This document analyzes the verbal and visual presentations on openly LGBT candidates' campaign websites compared to their opponents' websites. It was found that LGBT candidates discussed LGBT issues and progressive political ideology more frequently than their opponents. There were no significant differences found between the candidates in discussing religion, family, economic issues, or their visual presentations in images. The results provide empirical evidence for how LGBT candidates represent themselves and their constituencies through their online campaigns.

1. Mainstream Gay
Politicians Online
Verbal and Visual Presentations on LGBT Candidate Websites

2. Gay Politicians Winning
 More than 300% increase in openly LGBT U.S. politicians
holding public offices over past decade
 107 LGBT politicians elected to local, state, and national
offices in the 2010-midterm
 44 of the 54 LGBT candidates won races for state or the
U.S. Congress
 Emergence of LGBT politicians identifying as members of
a specific minority group and presenting themselves as
representative of a broader ideological constituency
(George, 2002; Victory Fund, 2012).

3. LGBT Candidates
 Although LGBT politicians have enjoyed increasing
successes recently, 30% of U.S. population would not
vote for a gay candidate (Gallup, 2012)
 LGBT candidates are conscious of this bias when
crafting their campaign communications
 This investigation attempts to describe the verbal and
visual content on openly LGBT candidates’ campaign
websites and compare it to their opponents’ websites to
account for endogenous effects

4. Theoretical Underpinnings
 Kaid and Davidson (1986) and Banwart (2002) used
social identity theory as the framework in their
development of VideoStyle and Webstyle, respectively.
 While prior studies explored gender differences
between candidates’ VideoStyles and WebStyles
(Bystrom, Banwart, Kaid, & Robertson, 2004), no
previous research on the presentations of openly LGBT
candidates on their campaign websites could be
located

5. Videostyle
 Kaid and Davidson (1986) used Goffman’s theory of self-
presentation as the theoretical basis for VideoStyle, the first
systematic protocol for analyzing candidate presentations in
televised political advertising.
 The goal of the candidates’ VideoStyle is similar to that of
the interpersonal communicator: impression management
and control of others’ responses.
 Based on the verbal, nonverbal, and production techniques
used in the television advertisement, candidates’
VideoStyles are purposefully constructed to achieve specific
cognitive, affective, and/or behavioral outcomes (Nesbitt,
1988).

6. Webstyle
 Banwart (2002) adapted VideoStyle constructs to develop
WebStyle in her analysis of gender differences in online campaign
communications.
 Websites offer “an unmediated, holistic, and representative portrait
of campaigns” (Druckman, Kifer & Parkin, 2009, p. 343).
 Unlike television advertising or news reports, websites give
viewers a unique opportunity to access a campaign directly,
enhancing unfiltered communication between politicians and the
electorate (Smith, & Smith, 2009).
 By 2010, nearly all major party candidates for the U.S. House or
Senate have campaign websites, and virtual presence has
become “a standard part of candidates’ tool kits” (Druckman, Kifer
& Parkin, 2010, p. 88).

7. Presentation & Representation
 The extent to which elected officials represent the
groups with which they identify, as well as their
broadest constituencies, has been analyzed using
Pitkin’s (1967) seminal explication of the theories of
representation.
 Descriptive representation refers to the extent to which
a politician “looks like, has common interests with, or
shares specific experiences” with those being
represented (Dovi, 2011).
 For LGBT politicians, descriptive representation may be
analyzed by determining whether LGBT candidates
present themselves as advocates of LGBT interests.

8. Presentation & Representation
 Pitkin (1967) also developed the theory of substantive
representation, which focuses analyses on the extent to
which the representative advances the policy
preferences of those he or she represents, but does not
necessarily share identity.
 LGBT politicians, along with other constituency and
candidate alternative variables, is a good predictor of
pro-LGBT policies at the local and state level (Heider-
Markel, 2010).
 No prior research comparing the presentations and
representative nature of LGBT politicians’ websites to
their opponents, however, could be found in the
literature.

9. Hypotheses and RQs
 H1: LGBT candidates will discuss LGBT issues more than their
opponents.
 H2: LGBT candidates will emphasize progressive political
ideology more than their opponents.
 H3: LGBT candidates will discuss religion less than their
opponents.

10. Hypotheses and RQs
 H4: LGBT candidates will discuss family less than their
opponents.
 RQ1: What are the differences between LGBT candidates and
their opponents’ emphasis of economic and social issues?
 RQ2: What are the differences in dress, facial expressions, eye
contact, and context between LGBT candidates and their
opponents in the images presented on their campaign
websites?

11. Method
 This investigation used SEO insights and trends analyses
to generate keyword search protocols for terms related to
the dependent variables:
 LGBT issues
 Progressive Political Ideology
 Religious Terms
 Family
 ANOVA revealed the number of pages on the websites of
LGBT and their opponents was not significantly different,
F(1, 91) = 0.02, p > .05.

12. Reliability
 Site search tool returns uniform results, 100%
agreement on coding of verbal content
 No significant disagreements on any specific
categories in the analysis of website images and
intercoder agreement on the visual content was 93.5%
 Overall, intercoder reliability across all verbal and
visual categories was determined using software that
calculated a Krippendorff’s Alpha of 0.96.

13. Results
 1,786 pages of online content analyzed to describe,
compare, and provide an empirical basis for theoretical
arguments about the verbal and visual presentation of
LGBT candidates’ websites
 54 LGBT Candidate Websites
 38 LGBT Opponents’ Websites
 No significant differences in the distribution of LGBT
candidates and their opponents across office level,
region, district partisanship, and candidate gender

14. Results
 H1: ANOVA revealed LGBT candidates use terms
associated with LGBT issues (M = 24.0, SD = 12.28)
more frequently than do their opponents (M = 4.64, SD
= 8.75), and this difference was significant, F(1, 90) =
12.48, p < 0.01. Supported
 H2: The results of an analysis of variance revealed that
LGBT candidates use terms associated with
progressive politics (M = 1.75, SD = 3.42) more
frequently than their opponents (M = -3.02, SD = 4.44),
and this difference was significant, F(1,90) = 8.27, p <
.01. Supported

15. Results
 H3: ANOVA revealed LGBT candidates use religious terms
(M = 6.98, SD = 5.41) marginally more frequently than their
opponents (M = 5.93, SD = 4.02), but this difference was
not significant, F (1, 90) = 0.70, p > .05. Not Supported
 H4: ANOVA revealed LGBT candidates use terms
associated with family members (M = 15.26, SD = 12.53)
marginally less than their opponents (M = 20.93, SD =
14.39), but this difference was not significant, F(1, 90) =
0.35, p > .05. Not Supported

16. Results
 RQ 1: Differences between LGBT candidates and their
opponents in discussions of deficit, immigration,
medical care,fiscal/economic, energy/transportation,
housing, military/defense, taxes, safety/crime,
jobs/employment, and education issues were not
significant, p > .05
 RQ 2: LGBT candidates were less likely to appear with
family members than their opponents.
 LGBT candidates and their opponents were equally likely
to be presented in casual dress, smiling, alone, and
making eye contact or looking directly into the camera.

17. Discussion
 Intuitive results provide first empirical basis for theoretical
arguments about the presentations of LGBT candidates on
their campaign websites
 Close relationship between social identity politics and
representation in regard to LGBT candidates
 “Democratic representation requires representatives who
share experiences, understand issues from the perspective
of disadvantaged groups, and who are able to constitute a
representative ‘voice’ within deliberations and decision-
making” (Urbinati & Warren,2008, p. 16).