practice
of cameramen. Eye-level angle or slightly low angle is mostly used to attain positive
reactions whereas high angle is connected to negative portrayals. The question that
remains, however, is which trait element of those politicians is more susceptible to the
camera angle manipulations.

Kilbum (2005) had the answer to the above question. He had reanalyzed the
data from National Election Studies (N ES) surveys and found that the ﬁve trait
perceptions gauged in presidential elections from 1980 to 2000 encompass
intelligence, knowledge, leadership, moral, and sympathy. Among these ﬁve traits, he
wrote, leadership perception was found to be most strongly inﬂuenced by
predisposition and was more susceptible to media portrayal of the political candidates.
One assumption derived from the literature of camera angle effects and visual bias in
politics then is that in the realm of political media arena, camera angle effects might
work speciﬁcally to affect the leadership trait of the politicians, or speciﬁcally male
politicians. It was very obvious that all the content analyses discussed in this section
only offer the visual assessment of the male politicians rather than that of the female
ones. The experimental studies in the previous section also overwhelmingly used male
rather than female confederates to test the effects of the camera angle manipulations.
Nowadays, however, as more and more women enter political leadership positions, a

need to test whether or not camera angles retain its effects for female politicians is

14

pronounced. In the following section, various issues regarding female leadership in
the political arena will be underscored.

Section 3
Female Political Leadership

The history of the United States has never seen a woman becoming either a
president or a vice-president. It was historic that the 1984 election saw a female vice-
presidential candidate Geraldine Ferraro from one of the two major political parties
for the ﬁrst and only time in the history of this country to date. Some scholars (i.e.,
Rosenwasser, Rogers, Fling, Silvers-Pickens, and Butemeyer, 1987) even went on to
conclude that Ferraro’s candidacy would have a potential and lasting impact on future
women and political process in the country. They speculated that more and more
women would seek higher political offices in the future because (1) the ﬁrst barrier
(i.e., self-selection) to women’s involvement in politics was greatly challenged by the
event and (2) future children, both boys and girls, would be socialized in a way apart
from the traditional role expectations. These speculations were suggested aﬁer
Rosenwasser and his colleagues failed to ﬁnd signiﬁcant support for their sexism
hypotheses in students’ attitudes toward female versus male politicians.

Rosenwasser and his colleagues were not alone in failing to ﬁnd signiﬁcant
support for the existence of sexism in vote choice. Analyses of election outcomes by
candidate’s sex have always yielded somewhat inconsistent ﬁndings. When Kaming
and Walter (1976) examined women’s election to city councils, they found little
evidence for voter bias against women. When Welch, Ambrosius, Clark, and Darcy
(1985) examined state legislative contests from 1970 to 1980, they found that a
candidate’s sex had little impact on election outcome. These researchers then

concluded that the primary reasons for lower representation of women compared to

15

men in state legislatures were not the existence of sexism among voters, but the
scarcity of female candidates and the fewer female incumbents.

However, right now, at a time when more women become increasingly active
in politics, it is still uncommon for a woman to be elected in the higher political
ofﬁce. In their investigations of candidacies for state office in Nebraska, Iowa, and
Missouri, Ambrosius and Welch (1984) reported (I) that female candidates averaged
11% fewer votes than male candidates in elections in the late 19705 and (2) that
women’s chances of holding ofﬁce decrease with the prestige of the office. These
ﬁndings were consistent with a Huddy and Terkildsen (1993) study. Huddy and
Terkildsen conducted an experiment with 297 undergraduate students to test whether
biases toward female candidates varied across office levels. In their study, subjects
were randomly assigned to hear about a woman or a man with either typical
masculine traits (e.g. assertive, coarse, self-conﬁdent, etc.) or feminine halts (e.g.,
warm, gentle, sensitive, etc.) running for national or local ofﬁce. The subjects were
then asked if they were willing to vote for that hypothetical candidate. They found
that male characteristics in both male and female candidates are preferred in higher
levels of political office and that voters punish both male and female candidates who
lack typical masculine traits, especially when seeking higher national or executive
office. They also found that typical feminine traits in both male and female candidates
are considered more suitable for lower or non-executive levels of office. The question
that remains then is what happens when visual information (i.e., photograph) rather
than verbal information concerning traits is available to voters. How could judgments
of feminine or masculine leadership traits be made?

Huddy and Terkildsen (1993) speculated that when information about

candidate’s traits is not present, female candidates are penalized directly when

16

running for higher or executive office solely because voters are more inclined to
stereotype them as typical women with feminine rather than masculine traits. This
speculation makes sense when either verbal or visual information is not available.
However, as discussed in section 1 and section 2 in this chapter, visual manipulations
(i.e., camera angle manipulations) have profound inﬂuence on perceived masculine
leadership traits (i.e., perceived attributes) of politicians. If visual information (i.e.,
photographs) is available, one possible explanation is that viewers would rely solely
on the available photographs to judge the degree of masculine leadership traits of both
male and female candidates. If so, what then is the most eminent masculine trait in the
leadership ﬁeld to affect judgment of suitability for an executive office position?
Eagly, Makhijani and Klonsky (1992) conducted a meta-analysis on 221
research studies on the evaluation of women and men who occupy organizational
leadership roles. They chose only the experiments in which the characteristics of
leaders other than their sex were held constant to investigate whether people are
biased against female leaders. They found that leadership evaluations were less
favorable for female leaders than for males (d=.005) and that slightly more than half
(56%) of the comparisons between male and female leaders favored men. The result
also showed that the tendency to favor men over women was signiﬁcantly larger
when the dependent variable was the leader’s competence (p<.001) than when the
dependent variables are other leadership elements (e.g., styles). This meta-analysis
seems to convey that the leadership competence is the prevalent masculine leadership
trait which is preferred in masculine-type leadership positions. Eagly and Johannsen-
Schmidt (2001) further supported the above meta-analysis results. They concluded at
the end of their study that one area in which gender stereotypes manifest themselves is

the attribution of leadership competence.

l7

To sum up, when only visual information is available, viewers might judge the
suitability (i.e., the leadership competence) of a candidate to an executive ofﬁce
position based solely on what they get from the photographs. As a result, one
assumption can be made; that is, manipulations of a candidate’s photographs (i.e.,
camera angle manipulations) might be equal to manipulations of viewers’ perceptions
of the candidate’s leadership competence. Since leadership competence is a more
masculine than feminine trait, another assumption is that camera angle effects will
work proportionally the same for both male and female candidates, with higher angle
associated With weakness (i.e., culturally-based feminine traits) and lower angle
associated with power and strength (i.e., culturally-based masculine traits).

Because there have already been quite a few studies on the relationships
between photographs and male political candidates, the present study will focus solely
on female political candidates. The female mayor candidacy is chosen for the present
study because of (l) the impossibility of testing the political leadership of the female
presidential or vice-presidential candidates thanks to their next to non-existence in the
history of the United States, (2) the categorization of the mayor position in executive
or so-called masculine-type office, and (3) Adams’ ﬁndings (1975) that a hypothetical
female candidate receives intermediate support when running for local executive
positions (i.e. mayor). The intermediate support of a female candidate in this
masculine-type office is what makes the mayor position suitable for the present study
because (1) subjects’ sexism is reduced to a certain extent thanks to the present
commonality of female candidacy at that office level and (2) the masculine-type
ofﬁce will make the more positive leadership judgment sway toward masculine traits
(i.e., strength, power, and competence) in either a male or female candidate. This is

where the camera angle manipulations in the photographs ﬁt in and also where the

18

effects of varied camera angles on a female political candidate’s photograph retain
their proportional consistency with the effects on male candidate’s photograph. With
no other information available, subjects will depend totally on the photographs (the
same photographs whose only difference is the variation of the camera angles) to help
them judge the photographed person.

It is then proposed that variations in camera angle (low, medium, high) will
give rise to different ratings along the “political leadership competence” dimension of
a female hypothetical mayor candidate, with the low angle condition having the
greatest competence, the high angle the least, and the medium angle producing a
rating in between the ﬁrst two. Three hypotheses are then derived:

H1: Low angle photograph of the hypothetical mayor candidate will be rated

to have the greatest leadership competence.

H2: High angle photograph of the hypothetical mayor candidate will be rated

to have the least competence.

H3: Medium angle photograph of the hypothetical mayor candidate will be

rated to have higher competence than the high angle photograph and lesser

competence than the low angle photograph.

Finally, Eagly and her colleagues (1992) found that studies using male
subjects produced a larger preference for male leaders than studies using female
subjects (p<.05). Given the subtle presence of sexism in judgment of male and female
leadership found in this meta-analysis, it makes sense to check out if the present study
could track any signiﬁcant trend of sexism. Another hypothesis is then proposed:

H4: Regardless of camera angle perspective, female subjects will produce

higher ratings along the leadership competence dimension for the hypothetical

female mayor candidate than male subjects.

19

CHAPTER 3
METHODOLOGY

Overview

A total of 156 participants looked at a photograph of a female confederate as
part of a classroom experiment. The manipulated independent variable in the study
was camera angle used to differentiate the three photographs in the study. Each
photograph was printed in color on a one-fourth A4 paper. After seeing the
photograph, the participants completed measures of perceived leadership competence
and perceived attractiveness of the confederate, together with the trend of their
political news consumptions and their demographic characteristics.

Participants

Participants in this study were undergraduate communication students at
Michigan State University who received one course credit for their participation.
Nearly all the participants (97.4%) were between 18 and 24, 49 (31.4%) of the 156
participants were male, and 136 (87.2%) were Caucasian. Both male and female
participants were given a random photograph which was taken from either a medium,
high, or low angle. This resulted in three main cells with slightly above 50
participants each. Male participants were more concentrated in low camera angle

condition (n = 20) than in either medium angle (n == 15) or high angle (n = 14)

 

 

 

condition.
Table 2
Biological Sex *Camera Angle Crosstabulation
Camera Angle

Medium High Low Total
Male 15 I4 20 49
Female 39 37 31 107
Total 54 51 51 156

 

20

About four ﬁfths (80.8%) of the 156 participants consume political news at
least once a week, with television (M = 3.65, SD = 1.14) and the internet (M = 3.44,
SD = 1.11) rated as the two mediums they depend on most and radio (M = 2.45, SD =
1.18) and magazine (M = 2.47, SD = 1.15) as the mediums they depend on the least.
Newspapers (M = 2.97, SD = 1.03) was rated in between all the four mediums.

Table 3
The Trend of Participants ' Political News Consumption

 

Question: Please circle the number that corresponds with your level of agreement to a
number of statements below about yourself.
I use (mediuml for acquiring political news on a regular basis.

 

 

 

 

 

 

Medium
Newspaper Internet Television Magazine Radio
Strongly 1 14 10 9 41 37
Disagree
2 36 19 19 39 55
3 53 46 27 41 30
4 46 54 64 31 25
Strongly
Agree 5 7 27 37 4 9
Table 4
Comparison of Mean and Standard Deviation of the Five Mediums
Medium
Newspaper Internet Television Magazine Radio
M 2.97 3.44 3.65 2.47 2.45
SD 1.03 1.11 1.14 1.15 1.18

 

21

All participants, regardless of the camera angle conditions they were exposed
to, were told that the photograph they were looking at was that of a political candidate
running for a mayor office in Portland, Oregon. Portland was chosen for its
northeastern geographical setting. Unlike cities in the southern regions, participants of
the study might have heard about the city and perceived it as not too different ﬁ-om
places in their own state. Moreover, it was highly possible that participants were not
very familiar with Portand’s local politics since it is in a region with which they were
not likely to have experienced. Hence, they were less likely to have connotations
associated with the city, and the possible bias of their responses would be reduced to a
great extent.

Manipulations: Confederate and Camera Angle

The model for the study, a thin mid-thirty-year-old American graduate student
of MSU was the same in all the three experimental conditions (low, high, and
medium). In all the three conditions, she wore the same formal attire (e.g. dark
business attire), looked the same, was in the same standing posture, and was in the
same studio with a white background.

Each photograph was a simple studio portrait (i.e., head shots) taken from
three different camera angles—straight on, from above, and from below. Like other
studies, camera angle in this study was operationally deﬁned as the degree to which a
camera lens varied on a vertical plane of 180 degrees from a normal eye level camera
shot of the model. As a result, in the medium angle condition, the model was
photographed with the camera on the horizontal plane of her eyes. In the high-angle
condition, the model was photographed with the camera at an angle of approximately
12 degrees above the horizontal plane of her eyes. Finally, in the low-angle condition,

the model was photographed in a frontal, low shot, from an angle of approximately 12

22

degrees below the horizontal plane of her eyes. All of the photographs were taken in
only one photography session with the camera positioned four and a half feet from the
model using a 90mm lens camera. The camera was positioned on a vertical pole in the
studio and was moved up and down to get the high and low camera angle. Consistent
with Tiemens’ (1970) methodology, to get the photographs ﬁ'om the high and the low
angle, the camera was positioned 12 inches above and below the eye-level plane.
There were four shots ﬁ'om each camera position, producing a total of 12 photographs
to choose from at the end of the shooting process. One photograph from each camera
angle position was selected by the investigator and the cameraman. The experimental
variations of camera angle described above create an experiment with three
photographs or three conditions.

The decision concerning the variation of the camera angle 12 degrees above
and below the horizontal plane of the model’s eyes was made in accordance with the
Mandell and Shaw (1973) study which was similar to the present study on two
components—one is the identiﬁcation of the confederate as a politician, and the other
is the purpose of the seeing the inﬂuence of camera angle in the journalistic context.
Moreover, Mandell and Shaw also found that using twelve degree variations caused
subjects to not be aware of any unusual camera angle shots in the photographs. These
criteria would make the present study successful in concealing the camera angle
manipulations. As a result, the fear of viewers’ awareness of deliberate manipulation

in the photography would be eliminated.

23

Measures

The three-page questionnaires solicited the following variables:

Assessment of Perceived Leadership Competence: Each photograph was
assessed regarding the perceived leadership competence of the person in the
photograph. Weinert’s Attribution of Leadership Competence Scale (Sczensny &
Kuhnen, 2004) was used. This scale consisted of 10 items. The ten items are nine
leadership characteristics (assertiveness, dominance, ability to c0pe with pressure,
responsibility, ability to convince others, ability to make decisions, possession of
initiative, self-conﬁdence, and possession of authority) and one question regarding the
subjects’ assessment of the degree to which the stimulus person was perceived as
being suitable for the mayor position. These traits were rated on a seven-point ranging
scale ﬁ'om not at all (1) to very much (7), and were indicated in Sczensny and
Kuhen’s ﬁndings to be valid predictors of the leadership competence. Consistent with
the Sczensny and Kuhen (2004) study which found the reliability of the items to be a

= .90, the present study found the internal reliability of a = .95.

24

Table 5
Means, Standard Deviations, and Corrected Correlation with Total Scale Score
of Competence-Scale Items

 

 

Competence Scale Items M SD r
1. The candidate is assertive. 4.33 1.57 .82
2. The candidate is dominant. 4.15 1.46 .75
3. The candidate is able to cope

with pressure. 4.35 1.43 .78
4. The candidate is responsible. 5.04 1.57 .76
5. The candidate is convincing. 4.38 1.37 .79
6. The candidate is decisive. 4.59 1.48 .83
7. The candidate possesses initiatives. 4.76 1.46 .80
8. The candidate is self-conﬁdent. 4.54 1.71 .80
9. The candidate possesses authority. 4.38 1.56 .83

10. The candidate is suitable for a
mayor position. 4.38 1.40 .76

 

Additional Variables: Participants were also asked to rate their perception of
the model’s attractiveness on a ranging scale from not at all (1) to very much (7) and
their levels of agreement to ﬁve statements about the regularity of their use of ﬁve
different mediums (i.e., newspapers, intemet, television, magazine, and radio) for
acquiring political news on a ﬁve ranging scale from strongly disagree (1) to strongly
agree (5). Finally, they answered one question about the ﬁ'equency of their political
news consumption.

Subject Attribzﬂ: The questionnaire ended by requesting information on

subjects’ attributes, including sex, age, and ethnicity.

25

Procedure

One questionnaire cover page with code number, one printed colored
photograph with code number, and one three-page questionnaire were stapled
together. All the code numbers were a four-digit number, the ﬁrst digit of which
identiﬁed the camera angle condition of each photograph. The ﬁrst digit for the
medium angle photographs is 1, that for the high angle photographs is 2, and that for
the low angle photographs is 3. A hundred and eighty questionnaires were generated
and randomly ordered.

The experimental study was conducted in three different classrooms with the
permission of the course instructors. At the end of a class lecture, the experimenter
entered the classroom, introduced herself as an MA. student in the Communication
Department and her project as a study on how photographs help viewers form
judgments of a female leader when other types of information about her was not
available. She also explained to the subjects that each of them would be given a
consent form and a questionnaire. Those who agreed to participate would print their
names on the consent forms and proceed to the questionnaires. They would get one
class credit for their participation. Those who didn’t want to participate were free to
leave the room. She also told the subjects that after completing their questionnaire,
they would bring their consent forms and questionnaires to the front table and put
them in two separate piles — one for the consent forms and the other for the
questionnaire packs. This was done to conceal the subjects’ identities. The subjects
were also told that after submitting their consent forms and questionnaires, they each
would take a piece of the papers on the table, which provided the contact information
of the investigators whom they could write to if they would like to see the result of the

study. Those papers were the debrief forms that tell the subjects about the true nature

26

of the study and the contact information of the investigators. The debrief forms
explicitly tell the subjects that they had been randomly shown one of the three
different photographs of one American woman that differed on levels of camera
angle, and that their answers would be compared with those who viewed the other
photographs. This was done to give the investigators an idea of the judgments of the
political leadership competence evaluation of all the three photos.

After the two-minute explanations, students were allowed to ask questions or
raise any of their concerns. Finally, the experimenter handed out the consent forms
and the questionnaires to the students. Student subjects then read through the consent
forms, printed their names, completed the questionnaires, handed in both pieces of
information to different piles, grabbed the debrief forms and left the room. They spent

less than 15 minutes in the whole process.

27

CHAPTER 4
RESULTS

Perceived Leadership Competence

ANOVA and t-test were used to test the hypotheses about perceived
leadership competence advanced in this study. A one-way ANOVA was used to test
the ﬁrst three hypotheses, and subsequent tests of mean differences were conducted
using Tukey HSD analysis with a set at p < .05. The last hypothesis was tested using
an independent-samples t—test.

Camera Angle and Perceived Leadership Competence: The ﬁrst three
hypotheses of the study predicted that variations in camera angle (medium, high, low)
would produce different ratings of the perceived leadership competence. A one-way

analysis of variance performed on perceived leadership competence as a function of

camera angle was signiﬁcant, F (2, 155) = 6.278, p = .002, riz = .08. Subsequent

Tukey HSD analyses reveal supports for hypothesis 1, which predicted that the
photograph taken from a low camera angle perspective would be rated to have the
highest leadership competence. The perceived competence of the low angle
photograph (M = 4.98, SD = 1.12) was signiﬁcantly greater than that of the high angle
photograph (M = 4.30, SD = 1.31) and that of the medium angle one (M = 4.20, SD =
1.20). Hypothesis 2, which predicted that the photograph taken from a high angle
perspective would be rated to have the lowest competence, was not supported.
Although the high angle photograph was rated to be less competent than the low angle
photograph, the ratings were indifferent when compared with the medium angle
photograph. Likewise, hypothesis 3, which predicted that the photograph taken from a
medium angle perspective would be rated to have the competence ratings between the

low angle and the high angle one, was not supported. The mean difference between

28

the competences in the medium and high angle conditions was not signiﬁcant. The
ﬁndings, then, were consistent with hypothesis 1, but not with hypothesis 2 and
hypothesis 3. Moreover, the direction of the result was also not consistent with the
prediction in H2 and H3. The high angle photograph was rated to have higher

competence than the medium angle one.

 

 

 

 

 

 

Table 6
Camera Angle Effects on Perceived Leadership Competence
Statistics
M SD Min Max
Medium Angle 4.20 1.20 1.70 5.90
High Angle 4.30 1.31 1.60 6.80
Low Angle 4.98 1.12 2.20 6.80
Figure 1
Camera Angle Eﬂects on Perceived Leadership Competence
5.00 ‘
4.90 J
g 4.80 ..
8
g 4.70 —
g 4 60
O . .
.9:
«E 4.50 r
E
0
3 4.40 t
4.30 ‘
4.20 "

 

 

 

High Medium Low

Camera Angle

29

Biological Sex and Perceived Leadership Competence: Hypothesis 4 predicted
that female subjects would produce higher ratings of the perceived leadership
competence than would male subjects. An independent-samples t-test was used to test

this hypothesis. The result did not yield a signiﬁcant difference between the two
sexes, t (154) = 1.85, 122 = .02, p = .07. The perceived competence rated by the female

subjects (M = 4.62, SD = 1.19) was not higher than that by the male subjects (M =

4.22, SD = 1.36). Therefore, the data were inconsistent with the fourth hypothesis.

 

 

The Eﬁects of the Raters ’ BiologicalrSEcheoZr Perceived Leadership Competence
Statistics
M SD Min Max
Male 4.22 1.36 1.60 6.80
Female 4.62 1.19 2.10 6.80

 

 

 

 

 

Figure 2
The Eﬂects of the Raters ’ Biological Sex on Perceived Leadership Competence

7 I- u——-
6

8 i

5

*6 5 -

a.

E

o

O 4 ‘

a.

E

E 3 -

0

'6

8

u-l 2 ..
I _

 

 

 

r I

Male Female

Biological Sex

 

,l

Biological Sex *Camera Angle 6114.1 Perceived Leadership Competence:

Additionally, a two-way analysis of variance was conducted on the perceived

competence as a function of raters’ sex and camera angle. There is no interaction

effect found between camera angle and raters’ sex, F (2, 155) = .92, p = .40, riz =

.012. Both male and female subjects did not differ in their ratings of the perceived

leadership competence of the female mayor candidate across all the three camera

angle conditions.

 

 

 

 

 

The Eﬁ'ects of the Raters ’ Biological Sl‘eib‘leCimera Angle on Perceived Leadership
Competence
Male Female
Camera Angle M SD M SD
Medium 3.83 1.24 4.34 1.70
High 3.68 1.33 4.54 1.24
Low 4.88 1.23 5.04 1.05
Figure 3

Biological Sex *Camera Angle Eﬂects on Perceived Leadership Competence

 

5.10 -
0
§ 4.80 r
ﬂ
2'
g 4.50 ‘
u
.E'
.E ..
c 4.20
0
'5
3 3.90—
3.60'l

 

 

 

l l
Medium High Low

Camera Angle

31

Biological Sex

Male """"

 

Female

Perceived Attractiveness
Perceived Attractiveneﬂnd Perceived Leadership Competence: To further
understand the effects of camera angle on perceived competence, a correlation check
between the perceived competence and attractiveness was examined. The results
showed that the correlation was signiﬁcant, r (154) = .31, p < .001. Perceived
attractiveness was positively related to perceived competence. The photograph which
was seen to be more attractive would be rated to have more competence. Does camera
angle then affect the attractiveness rating?
Figure 4

Correlation Check of Perceived Attractiveness and
Perceived Leadership Competence

 

O

6 g B
8

manomo

      

one 00 an mo OCD

 

Leadership Competence

4‘ 8
3
3‘° 3 3 2
o o o
48 8 o o o
2 o

O

R Sq Linear = 0.096

 

'1 2' '3 4 '5 is ?

Perceived Attractiveness

Camera Angle and Perceived Attractiveness: A one-way analysis of variance
was used to examine the difference in reported attractiveness based on the three

camera angles. The result yielded insigniﬁcant effects of the camera angle on
perceived attractiveness, F (2, 155) = 0.26, p = .77, 172 = .003. The low angle

photograph (M = 4.08, SD = 1.31) was not more attractive than either the medium
angle photograph (M = 4.02, SD = ‘1.22) or the high angle one (M = 3.90, SD = 1.21).

No difference in attractiveness rating was observed among all the three photographs.

32

,1

Camera angle effects at 12 degree variations affected the perceived competence, but

not the perceived attractiveness.

 

 

 

Table 9
Camera Angle Eﬂects on Perceived Attractiveness
Statistics
M SD Min Max
Medium Angle 4.02 1.22 1 6
High Angle 3.90 1.21 l 6
Low Angle 4.08 1.31 l 7

 

Figure 5

Camera Angle Eﬂects on Perceived Attractiveness

 

 

 

4.1'
§ 4.05‘
3
'5
U
a
E 4 4
<1
'8
.2.
8
b 3.95‘
A:

3.9 ‘

 

 

Medium

j

High Low

Camera Angle

 

33

I

 

CHAPTER 5
DISCUSSION

This study began with the expectation that variations in camera angle would
give rise to different ratings of perceived leadership competence of a female mayor
candidate, with the low angle condition having the greatest competence, the high
angle the least, and the medium angle producing a rating in between the two.

Findings were generally consistent with the assumptions that photographs
taken from a low camera angle perspective were more favorably-judged in terms of
the perceived competence than either from a medium or a high angle perspective.
These ﬁndings provide partial empirical support to the German cameramen’s claims
in the Kapplinger’s (1982) study that low camera angle can be used to communicate
the positive traits (i.e., rest, sympathy, ease, power, liveliness, and skill) of a
politician, but also limited these claims to only the perceived masculine traits (i.e.,
power, competence, dominance, and decisiveness) . It could then be concluded that
the low camera angle treatments might be capable of projecting the masculine
leadership trait of the photographed person. Hence, like their male counterparts,
female political candidates who run for an executive office might beneﬁt from a low
angle shot, which underlines their masculine-leadership traits to the viewers.

Interestingly, however, assumptions of high and medium camera angles were
not ﬁrlly supported. Inconsistent with Mandell and Shaw’s (1973) ﬁndings on
judgments of the chairman’s potency, which found that medium angle photographs
were judged to have higher potency than the high angle one, the present study found
that judgments of photographs of a female political candidate did not differ when
taken from either a medium or a high angle perspective. Since a high angle
perspective makes the photographed person appear shorter than the viewers, one

explanation of this phenomenon might be that both male and female viewers might

34

tend to punish a male political candidate, but not a female political candidate, for
being slightly shorter than them (i.e., in this case they seemed to look downward
toward the model whose photograph was taken from a 12-degree-higher-than-eye-
level-horizontal-plane perspective). This explanation makes sense speciﬁcally in
American context where height, for men, has historically been associated with social
power. As Jackson and Ervin (1992) found, discriminations in judgments of social
attractiveness, professional status, and physical attractiveness as a ﬁmction of height
were much more pronounced when the evaluated models were men than when they
were women. This argument brings about another possibility. Culturally, American
people punish men for being short and women for being fat. A double chin is
perceived as a disadvantage, especially for women. Technically, however,
photographs taken from a low angle will project the double chin of a person, while
those from a high angle will hide the so-called drawback. In this case, will body size
and camera angle interact to affect the judgments of female leadership competence?
Future research is needed to verify these speculations in regard to camera angle
manipulations.

One main limitation of the present ﬁndings, however, is embedded in the
operationalization of camera angle. Only a single camera angle perspective (i.e., 12
degree variations) was utilized for each condition of the present study, and subjects
were exposed to only one of the camera angle perspectives, not to mention they all
saw the same model. In real life, this phenomenon hardly exits, especially in a media-
saturated country like the United States. As found in the study, the most popular
medium from which the subjects got most of their political news were not newspapers
or magazines where still photographs were common, but television. In television, a

single shot is just a single element of a total program that includes compliments of

35

Al

differing shot perspectives of a person. Since the whole might be different from the
simple sum of its parts, a television presentation might be more than individual shots
working independently of one another. As McCain, et a1. (1977) wrote, medium, high,
and low camera angle shots might have different effects depending on how they are
employed in the context of a sequence of shots. Future research is then needed to
explain all these unanswered phenomena in the context of other mediums, speciﬁcally
the television medium.

The strength that the present study possesses is the careful selection of its
dependent variable. The investigator of the present study believed that, apart from the
differences in the degree variations of low and high camera angles, the inconsistencies
and the contradictions of previous experimental studies about the camera angle effects
might be understood in terms of the different dependent measures employed and the
way the models were identiﬁed in those studies. Thejudgments viewers make
concerning a person’s potency or competence might not appear to operate in the same
fashion as do judgments about that person’s credibility or trustworthiness. Like the
potency dimension used in the Mandell and Shaw (1973) study, the competence
dimension of the present study is most analogous to the power and dominance
judgments (i.e., traditionally-based masculine traits) alluded to by ﬁlm theorists. A
camera shooting upward toward a photographed person may increase his perceived
power, but not necessarily his likeability. One caution is not to be missed, though.
Even though low angle shots might be associated with perceived power and perceived
dominance, the judgments viewers make about a newscaster’s perceived dominance
might not be the same as that about a politician’s. This might partly explain why the
present study found low camera angle to affect the perceived dominance of a political

candidate while the Kappas, et al. (1994) study could not ﬁnd any support for the

36

,/

 

same hypothesis when the model in their study was not identiﬁed as anything besides
an actor. It would be useful for future research to take these elements into
considerations while designing their studies.

Like many other studies (Rosenwasser, Rogers, Fling, Silvers-Pickens, &
Butemeyer, 1987; and Kaming, & Walter, 1976) that tried to dig on the subtle
presence of sexism in judging political candidates, the present study also failed. Male
and female subjects’ judgments of the female model’s leadership competence in the
study did not differ at all. However, since the present study, as well as other studies
(Rosenwasser, Rogers, Fling, Silvers-Pickens, & Butemeyer, 1987; and Kaming, &
Walter, 1976), used college students as experimental subjects, cautions should not be
overlooked because this group of the population belongs to more recent cohorts and
might be better educated than the general population. This might be a big underlying
cause of them having more egalitarian and more similar sex-role attitudes. This study
is clearly limited by the above reason. Thus, further study is called for to test the
relationship between camera angle and other visual variables in other settings, and
with more diverse subject populations.

Not only was biological sex not found to affect the general judgment of
perceived leadership competence, but it was also not found to impact the judgments
across different levels of camera angle. This ﬁnding is consistent to that of the
Kappas, et al.’s (1994) study, in which only 16 subjects were used and the variations
of camera angles were set at 40 degree treatments. Coupling the two studies together,
it makes some sense to say that, regardless of the degree variation of the camera
angle, male and female subjects’ visual perceptions might be similarly distorted in the
same direction. It is possible that both American men and women might have been

socialized to have the same response to the camera angle treatments. After all,

37

viewers’ biological sex might not be an important moderator of the camera angle
effects.

The last interesting ﬁnding of the present study was that even though
attractiveness and perceived competence were positively correlated and the camera
angle manipulations worked to distort the perceived competence, they did not have
any signiﬁcant effects on judgments of the perceived physical attraction at all. This
ﬁnding is consistent with the McCain et al.’s (1977) ﬁnding when variations of
camera angle were set at 38 degrees. However, cautions are particularly needed here.
As Kappas, et a1. (1994) reasoned that facial shapes were distorted when different
camera angles were used, their ﬁndings also indicated that facial expressions (i.e.,
happy, sad, disgusted) werejudged to be more positive in the low angle treatment
than in the other two treatments when the variations of camera angle were set at 40
degrees. When judgments of facial expression were found to be affected by camera
angle treatments, it was somewhat unsatisfactory to ﬁnd that perceived attractiveness
of the person does not change. One explanation might lend itself to the fact that the
model used in the present study was asked to display a neutral facial expression rather
than the three expressions displayed in the Kappas, et a1. study. Thus, future research
should attempt to resolve these paradoxical ﬁndings by coupling the two dependent
variables (i.e., judgments of facial attractiveness and that of facial expressions) in the

study and vary the facial expressions of the models.

38

 

,/

 

CHAPTER 6
CONCLUSION

This study suggests that camera angle manipulations may be one among other
powerﬁrl determinants of perceived leadership competence of the politicians in the
executive offices. It speculates a useful possibility of testing camera angle effects in
two different leadership trait categories: feminine-trait and masculine-trait. Future
work should attempt to determine (1) how camera angle treatments work injudgments
of these two poles of leadership traits, in regard to the politician’s sex; and (2) how
camera angle treatments work at different levels of political ofﬁces. Moreover,
additional research is also needed to verify the present study’s ﬁndings before we can
fully understand all the relationships involved. The present study suggests that the
following factors should be considered in future investigations: (1) the variations of
the degrees of the camera angle and (2) the variations of the body sizes of the female
political candidate. Control and investigation of these variables will contribute
substantial knowledge to our understanding of the effect of camera angle on female
political leadership. Finally, the present study was limited to investigating some
effects of camera angle on female leadership with respect to a single candidate.
Additional research is then needed in situations where the audience view photographs

of two or more candidates at the same time.

39

 

\/

 

APPENDICES

40

INFORMED CONSENT FORM
Title of project: The Camera Angle Effects on Political Leadership Competence

Project contact: Chou Chea, M.A. (cheachou@msu.edu; 517-355-2816)
Michigan State University

John Sherry, Associate Professor (ishem@msu.edu, 517-355-6648)
551 ComArts, East Lansing, M148824

This study investigates how people form impression of political leadership
competence by just looking at the photo available to them. Participants will be asked
to look at a photo and answer a few questions about their opinions toward the person
in the photo. Participants will also be asked some basic demographic information. It
will take about 15 minutes to look at the photo and answer the questions.

Participation in this research is conﬁdential, and the privacy of participants will be
protected to the maximum extent allowable by law. In the event of publication of this
research, no personally identifying information will be disclosed and all data analyses
will be conducted at the aggregate level.

Participation is volunta_ry. Individuals are free to stop participating in the research at
any time, or to decline to answer any speciﬁc questions without penalty.

Participants will receive one research/class credit for participating in this study. If you
don’t wish to participate in the research study, you can see your course instructor for
non-research alternatives.

No risks are foreseen.

Participants may ask any questions about the research procedures, and these questions
will be answered. Simply contact Chou Chea at 517-355-2816 or
cheachou@msu.edu, or John Sherry at 517-355-6648 or jsherry@msu.edu.

The participants is 18 years of age or older.

If you agree to participate in this study, please indicate this by signing your name in
the space provided.

1 voluntarily agree to participate in the study about political leadership competence.

 

 

Participant Signature Date

 

 

Researcher Signature Date

 

If you have any questions or ooncems regarding your rights as a study
participant, or are dissatisﬁed at any time with any aspect of this study, you may
contact - anonymously, if you wish, Peter Vasilenko, Ph.D., Director of the
Human Research Protection Programs (HRPP) at Michigan State University:
(517) 355-2180, fax: (517) 432-4503, email: irb@msu.edu, or regular mail: 202
Olds Hall, East Lansing, MI 48824

 

 

 

41

MEASUREMENT
Please look carefully at the photograph that is attached to your questionnaire.

The photograph in your questionnaire is that of one political candidate running
for a mayor office in Portland, Oregon.

Please rate the candidate’s competence quality on a seven-point scale ranging
from not at all (1) to very much (7) to indicate your level of agreement with each
statement or question listed below. Please circle the number that corresponds

with your level of agreement.

1. The candidate is assertive.

Not at all 1 2 3 4 5 6 7 Very much
2. The candidate is dominant.

Not at all 1 2 3 4 5 6 7 Very much
3. The candidate is able to cope with pressure.

Not at all 1 2 3 4 5 6 7 Very much
4. The candidate is responsible.

Not at all 1 2 3 4 5 6 7 Very much
5. The candidate is able to convince others.

Not at all 1 2 3 4 5 6 7 Very much
6. The candidate is decisive.

Not at all 1 2 3 4 5 6 7 Very much
7. The candidate possesses initiative.

Not at all 1 2 3 4 5 6 7 Very much
8. The candidate is self-conﬁdent.

Not at all 1 2 3 4 5 6 7 Very much
9. The candidate possesses authority.

Not at all 1 2 3 4 5 6 7 Very much

10. To what degree do you consider the candidate to be suitable for a mayor
position?

Not at all 1 2 3 4 5 6 7 Very much
11. How physically attractive do you consider the candidate to be?

Very unattractive 1 2 3 4 5 6 7 Very
attractive

42

‘L/

 

 

Please answer the information below about yourself:

Please circle the number that corresponds with your level of agreement to
a number of statements below about yourself.

I use newspapers for acquiring political news on a regular basis.
Strongly disagree 1 2 3 4 5 Strongly agree
I use the intemet for acquiring political news on a regular basis.
Strongly disagree 1 2 3 4 5 Strongly agree
I use television for acquiring political news on a regular basis.
Strongly disagree 1 2 3 4 5 Strongly agree
I use magazine for acquiring political news on a regular basis.
Strongly disagree 1 2 3 4 5 Strongly agree
I use radio for acquiring political news on a regular basis.

Strongly disagree 1 2 3 4 5 Strongly agree

Check (‘1) the information below that applies to you.

 

How much do you consume political news on average?
At least once daily
At least several times a week
At least once weekly
At least once monthly
Less often than once monthly
What is your biological sex?
Male
Female
What is your age?
18 to 24 years old
25 to 35 years old
36 to 50 years old
Over 50 years old

43

 

 

v

 

 

What is your ethnicity?
African American
Asian American
Caucasian/European American
Hispanic

Other (please specify)

This is the end of the questionnaire.

44

,/

 

DEBRIEF FORM

Title of project: The Camera Angle Effects on Female Political Leadership
Competence

Project contact: Chou Chea, M.A. (cheachou@msu.edu; 517-355-2816)
Michigan State University

John Sherry, Associate Professor (isherry@msu.edu, 517-355-
6648) 551 ComArts, East Lansing, M148824

Thanks again for your participation in the study.

You werejust randomly shown one of the three different photos of one female
American confederate that differed on three levels of camera angle (low, high,
and medium).

Your answers will be compared with those people who viewed the other
photos in order to give us an idea of the judgment of female political
leadership competence evaluation of all the three photos.

If you have any questions about this study - or would like a copy of the study
results - please contact Chou Chea at cheachou@msu.edu.

45

 

:r I

k/

 

 

 

 

 

 

PHOTOGRAPHS

The low-angle photograph

 

The medium-angle photograph

 

The high-angle photograph

 

46

,/

 

BIBLIOGRAPHY

47

 

BIBLIOGRAPHY

Adams, W. C. (1975). Candidate characteristics, ofﬁce of election, and voter
responses. Experimental Study of Politics: 4, 76-91.

Ambrosius, M. M., & Welch, S. (1984). Women and politics at the grassroots:
Women candidates for state ofﬁce in three states, 1950-1978. Social Science
Journal: 21, 29-42.

Bishin, B. G.; Stevens, D.; & Wilson, C. (2006). Character counts? Honesty and
fairness in presidential elections. Public Opinion Quarterly: 70, 235-248.

Campbell, A.; Converse, P.; Miller, W.; & Stokes, D. (1960). The American voters.
Chicago: The University of Chicago Press.

Eagly, A. H., & Johannsen-Schmidt, M. C. (2001). The leadership styles of women
and men. Journal of Socﬁ Issues: 57, 781-797.

Eagly, A. H., Makhijani, M. G., & Klonsky, B. G. (1992). Gender and the evaluation
of leaders: A meta-analysis. Psychological Bulletin: 111, 3-22.

Edmonds, R. (1982). The sights and sounds of cinema and television: How the

aesthetic experience inﬂuences our feeling. New York: Teachers College
Press.

Entman, R. (1993). Framing: Toward clariﬁcation of a ﬁ'actured paradigm. Journal of
Communication: 43, 51-58.

Grabe, ME. (1996). The South Aﬁican broadcasting corporation’s coverage of the
1987-1989 elections; the matter of visual bias. Journal of Broadcasting and
Electronic Media: 40, 153-179.

Greenwood, K. (2005). Picturingpresidents: A content analysis of photographs of
presidents from pictures of the yea_r. Paper presented to the Political
Communication Division of the lntemational Communication Association,

New York, NY.

Huddy, L., & Terkildsen, N. (1993). The consequences of gender stereotypes for
women candidates at different levels and types of ofﬁce. Political Resealch
Quarterly: 46, 503-525.

Huss, R. G., & Silverstein, N. (1968). The ﬁlm experience. New York: Delta, 1968.

Iyengar, S., & Simon, A. F. (2000). New perspectives and evidence on political

communication and campaign effects. AnnuaiReview of Psychology: 51, 149-
169.

48

 

VI

 

Jackson, L. A., & Ervin, K. S. (1992). Height stereotypes of women and men: The
liabilities of shortness for both sexes. The Journal of Social Psychology: 132,
433-445.

Kappas, A., Hess, U., Barr, C. L., & Kleck, R. E. (1994). Angle of regard: The effect
of vertical viewing angle on the perception of facial expressions. Journal of
Nonverbal Behavior: 18, 263-280.

Kaming, A. K., and Walter, B. O. (1976). Election of women to city councils. Social
Science Oﬁrterly: 56, 605-613.

Kepplinger, H. M. (1982). Visual biases in television campaign coverage.
Communication Research: 9, 432-446.

Kilbum, h. W. (2005). Does the candidate really matter? American Politics Research:
33, 335-356.

Lee, T.; Ryan, W.E.; Wanta, W.; & Chang, K. (2004). Looking presidential: A
comparison of newspaper photographs of candidates in the United States and
Taiwan. Asian Journal of Communication: 14, 121-139.

Mandel, L. M., & Shaw, D.L. (1973). Judging people in the news-unconsciously:
Effect of camera angle and bodily activity. Journal of Broadcasting: 17, 353-
362.

McCain, T. A., Chilberg, J ., & Wakshlag, J. (1977). The effect of camera angle on
source credibility and attraction. Journal of Broadcasting: 21, 35-46.

Moriarty, S. E., & Popovich, M. N. (1991). Newsmagazine visuals and the 1988
presidential election. Journalism Quarterly: 68, 371-3 80.

Peterson, D. A. (2005). Heterogeneity and certainty in candidate evaluation. Political
Behavior: 27, 1-24.

 

Rosenstone, S., & Hansen, J. M. (1993). Mobilization. participation, and democraay
in America New York: Macmillian Publishing Company.

Rosenwasser, S. M., Rogers, R. R., Fling, S., Silvers-Pickens, K., Butemeyer, J.
(1987). Attitudes toward women and men in politics: Perceived male and
female candidate competencies and participants personality characteristics.
Political Psychology: 8, 191-200.

Schreiber, E. M. (1978). Education and change in American opinions on a woman for

President. Public Opinion Quarterly: 42, 171-182.

Scott, L. M. (1994). Images in advertising: The need for a theory of visual rhetoric.
Journal of Consumer ReseaﬂLt: 21, 252-273.

49

AI

 

Sczensny, S., & Kuhnen, U. (2004). Meta-cognition about biological sex and gender-
stereotypic physical appearance: Consequences for the assessment of
leadership competence. Socieﬂ for Personam and Social Psychology: 30, 13-
21.

Sczesny, S.,& Kuhnen, U. (2004). Meta-cognition about biological sex and gender-
stereotypic physical appearance: Consequences for the assessment of
leadership competence. Personality and Social Psycholgy Bulletin: 30, 13-21.

Tieman, RT (1970). Some relationships of camera angle to communicator
credibility. Joumaa of Broadaasting: 14, 483-490.

Tieman, RT (1978). Television’s portrayal of the 1976 presidential debates: An
analysis of visual content. Communication Monogzaphs: 45, 362-370.

Welch, S., Ambrosius, M. M., Clark, J., and Darcy, R. (1985). The effect of candidate

gender on electoral outcomes in state legislative races. Western Politics
Quarterly: 38, 464-475.

50