Вы находитесь на странице: 1из 89

WHAT VOTERS WANT FROM OFFICIAL POLITICAL CANDIDATE WEBSITES

A Thesis submitted to the Faculty of the Graduate School of Arts and Sciences of Georgetown University in partial fulfillment of the requirements for the degree of Master of Arts in Communication, Culture and Technology

Sean Soendker Nicholson, B.A.

Washington, DC April 29, 2008

TABLE OF CONTENTS
Chapter 1 Introduction .................................................................................................. 1 Chapter 2 Literature Review ......................................................................................... 4
The Political Scientists Perspective ........................................................................................ 7 The Candidates Perspective .................................................................................................. 16 The Web Professionals Perspective ...................................................................................... 19 The Voters Perspective ......................................................................................................... 25

Chapter 3 Definition of Concepts ................................................................................ 30 Chapter 4 Methodology ............................................................................................... 34


Survey Samples ...................................................................................................................... 35 Survey Instruments ................................................................................................................. 36 Data Transformations ............................................................................................................. 38

Chapter 5 Analysis ....................................................................................................... 40


Overview of data .................................................................................................................... 40 Hypotheses Testing ................................................................................................................ 47

Chapter 6 Findings and Discussion............................................................................. 56


Respondents cited obtaining information as their top priority ............................................... 56 Respondents wanted to support candidates by donating money, volunteering, and attending events ............................................................. 58 Respondents did not expect candidate websites to meet every online need ........................... 59 More public domain data is needed ........................................................................................ 60

Appendices .................................................................................................................... 61
Data Frequency Tables ........................................................................................................... 61 Survey Instrument .................................................................................................................. 74 The MindCanvas Interactive Survey Tool ............................................................................. 78

Bibliography .................................................................................................................. 80

ii

CHAPTER 1 INTRODUCTION
Ten years ago, few political candidates worried about creating websites to support their campaigns. Today, the candidate website has become a fundamental part of the twenty-first century campaign, and voters are interacting with contenders online offerings like never before. While consumption of candidate websites remains relatively limited when compared to other Internet information sources (Kohut 2008), voters are visiting candidate websites with greater frequency. For instance, a survey of approximately 2,000 New Hampshire voters conducted in the days before the states 2008 presidential primary showed that 40% of respondents had visited at least one of the candidates websites (Wheaton 2008). Such widespread online activity in the 2000 or 2004 elections would have been unthinkable. With relatively little time and expense, a political candidate can conceive, design and launch a website, opening the doors to communication with a virtually unlimited audience. However, little has been published on the subject of what voters actually want from candidate websites. Most published work on American voters online behavior have focused on the expansion of Internet access, where voters are getting their information and simply describing what an online campaign has looked like in past election cycles. While such data is invaluable in understanding the changing political landscape, and has most certainly inspired more candidates to adjust their

strategies to match their prospective constituents changing media consumption habits, it is not helpful in answering the question: Ive got this new websitenow what do I do with it? My goal throughout the thesis process has been simple: I want to know how to make powerful, effective websites for campaigns. Working for a political communications firm that specializes in persuasive direct mail and opposition research for the past several years, I have watched as other firms developed websites for our candidates and opponentssometimes in awe, sometimes in disgust. In late 2006, I began researching online political campaign strategies with the intention of turning my on-again/off-again web design hobby into a marketable skill. Like a sponge, I tried to absorb as much as I could, consuming websites and blogs devoted to online campaigning, reading what I could find in magazines and books, and cataloging campaigns websites. What I found was exciting and informative, but frustrating. There is a lack of comprehensive published research as to what target audiences want from sites. Instead, most reviews and suggestions are based an informal content analyses. These texts describe what (generally presidential) campaigns are offering on their sites, and extrapolate user interest based on emphasis on the site. In addition, many authors seemed smitten by the wow factor of new technologies, with little in the way of critical analysis. 2

In response to this vacuum in the literature, this thesis will examine official candidate websites from the end-users perspective, asking the seemingly simple question: What do voters want from official political candidate websites? Of course, this is not a simple question. For instance, how do different target audiences define their desired outcomes? Do desired outcomes vary when comparing presidential candidate websites to mayoral candidate websites? How does a voters attitude change when returning to a candidate website for a second visit? What about the tenth visit? How many visits does the average voter actually make? Building on previous research in political science, communication theory, and outcome-oriented consumer research, I conducted a pilot study exploring what voters want from political candidate websites. Using an interactive online survey tool, respondents were asked to rank potential site components in a variety of manners, and unrestricted, open-ended responses were solicited regarding sites most important features. Upon analyzing the data, the findings were clear: (1) more than anything else, respondents desired issue position and biographical information from candidate websites; (2) respondents wishing to actively support a candidate wanted to be able to donate money, volunteer and learn about campaign events and (3) respondents did not expect a candidate website to fulfill all of their online social interaction and entertainment needs.

CHAPTER 2 LITERATURE REVIEW


This chapter explores the body of research on political candidate websites, and considers the many sources within a simplified framework from three dominant perspectives: that of the political scientist, the candidate and campaign, and the web professional. The strengths and limitations of each perspective for creating better candidate websites are explored. In general, researchers and the public share a common understanding of what constitutes a candidate website: the content accessible on remote computers via the World Wide Web and identified by a unique address (technically the Uniform Resource Locator, or URL). Website activities are a subset of cybercampaigning (Gibson 2004), ecampaigning, and other broad categories of Internet campaigning which also include email, third-party websites and any other activity that involves information and communications technologies (Ward et al. 2003). This thesis is concerned only with what voters want from the official candidate websites (e.g., www.ronpaul2008.com or www.tedkennedy.com), though the full array of online opportunities (e.g., emails, social networking websites, online advertising, discussion boards, YouTube channels) clearly influence voters behaviors and attitudes when interacting with the official candidate websites. Klotz (2004) divides the history of Internet campaigning into 2 parts: (1) 19921999, when the stakes were fairly low because fewer than half of Americans were online, and (2) 2000-on when the Web had become firmly established as a mass medium. In 1992, the presidential campaigns used the Internet in only a limited capacity , and it was not until 1996 that campaigns began to use the World Wide Web

for mass campaigning (Ednres and Warnick 2004), most notably by presidential candidates Bob Dole and Bill Clinton. Soon, other federal candidates began to explore the medium: In 1998, 35% of Democratic and Republican Congressional candidates created websites, and 72% of Democratic and Republican Senate candidates created websites (Druckman et al. 2007). By 2006, 88% of 129 Senate candidates seeking 33 seats listed web sites, and 81% of 1,102 House candidates listed web addresses. Numbers are even higher for major party candidates: in 2006, 97% of Democratic Senate candidates, 94% of GOP Senate candidates, 88% of Democratic House candidates, 84% of GOP House candidates created websites (Gulati and Williams, 2007). Voter consumption of candidate websites has also increased. In 2003, the Pew Internet and American Life Project (as cited in Gibson 2004) reported that only 11 percent of respondents consumed candidate sites, though more than 60 percent looked for election information online. In 2000, the most popular individual US Senate candidate websites were estimated to have received 800,000 visitors. In 2002, traffic for some US Senate sites topped 1.6 million visitors (Druckman et al. 2007, citing Williams 2003). It is no longer a question of if a candidate will create a website, but how. As Foot and Schneider (2006) explains: With the integration of the Internet into every aspect of society, the Web has become a realm in which it is necessary for all kinds of political actors to have a presence-- at least in order to be credible among the growing population of Internet users in the U.S. In the words of a young citizen participating in a focus group on presidential campaign sites in 2000, I 5

expect the next leader of the free world to be able to pick an expert Web designer. Campaigns have increasingly responded to that logic. Candidate websites share many characteristics of other communication media. Like other outreach and communications efforts, a stop-start pattern of behavior is a major trait of candidates online work (Gibson 2004). Candidate websites typically share content, messaging and branding strategies, as will be discussed below. Candidate websites also provide two distinct categories of engagement: (1) increased horizontal communication among people and (2) increased vertical communication between citizens and their political elite (Stromer-Galley 2000). While some have suggested that online activity does not conform to the top-down dynamics of political communication (Krueger 2006), from the candidates perspective, the Internet remains a mostly one-way communication channel. Klotz (2004) provides a useful framework for considering the unique nature of Internet campaigning in relation to other media channels. First, there is a low incidence of accidental exposure (64). The candidate and her campaign know that visitors to their website have at least some interest in the campaign at the same time, it is hard to reach undecided voters who do not even know the candidate or candidate website exists, especially without advertising or traffic driven to the site by external means. Second, viewers are able to choose exactly when, where and how they consume candidate websites. Third, candidates can share their ideas with a large number of people while simultaneously receiving feedback from a large voting audience. As discussed in greater detail below, however, this potential for dialogue or even least the illusion of

dialogue can raise expectations for two-way communications that the candidate is not able to fulfill. Finally, Klotz notes that there is an unlimited and unrestricted online space for unfiltered communication, though this may not necessarily be contained on a candidates website.

The Political Scientists Perspective Academic research has been generally pessimistic about candidate websites. Presidential websites in 1996 were deemed unimaginative, and focusing too heavily on information and not offering greater interaction (Gibson 2004 citing Reavy and Perlmutter 1996; McKown and Plowman 1998; Stone 1996). Critics described 1998s websites as mostly brochureware and virtual billboards, simply replicating in electronic form materials already distributed in print (Foot and Schneider 2006, citing Sadow and James 1999, and Kamark 1999). Candidate websites in 2002 were strongly criticized for their online mobilization efforts, with only around one in ten making basic polling location information available and only seven percent of all sites succeeding in sending any email reminders to
5. 6. 7. 8. Table 1. The Political Scientists contribution to understanding candidate websites Guiding question:

1. How can a candidate website


contribute to a more informed and engaged citizenry? Strengths: 2. Concern with long-term democratic progress, not just imminent election 3. Quantitative and structured analysis, 4. Framework for multiple dimensions of online activity Limitations: Too much content analysis, which is simplistic and crude at times Limited attention to the demand side of the equation User interest is assumed to correlate to campaign activity Too much focus on elite campaigns

supporters to vote (Gibson 2004). However, the outlook has become more positive with the analysis of the 2006 federal election cycle. Trammel et al. (2006) exemplify this change in attitude, writing that candidate websites have evolved from mere brochure-ware to more engaging and interactive tools to inform and mobilize prospective voters and the media. It is appropriate to attribute this general pessimism to the political scientists unique perspective: the political scientists goal is to increase political action and discourse, and serve as a professional observer. While campaigns are concerned with winning an election or advancing their own political agenda, the political scientist concerns herself with increased political action and engagement from the entire citizenry, and is not necessarily concerned with the election of one candidate over another. Stromer-Galley (2000) distinguishes between (1) human interaction and (2) media interaction, asserting the importance of recognizing the technology itself in communication. To ignore media-interaction as a form of interactionis to risk ignoring the important structural components that facilitate or hinder human communication in many contexts [online], including the political. Making this distinction, Stromer-Galley asserts that candidates campaigns have exploited the media interactive capabilities of the medium but neglected the human interactive capabilities. (p. 119). For all of the Internets much-hyped promise as a flattening medium, candidate websites themselves have done little to actually increase direct communication between the citizenry and political elites. As Palser (2008) laments, a 8

true [online] dialogue requires a commitment that most candidates probably will not keep. This posture towards online political activity is reflected in the output of political scientists. Gulati and Williams (2007) provide a useful framework for evaluating the body of academic research on candidate websites, breaking down the literature into three major veins: (1) descriptive 1 , (2) comparative 2 and (3) historical analyses, 3 with quantitative content analysis the centerpiece of most analysis. Candidate websites are simply conducive to content analysis by nature (Klotz 2007), and there existed a need to describe and define the social phenomenon (cite); it is therefore not surprising to find many scales and numerical indicators to evaluate candidate websites. Many follow the powerful lead of Gibson and Ward, who in their 2000 article, A Proposed Methodology for Studying the Function and Effectiveness of Party and Candidate Web Sites, create numerical scales for (1) information and communication flow to assess functionality, and (2) site delivery, to assess how effectively a site delivers its contents. The pair measures information and communication flow with numerical indicators, including newsletter word counts, policy statement word counts, number of links to other websites, presence of website
1 Gulati and Williams (2007) cite Bimber and Davis 2003; DAlessio, 1997; Dulio, Goff and Thurber, 1999; Foot and Schneider, 2002; Klotz, 1997; Puopolo, 2001; Williams, Aylesworth and Chapman, 2002. 2 Gulati and Williams (2007) cite Greer and LaPointe, 2004; Herrnson, Stokes-Brown, and Hindman, 2007; Gibson, Margolis, Resnick, and Ward, 2003; Norris, 2001; Van Dijk, 2005; Ward, Gibson, and Lusoli, 2006; Conners, 2005; Druckman, Kifer, and Parkin, 2006; Endres and Warnick, 2004; Xenos and Foot, 2005. 3 Gulati and Williams (2007) cite Chadwick, 2006; Foot and Schneider, 2006; Bimber and Davis, 2003; Gibson and McAllister, 2006.

search functionality. Their site delivery scale is more complex, assigning numerical values to the presence of animation, HTML frames, size of home page in kilobytes, search engine rankings and other indicators (Please see Table 2).

Table 2. Gibson and Wards (2000) Indicators for Site Delivery Graphics/flashiness Multimedia/dynamism Freshness Accessibility (in principle) Accessibility (in practice) Navigability -> -> -> -> total number of images or pictures moving icons (1), audio (2), video (3), live streaming (4) updated daily (6), 1 to 2 days (5), 3 to 7 days (4), every 2 weeks (3), monthly (2), 1 to 6 months (1), more than 6 months (0) no frames option (+1), text-only option (whole site) (+1), text-only documents to download/print (+1), foreign language translation (+1), blind/visually impaired software (+1) (a) site working (1), inaccessible (0) (b) size of home page in Kb navigation tips (+1), number of search engines (+n), home page icon on each page (+1), major site area links/menu bar on each page (+1), site map/index (+1) same day (5), 1 to 2 days (4), up to 1 week (3), up to 2 weeks (2), up to 1 month (1), more than 1 month (0) number of words, (0) if irrelevant to query (a) number of links in (b) yahoo party index, present (1), absent (0)

-> ->

Responsiveness (speed) Responsiveness (quality) Visibility

-> -> ->

Foot and Schneider (2006) employ an analogous quantitative approach to assessing the informing, connecting, involving and mobilizing communications strategies in their comparative analysis of US candidate websites from 1998-2004. They identified six indicators of informing activity, seventeen indicators of connecting activity, ten indicators of involving activity and four indicators of mobilizing activity (Please see Table 3). Websites were evaluated with a simple binary scale (e.g., there (1) is or (2) is not a candidate biography, link to opponent website, photo of a campaign

10

event, etc.) to constitute the data for the authors conclusion that informing as an online practice was adopted almost universally by 2004, but that virtually none of the analyzed websites (for US House, US Senate and gubernatorial candidates) had established connecting or mobilizing as practices. Benoit and Benoit (2005) created a somewhat more design-oriented metric to evaluate the George W. Bush and Al Gore websites as captured two weeks before the 2000 election. Their scale, which they suggest could be used to analyze and evaluate all political candidate web pages, was comprised of mostly binary indicators of ten different dimensions: identification, navigation, readability, irritability, information accessibility, interest level, information breadth and depth, issues, support, adaptation to audience and interactivity. The political scientists perspective advances the understanding of candidate websites in several key areas: (1) it is concerned with long-term democratic progress and civic engagement, and not just who wins the next election cycle, (2) the quantitative and structured analysis provides useful data for understanding the evolution of online campaigning, which at its bests breaks down sophisticated behavior into core functions and (3) the political scientist reminds us of the multiple dimensions of online activity, while others may have a more limited interest. In addition, the academic literature has also begun to directly ask voters why they are visiting candidate websites, though in a limited manner. This will be discussed in more detail below.

11

Table 3. Foot and Schneiders (2006) framework for Web Campaigning Practices Informing activity a candidate biography issue positions campaign news candidate speeches Involving activity donation information contact information other than email email address volunteer sign-up email list sign-up photos of campaign events campaign calendar visitor comments online polls online events Mobilizing activity online structures that supported sending letters to editors of publications sending links as email from the campaign site to others availability of online material produced for offline dissemination downloadable electronic paraphernalia Connecting activity link to any other site party affiliation endorsements link to political site link to government site link to party site voter registration link to civic/advocacy group site link to press site position alignments position comparison to opponent link to local site comparison section link to citizen site link to opponent site

Note that not all of these indicators are very prevalent (only 1% of candidate sites linked to opponents sites), but this table is a useful example of a political scientists comprehensive construction of all important candidate website components.

The political scientists perspective falls short in several key ways. There has been (1) an overreliance on content analysis, (2) limited attention to user demand, (3) limited analysis of smaller campaigns, and (4) an assumption that candidate behavior is highly correlated to users interest. As noted above, political candidate websites are conducive to content analysis. It is relatively easy to quantify the number of links in a navigation menu, or note whether a biography page exists. However, existing research lacks a substantial analysis of the textual content itself. Indeed, with the increasing

12

availability of website design tools and content management systems, the variation in the existence of many of the measured components cited above is almost non-existent. While a content analysis is helpful in identifying which websites have employed certain techniques, it provides almost no insight into how a candidate may effectively and persuasively communicate with a target audience. For instance, it would be relatively easy to create a website that includes every component measured by Foot and Schneider (2006) above. However, simply because a website includes donation information, contact information, email addresses, volunteer signup forms, email list signup forms, photos of campaign events, a calendar, visitor comments, online polls and online events their entire list of involving activities one cannot conclude that a candidate has done everything possible to involve voters. Benoit and Benoits (2005) scale of interactivity includes a similar list of features, which are included on almost every 2008 presidential candidate website. Again, such a simplistic evaluation checklist does not mean we can conclude that all candidates sites are equally interactive. While one could reasonably respond to this criticism with a more sophisticated checklist of features, such an exercise is doomed to be outdated almost instantly. The use of HTML frames, a key measure of irritability in Benoit and Benoits analysis of 2000 websites, is no longer relevant. Social networking outreach activities of the 2008 presidential contenders could be measured by the number of links to third-party Web 2.0 applications 4 . Such an exercise may provide a glimpse into the willingness of the campaigns to explore social networking techniques, but it tells us nothing about how the
4

As of March 2008, JohnEdwards.com linked to 23 social networking websites. HillaryClinton.com linked to 6, and BarackObama.com linked to 16.

13

campaign may actually be interacting with voters on those sites. Additionally, online networking is most effective as a campaign tool when it translates online activism into concrete, actionable offline behavior. Using social networking sites, campaigns can organize a plethora of online communities; but these communities are only significant when they supplement, compliment and finance offline activism (Armstrong and Moulitsas 2006). Tellingly, similar metrics would not be considered reasonable for candidate communications in other media channels. For instance, television commercials are not evaluated on the total number of images or scenes, or by how many words are articulated by the narrator (a la Gibson and Ward 2000). Candidates direct mail pieces would not be judged to be involving simply because they include the date and location of upcoming events, or the campaigns contact information (a la Foot and Schneider). Instead, we consider communications in other media with a more complex and nuanced lens. Content analyses may be employed, but only as part of a larger methodological battery. Furthermore, the methodology would evaluate the actual language and imagery of the communication, and not just whether the language or imagery exists. Sometimes, without possessing anything but content analysis data, scholars have been willing to make broad declarations of what websites should or must be like. For instance, Williams et al. (2005) conclude their quantitative content analysis of the 2004 Bush and Kerry campaign blogs as follows: It is undeniable that a modern campaign website now requires websites that are interactive and invite user participation (emphasis added). The authors present no evidence for such a conclusion, beyond the 14

fact that Bush and Kerry used blogs and hyperlinks heavily in 2004. Though it is reasonable to argue that interactivity is desirable (and I would), the authors present no evidence of the utility of blogging or hyperlinking activity. They simply describe the frequency of the behavior in one unique historical moment. Furthermore, the authors are making conclusions about the entire universe of campaign websites based on an analysis of only two sites. It should also be noted that online activity by citizens is not always in line with what decision makers and impartial observers regard as proper political activity (McGregor 2006). If we measure online engagement using only the metrics of academics without sufficiently considering the voters perspective, we may be inadvertently undervaluingor overemphasizingspecific behaviors. For example, the Barack Obama campaign used an online phone bank system prior to the early 2008 primaries with greater success than rival campaigns (Stirland 2008). A simple content analysis of the competing efforts would fail to capture the unique dynamics of the Obama effort, and why others strategies were less effective. There are significant shortcomings in the body of content analysis research on candidate websites. In 2000 or even 2004, the limited engagement of candidates online (discussed more below) was appropriate. Now, however, with almost-universal adoption of at least limited Web campaigning, it is time for a more sophisticated perspective.

15

The Candidates Perspective From the candidates perspectiveas well as that of the campaign professional the goal is pretty simple: win the next campaign. The website is a tool, not an end in itself. On one hand, this means that candidate websites must be analyzed somewhat cynically as one analyses any other advertisement (Stromer-Galley 2000) 5 . Civic engagement, the goal of the political scientist, is a happy byproduct for the candidatehis or her true goal is to collect money, to turn out votes for the party, and to win. (Davis 2005) On the other hand, the candidate has the most to gain from an excellent website, and experiences firsthand the consequences of his or her choices in web design, content and form. Scholars have categorized candidates website goals in different ways. Bimber and Davis (2003) break website functions into (1) opinion reinforcement, (2) activism, (3) donating and (4)
Table 4. The Candidates contribution to understanding candidate websites Guiding question: 1. How can a website help me win? Strengths: 2. Has the most to gain by creating powerful, persuasive, websites 3. Comfortable with addressing needs of heterogeneous audience 4. Understands that professed desires may not correspond to actual behavior Limitations: 5. Looking out for #1, leading to choices like overemphasizing donations 6. Pulled in multiple directions by competing constituencies 7. Website is generally not top priority, and may not reflect best efforts

voter registration and mobilization. Rackaway (2007) expands on the work of Reich and Solomon (2007) to define the role of technology for the candidates/campaigns as (1) building community, (2) watching and listening as the campaign unfolds, (3)

Writing in 2000, Stromer-Galley asserts that candidate websites exist (1) primarily to provide information about the candidate in a controlled manner, and (2) to provide a facade of interaction through media interaction in a manner that poses little risk and some benefit.

16

mobilization and (4) the identification of voters. Foot and Schneider (2006), as noted above, break campaigns online strategies into four categories: informing, involving, connecting, and mobilizing. Online communication is superior for the candidate when compared to offline communication forms in some ways. It is (1) less expensive, (2) facilitates better communication with members of the press, and (3) allows for unfiltered communication with more control over content and presentation (Bimber and Davis 2003). However, despite the opportunities that come from creating and maintaining dynamic websites, candidates have been reluctant to embrace the medium. As Graff (2007) explains: Since its inception, the Internet has happened to candidates -- theyve always lagged behind in adopting new technology, failing to realize what it could do and thus being swept along with it as ordinary people step up and transform the process...The Dean 2004 Meetup groups were already growing and swelling even before Joe Trippi put a link on the campaigns home page. When it comes to the Web, the campaigns themselves are more often than not getting yanked out of the station as the train pulls away. (252) Stromer-Galley (2000) cynically explains this phenomenon, suggesting that candidates avoid interactive online channels because they are burdensome, risky in terms of losing control of the dialogue and incompatible with desires to maintain ambiguous discourse. In contrast, Michael Bassik (2008), a Democratic consultant with MSHC Partners who specializes in online political advertising, explains that there are legitimate, understandable reasons for candidates reluctance to invest time and money in online strategies. First, with scarce resources, candidates and campaign consultants will invest in communications strategies that have proven successful in previous campaigns (see also Kippen and Jenkins 2004). Second, there is a continuity gap, in 17

which campaigns are created, executed and disassembled in short time spans, and with ever-changing personnel. Third, online communications can be complex and hard to learn, though this is becoming increasingly less so. Fourth, there is a lack of published, quantitative research establishing which strategies are most effective. Finally, there exists a one-day sale reality for candidates, where marginal improvements in online communication efficacy may be eschewed in favor of large-impact strategies like television advertising. As elections are a zero-sum game with one desired outcometo convince enough voters to like the candidate enough to select him/her over the other choiceswe can assume that campaigns attempt to consider the wants and desires of their voters when designing and maintaining candidate websites. However, the population of voters is not a homogenous block of individuals. Instead, the campaign is constantly weighing the risks and benefits of different strategies, prioritizing the needs of individuals and groups of individuals according to perceive importance. Another significant, and perhaps unavoidable, limitation in the exploration of the candidates perspective is that the public may never be able to know the full candidates perspective. It would be truly shocking if the John McCain, Barack Obama or Hillary Clinton campaigns did not possess reams of data showing which strategies, website components or creative designs are more effective in eliciting desired responses. However, when these campaigns refuse to even disclose the most basic traffic metrics, we cannot expect to learn the finer lessons from their efforts.

18

At the end of the day, the candidate is concerned with winning, and not necessarily with what voters want. This focus on winning, instead of the general democratic engagement sought by the political scientist, sometimes manifests itself in manipulative strategies, desires to do just enough to win, or the prioritization of some website features over others. For instance, prime website real estate is given to monetary donation and email signup features, though such opportunities may not always (if ever) be important to a website visitor. Voters may want to know how a candidate has voted on a myriad of issues, but such disclosure may be detrimental to a candidates electoral success. At the same time, though, the candidate has a vested interest in providing the voter with enough of what he or she wants to inspire loyalty at the voting booth, or even a commitment of time or money.

The Web Professionals Perspective The web professionals perspective is a technical one: how to create a powerful and beautiful communication tool. They are craftspeople, concerned with technical details. This perspective is valuable, as these practitioners are experts in the medium, and almost all innovation and creativity is occurring in the commercial, not political space. The web professional can provide insight not only into the technical creation of Internet tools, but also into usability experience research and consumer behavior data. Voters experiences in the commercial and private online spaces inform and influence their behaviors with candidate websites, and their expectations of what should and should not be a part of candidates websites. While the different threads and suggestions 19

of web professionals could fill volumes, there are several key contributions to the understanding of political candidate websites: (1) focus on providing high-quality, timely information, (2) reach users where they are, using a variety of channels and (3) create websites using the design and navigation conventions familiar to the audience. Most guidance from web professionals is simple and to the point: focus on providing good content, update the content frequently, and make it easy for users to collect the information in the manner that suits them best. As Delany (2008) writes, content is king! In addition to this focus on substantive information, the web professional also entreats candidates to create websites from the users perspective, even going so far as to create conceptual profiles of various visitors to imagine how different types of individuals will interact with the website. Candidates creating websites should imagine what [voters] will be looking for, and what will engage their interest in the candidate and campaign (Ireland and Nash, 2001, p. 40). Sutton (2007) advises political candidates to consider micro-sites to offer the specific information sought by relatively small groups of voters. The web professional perspective brings an understanding into what voters do 20
Strengths: Expertise in online behavior Understanding of user response to web design conventions Strong movement within community to simplify and focus on key functions of sites Limitations: Temptation to emphasize flash and gadgetry over substance Limited understanding of the unique realities of political campaigning Unfair comparisons of political and commercial activity Skewed perspective of online behavior Table 5. The Web professionals contribution to understanding candidate websites Guiding question: How can the form and function of candidate websites be improved?

when they are not on candidate websites. More than any other perspective, the web professional brings quantitative analytic data of user behavior, and consistently refines online products to provide users with what they want. Finkelpearl (2008) suggests that candidates be creative conformists: The web works because key site elements appear in consistent locations across all well-designed sites. While you want your website to have a look and feel that is unique to your organization, make sure you also follow design conventions. For example, people expect to find logos in the top left position of a website. Primary navigation is often in the lefthand side of the page or horizontally positioned across the top of the page. Help links are often in the upper right of pages. Take a look at large well-known sites like EBay, Amazon, Apple, and Target and youll see these conventions repeated over and over. While sometimes these giants break from standard practice, they do so purposefully. For example, Amazon puts their search bar in Table 6. the center of the page rather to the right Colin Delanys 10 (+1) Ways to Build Traffic to a Website since so many people use this feature. Content is king! Amazon has also played with putting sale If content is king, for most sites, ads where their logo typically goes. They niche content is even more know people will glance there. So while kingier. (sic) there may be good reasons for you to break Feed the beast regularly. Be easy to find on search engines. from convention, nine times out of ten you Dont miss an opportunity to put should stick with what people expect to your website address in front of a potential audience. see. Trust us, the good karma will return to Stay in touch. Email signups, RSS, you. This emphasis on creating a website that looks and behaves like traditional, commercial websites reduces a visitors need to learn new navigation rules, and helps users develop confidence and trust in the website (Smashing, 2008).
and content widgets are all ways to help people remember that you exist and that they should read you. Let your readers help. Participate in the online discussion. Contact opinion leaders in your field directly. Consider producing pieces for outside sites. If you have the resources, paid search (Google) ads and carefully targeted blog ads can pay off.

21

Where web professionals provide specific guidance, it tends to be technical in nature, suggesting the employment of a particular tool or strategy. For instance, Zeigler (2006) and Delany suggest that campaigns use syndicated feeds and other delivery technologies for delivering website content to users. Finkelpearl (2008) advises campaigns to pay close attention to typeface and choices, advising them to employ simple, easy-to-read text options guided by cascading style sheets (CSS). A review of literature from web professionals also suggests a strong preference for multimedia content and social networking tools. Sutton (2007) suggests that campaigns use their online social networks to expand the contribution pipeline and volunteer pool, and communicate targeted event news and fundraising ideas, eliminating the need to send spam email to an already connected audience.
Figure 1 Daltons (2007) Analysis of Major Presidential Candidate Websites in PC Magazine

However, some web professionals publications reviewed in the preparation of this thesis offered only superficial evaluations of candidate websites. For examples, Dalton (2007) evaluated the websites of presidential candidates Hillary Clinton, John

22

Edwards, Barack Obama, Rudy Giuliani, John McCain and Mitt Romney (see below) on only three general criteria is an attempt to discern the best campaign site: (1) design and navigation, (2) multimedia and (3) overall. He concluded that Edwards had the best, most full-featured candidate site, because of use trendy colors, easy navigation, and tons of video, audio, podcasts, downloads, and links to social networks. Obama and McCain were given the next best ratings (4 out of 5 dots). Regarding Obama, Dalton writes: good combination of issue-related content, multi-media and social networking features. Regarding McCain, he writes: Well designed and informative. But all the Contribute and Join Us buttons become annoying. There was no discussion of how the websites met the needs of their respective primary audiences in the months before any votes were even cast, or whether While web developers excel at improving
Table 7. The Bivings Groups 10-Point Guide to Build[ing] a Better Political Campaign Website 1. Createagraphicdesignthatsays somethingaboutthecandidate. 2. Givevisitorsstufftodobesides justgiveyoumoney. 3. Recognizeyourbestvolunteers. 4. Makefundraisingpitches specificandtiethemtoeventsin thenews. 5. Cutdownonthenumberof emails,particularlytheones beggingformoney. 6. Don'tblogunlessyouaregoing toembracethespiritof blogging. 7. Publishasmuchcontentas possibleviaRSSfeeds. 8. Ifyougonegative,trytobe cleveraboutit. 9. Provideuserswithabehindthe sceneslookatyourcampaign. 10. Createacommunityaround yoursite.

the aesthetic design and organization of website content, they offer no substantive insight into what content and tools are desired by the voter. Specifics are rarely offered on how good or high-quality content should be constructed, or what makes some content better than others. Moreover, most web professionals possess no real political knowledge or understanding of a candidates unique electoral environment, resulting in 23

analysis and critiques that focus on style and technical details without considering what the candidate may need to communicate to voters to be successful on Election Day. While this limitation of the perspective is significant, it is not surprising. After all, the perfection of content is generally not assigned to the website developer any more often than the writing of a script is assigned to the production manager who edits a television advertisement. 6
Table 8. Ireland and Nashs (2001) Ten Secrets of Successful e-Campaigns and Essential Features of an Effective Website Essential Features: Above all-ease of use Good domain name Regular professional maintenance Download time in fifteen seconds or less Secure online fundraising Volunteer sign-up form Short form to collect email addresses Email-to-friends form (to help visitors tell others) Issues pages Media kits, including photos and press releases Top-quality photos and graphics Events calendar Endorsements page Voter registration info page Well-written Home and Biography pages Privacy policy Site map Contact information for your visitors Ten Secrets: Use the site to promote your whole campaign Your offline campaign must promote your site Respond to emails within 48 hours Give people a reason to come back Learn about your visitors Online donations: the One Percent rule Make volunteering online easy and rewarding Permission campaigning, not spamming Viral campaigning: let your visitors help Dont keep the genie in a bottle

Some talented individuals certainly write scripts and then produce their own spots, but this is not the norm.

24

The Voters Perspective Combining the three perspectives discussed abovethe political scientist, the candidate and campaign, the web professionalwe are able to paint a basic picture of what voters want from political candidate websites. Clear themes emerge: voters are interacting with candidate websites because they want information, they want to actualize their policy goals and participate in the democratic process, they want to be a part of an online community and they want to find enjoyable, novel experiences. That said, it is important to note that there is no single Internet effect on an online audience (Farnsworth and Owen 2004), and that little has been done (at least in the published research) to weigh the relative importance of each goal for target audiences. Indeed, the evidence suggests that desired outcomes evolve with repeat visits to the website and changes in the larger political environment. However, building upon Farnsworth and Owens uses and gratifications framework, the published literature establishes the desired outcomes for voters visiting candidate websites as detailed below. Desire for information. A desire to hear directly from candidates without a media filter is
Table 9. The Voters contribution to understanding candidate websites Guiding questions: How should do I vote? How can I support candidates? Strengths of voter perspective: Winning the support of the voter at the ballot box is the ultimate campaign goal Meeting the needs of the voter visitor provides an opportunity to convert passive supporter into an active one Limitations: There is a wide range in online skills It is impossible to satisfy all voters; different voters have competing and conflicting demands Limited patience for the limitations and constraints of modern campaigning Unfair comparisons of political and commercial activity

25

the most common reason for visiting a campaign website (McGill et al. 1997, as cited in Klotz 2004). Bimber and Davis (2003) found that browsing was the most cited reason for visiting the 2000 presidential candidate website. Farnsworth and Owen (2004) confirm that the desire for information is a primary reason for visiting websites, along with interactive functions and becoming involved in campaign activities (discussed in more detail below). In fact, the desire for information seems to be even more pronounced with undecided voters and return visitors: Among people inclined to vote, but not strongly decided, for Bush or Gore, 45 percent listed learning about issue positions and 9 percent listed learning about the individual as their reason for visiting the candidates site, compared with 26 percent and 3 percent, respectively, for those who already had a strong candidate preference. (Bimber and Davis 2003, 116) and Across all of the candidates sites, far and away the most common reason for return visits was to learn about issues....Repeat visitors are knowledgeable, well-educated, partisan and seeking more issue information in support of their political preferences. (Bimber and Davis 2003, 130) In addition, when seeking candidate information, voters want to control how and when they access said information (Benoit and Benoit 2005, 234). From his research on candidate blogs, Trammell (2006) found that audiences desire personal information of candidates, suggesting that (1) direct addresses, (2) calling a candidate by his or her first name or (3) posting of content written by the candidate contribute to humanized, text-based interactivity (Trammell et al. 2006).

26

These techniques promote immediacy, personal presence, and multivocality (Endres and Warnick, 2004, as cited in Trammell et al. 2006). Desire for participation in the democratic process. The opening up of the democratic process is one of the most discussed opportunities of Web campaigning, in both academic and popular literature. Farnsworth and Owen (2004) found that interactive functions and opportunities to partake in campaign activities were major goals of users engaging in politics online. The revolution in online politics is about participation, Garrett Graff writes in The First Campaign (2007). Jerome Armstrong and Markos Moulitsas, prominent Democratic bloggers, in their Crashing the Gate manifesto (2006) write that their activism is focused on building a broader movement. Though it is unclear how this enthusiasm for democratic engagement manifests itself on candidate websites differently than it does in other Internet forums, websites do help engage supporters through reinforcement, solicitations for activism, donating and mobilization on election day (Bimber and Davis 2003). Gibson suggests that simply visiting a candidate website may contribute to a sense of involvement in the political process: The requirement of finding the website, although it may raise the barrier to reaching a mass audience, may also be vital to creating a sense of ownership and control among users once they arrive at the site (Gibson 2004). Peng et al (1999) write that voters want to do more than just consume content; instead they want to control their access to content through the use of hyperlinks, have the opportunity to contribute to a site, and go beyond passive exposure (quoted in Williams and Trammell 2005). 27

Citing the Kettering Foundations focus groups from 1991, Stromer-Galley (2000) asserts that Americans desire more access to participation in the process of setting the public policy agenda than accountability of politicians. Desire for community. Leadbeater (2008) writes that online political activists, especially those that engage in political blogging and online political communities, want to [socialize] and get recognition. Most literature discussing the desire for community in online political spaces focused on non-candidate political websites (e.g. Redstate.com, DailyKos.com), though Teachout and Streeter (2007) convincingly argue that supporters of Howard Deans 2004 presidential campaign built a community around Deans candidate website, using social networking websites, email and other online tools. Desire for novelty, enjoyment. Farnsworth and Owen (2004) also suggest that people are especially drawn to the unique functions and uses of the Internet that are either not available or less accessible via other media. Candidate websites allow for unique presentations of information that are simply not feasible in more traditional media. Also, there are no time limits for Internet communication, allowing the producers and consumers of content to interact for as long as the consumer desires. Factors influencing voters perspectives. There are few systematic differences between [decided and undecided voters] in motivations for first visiting a Web site. (Bimber and Davis 2003). More from Bimber and Davis (2003): Another factor affecting motivations was overall level of interest in the campaigns...[we] found that those voters most likely to visit a campaign site in order to learn about issues were not those most interested nor 28

those least interested, but those with a moderate level of interest in the campaigns. We do not have a clear understanding of why this is the case. The most likely possibility revolves around the observation that issue information has different meanings for voters than personal information about candidates (116). Political interest and online skill are the most powerful independent factors on Internet political participation (Krueger 2006). Those with the characteristics that predict conventional mobilization continue to hold a mobilization advantage in the new technological environment; only the mechanism by which they gain an advantage changes (Krueger 2006). We cannot discount the disparities in online access that impact different populations interactions with the online political process. As time passes, limitations on broadband access are reduced, but there will continue to be inconsistencies in Internet reach. For one, there are massive racial and socioeconomic differences a candidate confronts when moving a campaign online (Graff 2007). In addition, some individuals voluntarily choose not to actively participate in Internet activity (Murdock and Golding 2004). Building upon this body of research, interviews were conducted with a purposive sample of individuals to aide in the construction of a survey instrument. The goal of development, deployment and analysis of the survey instrument was twofold: to contribute to the body of knowledge regarding what voters want from candidate websites, and to develop a pilot study to guide future research on the subject by researchers with the means to collect data from a sample more representative of the full voter population. 29

CHAPTER 3 DEFINITION OF CONCEPTS


This chapter outlines conceptual and operational definitions of the key concepts discussed in the thesis. The methodology chapter that follows provides a comprehensive account of how these concepts were measured in the survey instrument. The political candidate website. This thesis is concerned only with what voters want from the official candidate websites (e.g., http://www.ronpaul2008.com or http://www.tedkennedy.com). While websites must be considered within the full array of online communication channels and opportunitiesemail, social networking tools, online advertising, discussion boards, media sharing websites, etcall questions in the survey instrument were focused on the candidate website. The instrument uses the language political candidate website separately from political information. Voters. For the purposes of this thesis, any individual at least eighteen years old is considered a voter, on the assumption that any of these individuals could register to vote for a future election if needed. It was assumed that respondents not from the United States would be eligible to vote in their native country if they were at least eighteen years old. It was also assumed that no respondents were disenfranchised because of felony convictions. Issue information. The role of issue information in political candidate websites is a central part of this thesis. Conceptually, issue information refers to any statement by the candidate of belief or intention regarding a public policy matter. Though usually found in a clearly marked issue section of a candidate website, I did not assume that voters desired candidate websites to be organized in a conventional manner when 30

constructing the survey. In the survey instrument, issue information is referred to as learning how a candidate feels about issues or how a candidate feels about specific issues. When coding the open-ended responses, any reference to issue positions, position statements, how a candidate would vote, voting records or white papers was considered issue information. Biographical information. Biographical information refers to any text on a political candidate website detailing the candidates life story, previous experience, tenure as an elected official or list of positions, occupations and associations. As noted above, a candidates voting record was considered issue information for the purposes of analysis. Biographical information was defined for the respondent as the candidates background or the candidates biography in the survey. Political information. Political information consumed online refers to any content consumed on the Internet involving public policy, elections or other political issues. The survey instrument was intentionally written so as not restrict the respondents understanding of political information. Monetary donations. The donation of money to political campaigns is governed in the United States by the Federal Elections Commission. With some exceptions, 7 most citizens may donate their own money to benefit the candidacy political candidacies. This function on political candidate websites is referred to as making monetary donations or donating money in the questionnaire. It was

Basic information regarding donations to federal candidates may be found at http://www.fec.gov/pages/brochures/citizens.shtml. Laws vary for candidates in state and local elections.

31

assumed in the coding of open-ended responses that phrases such as accepting contributions referred to monetary donations, and not the receipt of goods or services. Volunteering. Volunteering refers to any unpaid service done on behalf of a political candidate. While some political candidate websites list specific ways an individual may support a candidacy (e.g., assisting with canvassing tasks, making phone calls to prospective donors or voters, driving voters to polls on Election Day, etc.), specific examples of volunteering activity were not listed in the questionnaire. Event information. Event information refers to any logistical data that pertains to public events conducted to support a political candidacy. Open-ended responses coded as event information included phrases such as a list of upcoming events and the candidates schedule. Social networking. In a broad sense, social networking refers simply to the active expansion of an individuals social contacts, whether at physical events, through correspondents, or in the online space. As used in the questionnaire and in this thesis, social networking refers specifically to the genre of websites that are devoted to the meeting of new individuals and maintenance of social ties between individuals and among groups in the online space. Political candidates are able to integrate some functions of social networking websites into their own website, allowing users to publicly support the candidate, choose to receive information from the campaign, and other activities. Facebook, Twitter and MySpace were listed as specific examples of social networking websites in the questionnaire, though there are literally hundreds of different social networking websites with which a candidate and voter may interact. 32

Video. Video, as used in this thesis, refers to any non-animated movie or film, regardless of where the movie file is hosted on the Internet. Any open-ended responses that referred to YouTube or other video hosting websites were coded as video.

The above conceptual and operational definitions were used to guide the development of the survey instrument, as detailed below. The methodology chapter that follows provides a comprehensive account of how these concepts were measured in the survey instrument.

33

CHAPTER 4 METHODOLOGY
The following chapter details the construction and deployment of the pilot study survey instrument. A discussion of the convenience survey samples and data transformations is also included. To explore the research questionwhat do voters want from political candidate websitesI conducted a pilot study using an online survey instrument. Early in the research process, I chose to use MindCanvas online research service 8 instead of more traditional online survey tool because of its unique interactive card-sorting and gamelike approach to gathering data from respondents. The MindCanvas tool allowed me to employ two unique tools not available with a Survey Monkey (or similar) survey tool: an online card sorting exercise and a Divide-the-Dollar exercise that asked respondents to prioritize possible candidate website features in a game-like interface. 9 To help refine questions and language for the survey instrument, I interviewed four individuals with diverse experiences with candidate websites: (1) a Washington, DC political beat reporter, (2) a campaign manager who had recently completed a state legislative campaign, (3) a campaign consultant with a background in political journalism, and (4) a second campaign consultant who is active in several local-level political organizations and formerly served as a Congressional campaign staffer. The survey instrument included questions about respondents experiences with candidate websites, game-like interactive exercises that asked respondents to prioritize
8 9

More information may be found at http://www.themindcanvas.com/how-it-works/. Please see the Appendix for screenshots of the survey instrument.

34

possible website components and opportunities, political ideology and partisanship questions, and basic demographic questions (sex, race, education).

Survey Samples In November 2007, an initial survey was deployed using a convenience sample. Respondents were recruited via email, tapping available social networks in the Washington, DC and Northeast, Missouri regions. Initial contacts were encouraged to share the survey with their contacts, and individuals from many parts of the United States, plus several in Europe, completed the initial survey. After evaluating the Fall 2007 survey, several questions were added to the instrument, and a second round of respondents was recruited in March 2008. To prevent an overlap in samples, respondents in 2008 were recruited only via the Facebook social networking platform, an email listserv not used in 2007, and several online discussion boards: DailyKos.com (a liberal political blog), RedState.com (a conservative political blog), OpenLeft.com (a liberal political blog) and HoyaSaxa.com, an online bulletin board targeting Georgetown University students and alumni. Both surveys were in the field for approximately three weeks. Though the combined convenience samples makes it difficult to make firm predictions about the nature of the full voter population, the politically-engaged nature of my sample is useful in understanding the desires and behaviors of visitors to political websites, who are more likely to already to be politically engaged.

35

Survey Instruments Online survey instruments 10 distributed to the two samples described above included the following key questions and activities: Two multiple-choice questions asking about past interactions with candidate websites, One multiple-choice question asking how often respondents consume political information on the Internet, One card-sorting activity that asked respondents to drag individual cards, each with a discrete website component (see Table 10), into five different categories: definitely include in a candidate website, probably include in a candidate website, probably avoid in a candidate website, definitely avoid in a candidate website, and not sure. One coin-allocation activity in which respondents were asked to allocate twenty coins (each labeled as nickels, totaling one dollar) among an array of discrete website components (see Table 10). Respondents could assign any number of their twenty coins to any component, resulting in values of $0.00 to $1.00 in five cent increments. Seven questions asking respondents to agree or disagree with the following statements on a five-point scale, where 5=strongly agree, 4=agree, 3=neither agree nor disagree, 2=disagree, 1=strongly disagree: o Obtaining information is my primary goal when visiting a campaign website.
10

Please see Appendix II for exact instrument language, and Appendix III for screenshots of cardsorting and coin-allocation activities.

36

o I really like video on a campaign website. o I enjoy multimedia offerings on campaign websites, but not at the expense of information. o I prefer a simple site to a complex site. o I would not vote for a candidate who did not have a website. o Websites affect the way I vote. o I would design candidate websites differently. (Note: These questions were not included in the Fall 2007 instrument) One open-ended question asking respondents about recollections of specific websites they liked and did not like. One open-ended question asking respondents what they considered to the most important part of a candidate website. One open-ended question asking respondents what advice they would give in three sentences to campaigns developing a candidate website. Demographic questions about age, ZIP code, gender, race/ethnicity, education level, voting behavior, ideology and partisan affiliation. (Note: Voting behavior and partisan affiliations were combined into one question in the Fall 2007 instrument)
Table 10. Website Components Listed in Card-Sorting and Coin Allocation Activities*

Connecting with other volunteers Integration with social network sites Learning about campaign events Learning about the candidates background

Learning why people like you are supporting the candidate Making monetary donations Purchasing merchandise Reading news stories related to campaign

37

Learning how a candidate feels about issues Learning how to support the candidate with your friends Learning simple ways to support the candidate

Reading or watching candidate speeches Registering to vote Seeing photos of the candidate Signing up for campaign emails Volunteering for the campaign Watching videos of the candidate

*The display order of these choices was randomized in both the card-sorting and coin allocation activities.

Data Transformations Data was collected on the MindCanvas website, downloaded and aggregated in SPSS. Data transformations included the following: Reponses from the card-sorting activity were transformed into a 5-point scale, where 5=definitely include, 4=probably include, 3=neutral, 2=probably avoid, 1=definitely avoid. Qualitative responses about candidate websites that respondents liked and did not liked were coded to measure the frequency of candidates notes (e.g., the number of respondents who liked John McCains website). Descriptions of what was liked and disliked about said websites were not consistent enough to develop a useful coding system. Qualitative responses regarding what individuals considered to be most important in candidate websites and advice for developers of candidate website were systematically coded. As appropriate, responses were coded according to components deemed most important (e.g., focus on issue positions, or focus on donations) or stylistic recommendations (e.g., keep the site simple or use clear language). Responses were assigned with as many codes as appropriate to capture the thrust of the individual thoughts.

38

The creation of a folded ideology scale, where respondents who identified themselves as very liberal and very conservative were re-coded as having a strong ideological preference. Self-identified liberals and conservatives were re-coded as having a weak ideological preference and independents were re-coded as having no ideological preference. The resulting 3-point scale was as follows: 3=strong ideological preferences, 2=weak ideological preference, 1=no ideological preference.

Data collected from the Fall 2007 and Spring 2008 instruments are analyzed in the following chapter. Top-line frequency information is followed by the testing of the thesis hypotheses.

39

CHAPTER 5 ANALYSIS
This chapter summarizes the data and statistical analysis of the combined survey sample, focusing especially on responses regarding professed desires in political candidate websites. Top-line frequency information is followed by the testing of various hypotheses. Data were supportive of the working theories that consuming information was the top goal of voters, and that donating money and volunteering are the most important ways voters wished to support candidates. Overview of data Of 387 respondents, just over 80% (n=314) have previously visited candidate websites. Most respondents had consumed websites for both candidates they already supported (69.8%) and candidates they did not support (59.9%). Interestingly, only 55% of respondents (n=212) said they consumed a candidate website to assist in their decision about an upcoming election, suggesting that existing supporters constitute a significant portion of a candidates website audience. Communication between respondents and candidate campaigns via the websites was, for the most part, onedirectional. While 72% of respondents read issue statements and 71% read
Table 12. How often do you look for political information on the Internet? Response Every Day A few times a week Almost never A few times a year A few times a month Once a month Total 183 67 61 34 32 10 % 47.3 17.3 15.8 8.8 8.3 2.6 Table 11. Respondents looked at candidate websites Response To learn about a candidate they did support To learn about a candidate they did not support To get information to help with decision about an upcoming election Have not looked at candidate websites Total 270 232 212 74 % 69.8 59.9 54.8 19.1

40

candidate biographies, only 21% used the website to share opinions or ideas with candidates. A larger number of respondents submitted information for donation (41%) or volunteer (30%) purposes. Information gathering was the primary activity of respondents, as indicated by multiple measures. Almost 72% (n=278) of respondents visiting websites said they read issue statements, and 71% (n=273) read candidate biographies. When asked to name the most important feature of candidate websites in an open-ended question, 68.2% of respondents (n=264) replied candidates issues positions. Tellingly, the next most popular response, donation features, was stated by only 85 respondents (22.0%). In the coin sorting activity, 11 issue statements were overwhelming judged to be most valuable to voters (x =$0.20), which was almost twice as valuable as the next highest feature (candidate biographies x =$0.11). It should be noted that the sample was more Democratic, overwhelmingly white and more interested in political information than the full voting population. Almost half of the combined
Table 13. When visiting campaign websites, respondents. Response Read issue statements Read the candidates biography Just browsed Signed up for email updates Obtained information about campaign events Donated money Read the campaign blog Signed up to volunteer Requested a sign, bumper sticker, etc Told the candidate your opinion Met others Total 278 273 263 189 189 159 123 116 102 80 42 % 71.8 70.5 68.0 48.8 48.8 41.1 31.8 30.0 26.4 20.7 10.9

11

As described above, respondents were asked to distribute twenty nickels (totaling $1.00) to represent how they would prioritize website features. Respondents could allocate all or none of their coins to any of seventeen options. A full data chart may be found in the Appendix I.

41

sample (n=183, or 47.3%) consumed political information online on a daily basis. Sixty percent of respondents have completed some post-graduate coursework towards a master's or other professional degree, and only fifteen percent of the sample had not completed a bachelors degree. Almost ninety percent of respondents were white, with other constituting the next largest ethnic group. Almost sixty percent of respondents described themselves as Democratic voters, though almost twenty percent said they were neither liberal nor conservative. To define the different dimensions of voters goals when visiting candidate websites, I employed a principal components analysis with a varimax rotation using responses from the card sorting activity . Factors were extracted with eigenvalues of 1 or greater, revealing five distinct dimensions: (1) consuming candidate
12

Table 14. Rotated component matrix for factor analysis of card-sorting activity
Component 1 Donating Money Email Signup Volunteering Event Info Issue Statements Candidate Bio Watching Speeches Merchandise Online News Connecting With Other Volunteers Social Networking Simple Ways To Support Candidate Ways To Support With Friends Why Others Support Candidate Voter Registration .816 .778 .771 .720 2 .847 .845 3 .686 .572 .529 .529 .461 4 .777 .702 .572 5 .884

Extraction Method: Principal Component Analysis Rotation Method: Varimax with Kaiser Normalization Rotation Converged in 7 iterations

12

See Appendix 2 for language and Appendix 3 for screenshots of online sorting activity.

42

issue and biographical information, (2) using the website as a channel to offline action, (3) interacting with candidacy in the online and social spaces, (4) learning how to support candidacy, and (5) voter registration and voting.

Consuming candidate issue and biographical information The clean definition of the issue statement and candidate biography dynamic in the factor analysis is supported by data from all parts of the survey. Reading issue statements and reading candidate biographies were the top two activities respondents reported doing on candidate websites71.8% and 70.5%, respectively (n=387). When asked to assign values to possible website components in the card sorting activity, the candidate biography ranked first in overall importance, and issue statements ranked third (campaign events, discussed below, ranked second). A total of 280 respondents said candidate biographies should definitely be included (x =4.67). 13 Only six respondents said issue statements should be avoided (x =4.38). Looking at the aggregated data for the coin allocation exercise, issue statements were deemed the most valuable (x =$0.20) by an overwhelming margin.
Table 15. Obtaining information is my primary goal when visiting a campaign website.
Response Strongly Agree Agree Neutral Disagree Strongly disagree Frequency 165 60 8 5 3 Percent 68.5 24.9 3.3 2.1 1.2

N=241, Mean=4.57

13

Responses in the card-sorting activity were re-coded in SPSS on a 1 to 5 scale, with 1=Definitely Avoid, 2=Avoid, 3=Neutral, 4=Include, 5=Definitely Include.

43

Candidate biographies were the second most valuable options (x =$0.11). Indeed, when Spring Sample respondents were asked directly at the end of the survey about their agreement with the statement, obtaining information is my primary goal when visiting a campaign website, an overwhelming 93% either agreed or strongly agreed (n=241, x =4.57).

Using the website to generate offline action As shown above, a second strong dynamic revealed in the factor analysis includes (1) donating money, (2) volunteering, (3) signing up for email alerts and (4) learning about upcoming campaign events. In fact, a significant portion of the sample had previously engaged in such activity on candidate websites: 48.8% had subscribed to campaign email lists, 48.8% obtained information about campaign events, 41.1% donated money and 30.0% signed up to volunteer online. The inclusion of campaign events scored highest in the card-sorting exercise (x =4.61), followed by volunteering (x =4.29), donating money (x =4.22) and signing up for campaign emails (x =4.18). Among the potential website features for generating offline action, the ability to donate money and to sign up for volunteer opportunities were deemed most valuable in the coin allocation exercise (x =$0.09 and x =$0.07, respectively), followed by obtaining
Table 16. Top Ten Features in Coin Allocation Exercise Feature Issues Biography Donating money Volunteer opportunities Speeches Event information Simple support strategies Voter registration News stories Email signup x 0.20 0.11 0.09 0.07 0.07 0.06 0.06 0.06 0.06 0.05

44

event information (x =$0.06) and email subscriptions (x =$0.05). When asked to provide advice for campaigns and candidates creating websites, focusing on donations and volunteer opportunities was suggested by 16.5% and 14.5% of respondents respectively, more than any other campaign support activity. Only suggestions to focus on issue statements (advised by 36.2% of respondents), make the site easy to navigate (26.6%) and keep the site simple (26.6%) were offered more frequently.

Interacting with the candidate in online and social spaces The third dimension of responses revealed in the factor analysis includes (1) watching speeches online, (2) purchasing merchandise, (3) reading news stories, (4) connecting with other volunteers and (5) integration with social networking sites. While the merchandise and news stories variables smeared across other dimensions more than any others in the factor analysis, these five variables can be understood as a general desire by voters to interact with campaigns in virtual spaces. Of these five website components, the ability to watch speeches and consume news information were most important to respondents, though considerably less important than issue information and biographical information, as discussed above. Still, the desire for speeches ranked fifth in the
Table 17. Top Ten Features in Card-Sorting Exercise Feature Candidate biography Learning about campaign events Issue statements Volunteering Learning simple ways to support candidate Donating money Reading or watching speeches Signing up for campaign emails News stories Voter registration info x 4.67 4.61 4.38 4.29 4.23 4.22 4.20 4.18 4.14 4.12

45

coin allocation exercise (x =$0.07) and sixth in the card-sorting exercise (x =4.20). Attitudes toward merchandise and integration with social networking sites were far more ambivalent. In fact, integration with social networking websites ranked last in the coin allocation exercise (x =$0.02).

Learning how to support a candidate In reviewing data from the card-sorting exercise, it is clear that voters want to learn simple ways to support the candidate (x =4.23). When combined with learning how to support the candidate with friends (x =3.91) and learning why other people are supporting the candidate (x =3.75), these three variable constitute the fourth dynamic revealed by the factor analysis. These site components were deemed relatively unimportant by respondents in the coin allocation exercise. With such a limited number of respondents expressing a desire to learn why other people support the candidate (x =$0.04), connect with other volunteers (x =$0.03) or learn how to support candidates with friends (x =$0.02), there simply is not enough data in this sample to create a statistically-supported profile for the kind of voters who would be interested in this information.

Voter registration and voting The final dimension revealed by the factor analysis was the desire for voter registration information, which stands alone in the statistical model. Respondents were

46

clearly interested in voter registration information (card-sorting x =4.11, coin allocation x =$0.06), but it was not considered to be as fundamental to the candidate website as the components discussed above.

Hypotheses Testing As discussed above, the survey instrument was designed to test three guiding theories and their corresponding hypotheses. First, that obtaining information about the candidate is the primary goal of website visitors. Second, that signing up to volunteer and donating money are the most important actions to candidate website visitors. And third, that while voters may enjoy multimedia features as a means for delivering information, voters consider such features as means to an end. Formal hypothesis and relevant statistical analysis to test these theories are detailed below.

Working Theory 1: Obtaining information about the candidate is the primary goal of website visitors. Part 1: Comparing mean issue information and biographical information scores to mean scores for other variables. To explore this first working theory, I will test the following hypotheses related to the key variables that constitute the information dynamic and the marker variables for the four other components from the factor analysis (discussed above):
H1a There is statistically significant difference between the desire for issue information and the desire to for biographical information.

47

H1b H1c H1d H1e H1f H1g H1h

There is statistically significant difference between the desire for issue information and the desire to donate money. There is statistically significant difference between the desire for issue information and the desire to read or watch speeches. There is statistically significant difference between the desire for issue information and the desire to learn simple ways to support a candidate. There is statistically significant difference between the desire for issue information and the desire for voter registration information. There is statistically significant difference between the desire for biographical information and the desire to donate money. There is statistically significant difference between the desire for biographical information and the desire to read or watch speeches. There is statistically significant difference between the desire for biographical information and the desire to learn simple ways to support a candidate. There is statistically significant difference between the desire for biographical information and the desire for voter registration information. There is no statistically significant difference between the desire for issue and biographical information and other website components.

H1i

H0

A paired samples t-test was used to compare means from both the card-sort and coin allocation activities. The difference in means between issue information (x = $0.1973) and each of the other coin allocation variables was found to be statistically significant at the p=.000 level. The difference in means from the card-sorting was less pronounced because of the activitys more condensed 5-point scale, but still supportive of H1a. The difference between the issue information mean and most of the other variables means was significant at the p=.000 level. Exceptions included donating

48

money (p=.006), simple ways to support a candidate (p=.006), and volunteering (p=.116). The data from the collected samples support H1a, H1b, H1c, H1d, and H1e. The data is similar regarding the biographical information hypotheses. As noted above, biographical information ranked highest in the sample in the card-sorting activity. While deemed considerably less important than issue information in the coin allocation exercise, it was still the second-most important website feature. The difference between the mean biographical information score in the coin allocation exercise was significant at the p=.000 level for almost all variables. Only the paired samples test between biographical information and donating money was different, though still significant at the p=.019 level. The difference in means was significant for all comparisons in the card-sorting activity at the p=.000 level, except for the paired samples test with event information, which was not statistically significant (p=.102). The data from the collected samples support H1f, H1g, H1h, and H1i, allowing us to reject the null hypothesis. An examination of the greatest differences of means is interesting. Looking at the sample data, the greatest difference in means for the card-sorting activity was between issue information and social networking (a difference of 1.369). Merchandise, connecting with other volunteers, learning why similar people are supporting the candidate and photos rounded out the five features shown to be least important in the sample. Their relatively low mean scores were the result of respondents ambivalence more than active disgust. It is worth noting that the lowest mean was 3.46 (with 3.00 at

49

the middle of the five-point scale), suggesting that respondents are not turned off by any of the proposed features, at least in the abstract, but simply ambivalent to their presence. Part 2: Correlating the desire for general political information to the desire for candidate-specific information. To further explore Working Theory 1, I tested the following hypotheses related to the frequency of political information consumption online and the desire for issue, biographical and event information:
H2a H2b H2c H0 As online consumption of political information increases, desire for issue information increases. As online consumption of political information increases, desire for biographical information increases. As online consumption of political information increases, desire for event information increases. As online consumption of political information increases, desire for event information increases.

To test this hypothesis, the frequency of consuming online political information was correlated with the desire to learn about issue statements, the desire to learn about biographical information and the desire to learn about campaign events. A statistically significant positive relationship was found between the frequency of consuming political information online and the desire for issue information (Pearsons R=.143 at p=.005) and the desire for event information (Pearsons R=.167 at p=.001). There was a weak, but not statistically significant relationship between
Table 18. Correlations Between Frequency of Online Political Information Consumption and Desires for Various Website Information Components Frequency of consuming political information vs. Issue information Biographical information Event information Pearsons R=.143 p=.005 Pearsons R=.043 p=.398 Pearsons R=.167 p=.001

50

the frequency of online political information consumption and the desire for biographical information (Pearsons R=.043 at p=.398). H1e and H1g were supported; H1f was not supported.

Working Theory 2: Signing up to volunteer and donating money are the most important actions to candidate website visitors, especially to strong partisans and ideologues. Part 1: Comparing mean monetary donation and volunteering scores to mean scores for other variables. To explore this second working theory, I will test the following hypotheses related to the volunteering and monetary donation variables:
H3a H3b There is a statistically significant difference between the desire for monetary donation opportunities and the desire for issue information. There is statistically significant difference between the desire for monetary donation opportunities and the desire to read or watch speeches. There is statistically significant difference between the desire for monetary donation opportunities and the desire to learn simple ways to support a candidate. There is statistically significant difference between the desire for monetary donation opportunities and the desire for voter registration information. There is statistically significant difference between the desire for volunteer information and the desire for issue information. There is statistically significant difference between the desire for volunteer information and the desire to read or watch speeches. There is statistically significant difference between the desire for volunteer information and the desire to learn simple ways to support a candidate. There is statistically significant difference between the desire for volunteer information and the desire for voter registration information.

H3c

H3d

H3e H3f H3g

H3h

51

H0 H0

There is no statistically significant difference between the desire for issue and biographical information and other website components. There is no statistically significant difference between the desire for monetary donation and volunteer opportunities and other website components.

A paired samples t-test was employed to compare differences of means in the evaluation of H3. When looking at the coin allocation activity, the difference in mean scores between volunteer information and every other variable was found to be statistically significant (p=.000 for all paired samples t-tests, except for the comparison of donations and biographical information, where p=.019). Results were mixed with the card-sorting exercise. H3a was supported (p=.006), but H3b, H3c, and H3d were rejected, with p values of .846, .802 and .146, respectively. When looking at the coin allocation activity, the difference in mean scores between volunteer information and almost every other variable was found to be significant at the .000 level. Exceptions included the head-to-head test between volunteer information and speeches (p=.398), volunteer information and voter registration information (p=.010), volunteer information and event information (p=.005), and voter information and campaign news (p=.001). The difference between the mean volunteer information score from the card-sorting activity and the following variables was not significant: reading or watching speeches (p=.135), learning simple ways to support a candidate (p=.239), issue information (p=.116) and biographical information (p=.096). Accordingly, H3e, H3f, and H3g were rejected. H3h was supported. Data were mixed for H3, and were not sufficient to reject the null hypothesis.

52

Part 2: Correlating the desire for fundraising, volunteering, email signup and event information capabilities with a folded political ideology variable. To further explore Working Theory 2, I tested the following hypotheses related to the intensity of ideological preference and the desire for fundraising, volunteering, email signup and event information capabilities:
H4a Voters with strong ideological preferences will be more likely to desire fundraising capabilities than voters without strong ideological preferences. Voters with strong ideological preferences will be more likely to desire email subscription capabilities than voters without strong ideological preferences. Voters with strong ideological preferences will be more likely to desire volunteer signup opportunities than voters without strong ideological preferences. Voters with strong ideological preferences will be more likely to desire campaign event information than voters without strong ideological preferences.

H4b

H4c

H4d

To test the relationship between ideological preference and the desire for various candidate support functions, a folded ideological variable was created. Respondents at both ends of the spectrum those who identified themselves as very liberal and very conservativewere re-coded as
Table 19. Correlations Between Ideology and Desires for Various Website Components Folded Ideology Scale vs. Donating money Email signups Volunteering Event information Pearsons R=.175 p=.001 Pearsons R=.128 p=.013 Pearsons R=.203 p=.000 Pearsons R=.069 p=.177

53

having a strong ideological preference. Self-identified liberals and conservatives were re-coded as having a weak ideological preference and independents were recoded as having no ideological preference. The resulting 3-point scale was as follows: 3=strong ideological preferences, 2=weak ideological preference, 1=no ideological preference. Statistical tests were then conducted to determine the correlation between this folded ideological variable and the card-sorting variables for donating money, volunteering, email signups and event information. The tests revealed a statistically significant positive correlation between ideological strength and the desire for online donation (Pearsons R=.175 at p=.001), supporting H4a. A statistically significant positive correlation was also found between the desire for email signups (Pearsons R=.128 at p=.013) and the desire for volunteer opportunities (Pearsons R=.203 at p=.000). However, there was not a statistically significant relationship between ideology and the desire for event information. H4a, H4b, and H4c were supported by the data. H4d was not supported. The relationship between ideology and the desire for volunteering information was strongest. Combining the frequency data, factor analysis, correlation tests and paired sample t-tests, the data from my sample supports my first working theory that obtaining information about the candidate is the primary goal of website visitors. The first batch of hypotheses were supported concerning the statistical significant different between the desire for issue and biographical information when compared to other website components. Indeed, the desire for issue information was most profound in the coin 54

allocation activity, where respondents considered issue information approximately twice as valuable as any other website component. Biographical information and issue information were closely related in respondents minds, per the factor analysis, and biographical information actually had a slightly higher mean score in the card-sorting activity. Support was mixed for the second batch of hypotheses concerning the relationship between the frequency of online political information consumption and desire for issue information, biographical information and event information. A relationship was found in the sample data between frequency of general political information consumption and the desire for issue and event information, but not biographical information. Data for my second working theory regarding respondents most valued candidate-supporting actions were mixed. Though the ability to volunteer and the ability to donate money were shown to be most important in evaluating response frequencies, relatively few statistically significant differences in means were found. That said, the hypotheses concerning the positive relationship between strong ideological preferences and the desire for donating money, signing up for emails and volunteering were supported. The following chapter explores these data in detail, incorporating the qualitative open-ended responses and juxtaposing my research with the literature reviewed in Chapter 2.

55

CHAPTER 6 FINDINGS AND DISCUSSION


This pilot study suggests several key takeaways regarding what voters want from candidate websites. Obtaining knowledge (i.e. issue and biographical information) was the top priority for survey respondents. Donating money, signing up to volunteer and learning about campaign events were also shown to be very important in the collected data. Video, multimedia offerings, social networking integration and other online strategies were downplayed in importance by respondents, suggesting that candidate websites that meet the most important desires of votersat least the voters in this pilot studyare completely accessible to even the most resource-strapped candidate campaign. Respondents cited obtaining information as their top priority Combining the frequency data, factor analysis, correlation tests and paired sample t-tests, the data from my sample strongly supports the working theory that obtaining information about the candidate was the primary goal of website visitors. The desire for issue information was most profound in the coin allocation activity; in fact, respondents considered issue information approximately twice as valuable as any other website component. After considering possible website components in the card-sorting and coin allocation activities, an overwhelming 93% of respondents 14 affirmed that obtaining information was their primary goal when visiting a candidate website. In fact, almost 67% of strongly agreed that obtaining information was their primary goal in visiting websites. The frequency data and statistical analyses were supported by respondents open-ended responses. When asked to give just three sentences of advice to prospective candidates creating websites, more than a third explicitly mentioned issue positionsno other response came close in terms of frequency. When asked in an open14

This question was only asked on the Spring 2008 survey instrument.

56

ended question about the single most important feature of a candidate website, more than 68% of respondents named issue statements. In contrast, the second-most prevalent responsedonation featureswas named by only 22% of respondents. Furthermore, the emphasis on videos, multimedia offerings and social networking capabilitiesall discussed with frequency and fervor in recent popular literature regarding candidates online campaigningwere downplayed by survey respondents. While 67% of respondents reported enjoying video on candidate websites, the desire for videos was not nearly as intense as that for informationonly 26% of respondents strongly agreed with the statement, I really like video on a candidate website. More importantly, four out of five respondents affirmed that they enjoyed multimedia offerings on candidate websites, but not at the expense of information. As noted in Chapter 5, the data was also revealing in what respondents did not ascribe with value. In the coin allocation exercise, social networking capabilities had the lowest mean score of all options (x = $0.02). In fact, some respondents actually advised candidates against including social networking. Clearly, the choice to include such networking or flashy features does not necessarily mean that substantive content suffers, but these were only tertiary concerns for survey respondents. Though the desire for more substantive information was clear, there is not sufficient data from this pilot study to guide candidates in exactly how to construct and present issue or biographical information in a manner that satiates the typical voters desire for said content. Only the limited responses admonishing candidates to avoid negativity in their online communications and pleading for honesty were offered as 57

guidance for candidates creating websites. It remains to be seen exactly what conditions must be met for a voter to consider their desire for information fulfilled, if the mere illusion of substantive information is sufficient, and if a website could ever have too much information. In addition, it is worth exploring whether the assumptions of traditional campaign opinion polling need to be modified for guiding online communications.

Respondents wanted to support candidates by donating money, volunteering, and attending events While candidate websites communicate with the entire population of voters, individuals who already support a candidate and individuals predisposed to supporting a candidacy are two very important groups within the full population. Respondents in this pilot study clearly wanted to be able to donate money, sign up to volunteer and learn about upcoming campaign events via candidate websites. The desire for these three supporting activities was consistent across the card-sorting exercise, coin allocation exercise and the open-ended questions. None of the respondents provided specific language in the open-ended responses regarding what sort of volunteer opportunity or event information might be most desirable; more research in this vein would certainly help candidates know the best ways to solicit commitments from voters on their websites. The ability to translate online support into offline action is becoming an increasingly essential component of 21st century campaigns. Further research examining

58

the best methods for translating this virtual enthusiasm into direct action, whether phone banking, canvassing, or e-mail blasts to supporters, is critical.

Respondents did not expect candidate websites to meet their every online need An analysis of the data from this pilot study suggests that social networking websites and advanced video components, though consumed by a large number of Internet users and the subject of much media coverage, would not necessarily be required for the creation an effective candidate website. This is perhaps even more surprising given the fact that half of the convenience sample was recruited via Facebook and political community websites. Indeed, it seems that as visitors are very comfortable moving between different websites, and do not expect their political candidates to serve as portals or final Internet destinations.

Effective websites can be created by candidates at all levels. Assuming that the trends in the data from this pilot study hold true for the full population, the lessons are profound. Perhaps most excitingly, the data suggests that effective websitesthat is, websites that satisfy the most important desires of the voting publicare completely within reach for candidates at all levels of office. Moreover, with new open-source content management system technologies that dramatically reduce barriers to creating websites, content can be created and edited by individuals with no web expertise whatsoever. This is not to say that the engagement of web

59

development professionals is diminished, but we are approaching the point at which there is no excuse for not creating an effective candidate website.

More public domain data is needed The goal in developing, deploying and analyzing the survey instrument for this thesis has been twofold: (1) to contribute to the body of knowledge regarding what voters want from candidate websites, and (2) to develop a pilot study to guide future research on the subject by researchers with the means to collect data from a sample more representative of the full voter population. Clearly, more research in this area is needed. A simple and straightforward instrument similar to the one used in this project, but with a random sample, would do much to advance the academic and professional communities interested in building and maintaining powerful and effective online political websites. With a clear outline of what the broad voting population wants from political candidate websites, additional research could commence examining the attitudes and behaviors of different voting subgroups, voters as they return to candidate websites for repeat visits, and many other avenues. It is no longer a question of if a candidate will create websites to further their political goals, but how. Understanding exactly what voters want from political candidate websites will undoubtedly lead to improved user experiences, increased political engagement and higher degrees of success for the candidates themselves.

60

APPENDICES Data Frequency Tables


Respondents have looked at candidate websites
(n=387) Response To learn about a candidate they did support To learn about a candidate they did not support To get information to help with decision about an upcoming election Have not looked at candidate websites Frequency 270 232 212 74 Percent 69.8 59.9 54.8 19.1

When visiting campaign websites, respondents.


(n=387) Response Read issue statements Read the candidates biography Just browsed Signed up for email updates Obtained information about campaign events Donated money Read the campaign blog Signed up to volunteer Requested a sign, bumper sticker, etc Told the candidate your opinion Met others Frequency 278 273 263 189 189 159 123 116 102 80 42 Percent 71.8 70.5 68.0 48.8 48.8 41.1 31.8 30.0 26.4 20.7 10.9

How often do you look for political information on the Internet?


(n=387) Response Every Day A few times a week Almost never A few times a year A few times a month Once a month Frequency 183 67 61 34 32 10 Percent 47.3 17.3 15.8 8.8 8.3 2.6

61

How respondents sorted cards with website components


Total Positive 373 361 297 312 329 296 317 303 314 289 254 250 221 219 211 176 106

Definitely Avoid

Definitely Include

Neutral

Include

Avoid

Candidate biography Learning about campaign events Issue statements Volunteering Learning simple ways to support candidate Donating money Reading or watching speeches Signing up for campaign emails News stories Voter registration info Learning how to support the candidate with friends Video of the candidate Candidate photos Learning why people like you support the candidate Connecting with other volunteers Merchandise Integration with social networking sites

2 3 2 1 1 3 0 1 2 5 2 1 2 0 0 2 6

1 1 4 4 7 15 2 14 10 28 5 6 6 16 10 28 25

11 22 75 69 47 73 68 68 61 64 125 130 158 150 165 181 104

93 92 65 121 175 100 166 134 173 109 147 153 128 135 142 125 63

280 269 232 191 154 196 151 169 141 180 107 97 93 84 69 51 43

4.67 4.61 4.38 4.29 4.23 4.22 4.20 4.18 4.14 4.12 3.91 3.88 3.79 3.75 3.70 3.50 3.46

Mean

Function

62

Total Negative 3 4 6 5 8 18 2 15 12 33 7 7 8 16 10 30 31

Respondents responses when asked to name the most important features of candidate websites
(Coded open-ended responses) Response Issue statements Donation features Candidates biographical information Volunteer opportunities Event information Video Ease of use/navigation News Information (general) Photos Email signup Connect with other supporters Speeches Aesthetics Soliciting feedback Interactivity Blog Social networking Voting information Contact information Offline action Merchandise Novelty/fun Frequency 264 85 74 74 38 21 17 16 15 15 13 12 12 9 8 7 4 4 4 3 2 1 1 Percent 68.2 22.0 19.1 19.1 9.8 5.4 4.4 4.1 3.9 3.9 3.4 3.1 3.1 2.3 2.1 1.8 1.0 1.0 1.0 0.8 0.5 0.3 0.3

63

How respondents allocated $1.00 for website components


Mean $0.00 $0.05 $0.10 $0.15 $0.20 $0.25 $0.30 $0.35 $0.40 $0.45 $0.50 $0.55 $0.60 $0.65 $0.70 $0.75 $0.80 $0.85 $0.90 $0.95 Features Issues Biography Donating money Volunteer opportunities Speeches Event information Simple ways to support candidate Voter registration News stories Email signup $1.00 1 1 1 -

0.20 0.11 0.09 0.07 0.07 0.06 0.06 0.06 0.06 0.05

28 54 97

38 104 108

48 96 71 83 75

78 61 59 47 33 23 30 22 26 19

59 37 26 20 12 4 6 15 8 8

52 16 12 6 3 2 8 1 1

30 10 5 1 1 2

18 5 1 -

16 1 1 1 -

5 1 2 -

7 5 1 1 -

2 2 1 -

2 -

1 -

2 -

105 126 114 146 98

145 115 86 51 74 57

114 151 153 136 137 139 146 154

64

How respondents allocated $1.00 for website components


Mean $0.00 $0.05 $0.10 $0.15 $0.20 $0.25 $0.30 $0.35 $0.40 $0.45 $0.50 $0.55 $0.60 $0.65 $0.70 $0.75 $0.80 $0.85 $0.90 $0.95 0 Features Videos Why other people support candidate Candidate photos Merchandise Connecting with other volunteers Ways to support candidate with friends Social networking $1.00 1 -

0.05 0.04 0.03 0.03 0.03

188 124 216 113 216 138 233 126 269 81

46 31 18 20 23

17 17 8 4 8

7 8 4 3 2

2 2 2 2

1 0

1 1

1 1 0

1 0

0.02

272

96

11

0.02

159

58

20

65

Obtaining information is my primary goal when visiting a campaign website.


Response Strongly Agree (5) Agree (4) Neither agree or disagree (3) Disagree (2) Strongly disagree (1) Frequency 165 60 8 5 3 N=241, Mean=4.57 Percent 68.5 24.9 3.3 2.1 1.2

I really like video on a campaign website.


Response Strongly Agree (5) Agree (4) Neither agree or disagree (3) Disagree (2) Strongly disagree (1) Frequency 63 98 49 26 5 N=241, Mean=3.78 Percent 26.1 40.7 20.3 10.8 2.1

I enjoy multimedia offerings on campaign websites, but not at the expense of information.
Response Strongly Agree (5) Agree (4) Neither agree or disagree (3) Disagree (2) Strongly disagree (1) Frequency 74 129 17 20 1 N=241, Mean=4.06 Percent 30.7 53.5 7.1 8.3 0.4

66

I prefer a simple site to a complex site.


Response Strongly Agree (5) Agree (4) Neither agree or disagree (3) Disagree (2) Strongly disagree (1) Frequency 77 88 34 37 5 N=241, Mean=3.81 Percent 32.0 36.5 14.1 15.4 2.1

I would not vote for a candidate who did not have a website.
Response Strongly Agree (5) Agree (4) Neither agree or disagree (3) Disagree (2) Strongly disagree (1) Frequency 37 47 37 80 40 N=241, Mean=2.84 Percent 15.4 19.5 15.4 33.2 16.6

Websites affect the way I vote.


Response Strongly Agree (5) Agree (4) Neither agree or disagree (3) Disagree (2) Strongly disagree (1) Frequency 16 49 66 82 28 N=241, Mean=2.76 Percent 6.6 20.3 27.4 34.0 11.6

67

I would design candidate websites differently.


Response Strongly Agree (5) Agree (4) Neither agree or disagree (3) Disagree (2) Strongly disagree (1) Frequency 14 42 129 54 2 N=241, Mean=3.05 Percent 5.8 17.4 53.5 22.4 0.8

68

Respondents Advice for Designing Candidate Websites


(Coded open-ended responses, n=387) Response Focus on issue positions. Make the site easy to use/navigate. Keep the site simple. Focus on donations. Focus on volunteer opportunities. Focus on information (in general). Use clear language. Update the site often. Be honest. Focus on aesthetics/design. Focus on the candidates biography. Focus on videos. Focus on collecting visitors email addresses. Avoid negativity. Include photos. Help visitors connect with other supporters. Use the site to humanize the candidate. Solicit feedback from visitors. Use the site to rise above sound bites and media cycle constraints Make the site interactive. Deemphasize donations. Make the site fun. Focus on the blog. Focus on campaign news. Include the candidates voting records. Focus on campaign event information and calendars. Consider the many kinds of visitors when designing the site. Include voter registration information. Make the accessible to all users. Frequency 140 103 103 64 56 54 43 34 33 30 29 29 25 20 18 15 15 12 11 10 10 10 8 7 7 6 6 6 5 Percent 36.2 % 26.6 % 26.6 % 16.5 % 14.5 % 14.0 % 11.1 % 8.8 % 8.5 % 7.8 % 7.5 % 7.5 % 6.5 % 5.2 % 4.7 % 3.9 % 3.9 % 3.1 % 2.8 % 2.6 % 2.6 % 2.6 % 2.1 % 1.8 % 1.8 % 1.6 % 1.6 % 1.6 % 1.3 %

69

Respondents Advice for Designing Candidate Websites


(Coded open-ended responses, n=387) Response Involve visitors in campaign. Allow users to purchase merchandise. Include speeches. Include contact information. Do not pander. Deemphasize video/flash. Personalize the experience for visitors. Integrate the site with social networking websites. Invest in staff to update and maintain the website. Promote the site. Use the site to rebut attacks from opponents. Value every visitor. Include a chat room or forum. Coordinate with other campaign functions. Include real-time counters. Include links to other sites. Deemphasize social networking. Frequency 5 5 5 4 4 4 4 4 4 3 3 3 2 2 2 2 2 Percent 1.3 % 1.3 % 1.3 % 1.0 % 1.0 % 1.0 % 1.0 % 1.0 % 1.0 % 0.8 % 0.8 % 0.8 % 0.5 % 0.5 % 0.5 % 0.5 % 0.5 %

70

Gender
Response Female Male No valid response Frequency 216 167 2 Percent 56.1 43.4 0.5

Race/Ethnicity
Response White Asian Black Latino Other No valid response Frequency 338 7 9 7 16 10 Percent 87.3 1.8 2.3 1.8 4.1 2.6

Education
Response I havent graduated from high school. Im a high school graduate Ive completed some post-high school education, but it wasnt at a college or university. Im completed some college. Im a college graduate. Ive completed some post-graduate work (towards an advanced professional degree, Masters Degree, etc.) Frequency 3 2 5 47 100 230 Percent 0.8 0.5 1.3 12.1 25.8 59.4

71

Voting Behavior
Response I generally vote for a Democratic candidate. I generally vote for a Republican candidate. I generally vote for a Green Party candidate. I generally vote for a Libertarian candidate. Im an Independent, and have a hard time placing myself on the partisan spectrum. Im an Independent, but would generally vote for a progressive or liberal candidate Im an Independent, but would generally vote for a conservative candidate. Frequency 137 32 1 2 14 24 20 Percent 59.6 13.9 0.4 0.9 6.1 10.4 8.7

Ideology
Response Very Conservative (5) Conservative (4) Neither liberal nor conservative (3) Liberal (2) Very Liberal (1) Frequency 9 55 68 164 84 N=380, Mean=2.32 Percent 2.4 14.5 17.9 43.2 22.1

Partisan Affiliations
Response Democratic Republican Independent Green Libertarian Progressive Frequency 140 39 57 1 2 2 N=380, Mean=2.32 Percent 53.6 14.9 21.8 0.4 0.8 0.8

72

Folded Ideology
Response Strong Ideological Preference Weak Ideological Preference No Ideological Preference Frequency 93 219 68 N=380 Percent 17.6 56.6 0.18

73

Survey Instrument

Have you ever looked at a political candidate website to...? (Select all that apply.) o obtain information about a candidate you already supported? o learn about a candidate you did not support? o get information to help you make a decision about an upcoming election? o I havent really looked at a political candidate website. When visiting a political candidate website, did you ever....? (Select all that apply.) o sign up for campaign emails o donate money o request a bumper sticker, yard sign or other items to show your support o tell a candidate or campaign your thoughts on an issue o just browse o read how a candidate feels about specific issues o read a candidates biography o volunteer o get information about campaign events o get information about campaign events o watch videos o look at pictures o read How often do you look for political information on the Internet? o every day o a few times a week o a few times a month o once a month o a few times a year o almost never Thinking back, what candidate or campaign websites (if any) have really made an impression on you? How so? o open response How important are the following opportunities for you in a campaign website? o signing up for campaign emails o making monetary donations o volunteering for the campaign 74

o o o o o o o o o o o o o o o o o o o o o

watching videos of the candidate registering to vote learning about the candidates background learning how a candidate feels about issues learning about campaign events learning how to support the candidate with your friends seeing photos of the candidate connecting with other volunteers learning why people like you are supporting the candidate reading or watching candidate speeches reading news stories related to campaign purchasing merchandise (e.g., t-shirts, buttons, bumper stickers) learning simple ways to support the candidate integration with social network sites (e.g., Facebook, Twitter, MySpace) (open response) Very important Important Dont really care one way or the other Should probably be avoided Should definitely be avoided Not sure what that means

Heres a different way of thinking about a previous question: You have one dollar to spend -- how much of that dollar would you spend on the following features? o signing up for campaign emails o donate money o volunteering for the campaign o watching videos of the candidate o registering to vote o learning about the candidates background o learning how a candidate feels about issues o learning about campaign events o learning how to support the candidate with your friends o seeing photos of the candidate o connecting with other volunteers o learning why people like you are supporting the candidate o reading or watching candidate speeches o reading news stories related to campaign o purchasing merchandise (e.g., t-shirts, buttons, bumper stickers) o learning simple ways to support the candidate

75

Can you think of any candidate or campaign websites that really bothered you? What was it about those sites that you did not like? o open response Generally speaking, what would you consider to be the most important parts of a campaign website? o (open response) Do you agree or disagree with the following statements? o Obtaining information is my primary goal when visiting a campaign website o I really like video on a campaign website o I would not vote for a candidate who did not have a website o Websites affect the way I vote o I would prefer a simple site to a complex site o I enjoy multimedia offerings on campaign websites but not at the expense of information. o If it were up to me, I would design campaign websites differently. o Strongly agree o Agree o Not sure o Disagree o Strongly disagree If you were asked to consult a candidate create a new website, but could only offer up to 3 sentences of advice...what would you say? o (open response) Now its time for the fun demographic questions. For starters: o In what year were you born? o What is your ZIP code? o What is your gender? o How would you describe your race or ethnicity? Which of these options best describes your level of education? o I havent graduated from high school. o Im a high school graduate o Ive completed some post-high school education, but it wasnt at a college or university. o Im completed some college. o Im a college graduate. o Ive completed some post-graduate work (towards an advanced professional degree, Masters Degree, etc.) 76

Generally speaking, which of these choices best describes your voting behavior? o I generally vote for a Democratic candidate. o I generally vote for a Republican candidate. o I generally vote for a Green Party candidate. o I generally vote for a Libertarian candidate. o Im an Independent, and have a hard time placing myself on the partisan spectrum. o Im an Independent, but would generally vote for a progressive or liberal candidate. o Im an Independent, but would generally vote for a conservative candidate. o Not sure Generally speaking, how would you describe your political views? o Very liberal o Liberal o Neither liberal or conservative o Conservative o Very Conservative o Im not sure How would you describe your party affiliation? o Democratic o Republican o Green Party o Libertarian o Independent o Other

77

The MindCanvas Interactive Survey Tool


The following screenshots provide examples of the MindCanvas survey user interface.

This is a screenshot of the card-sorting exercise for which respondents were asked to drag cards with text of possible website components into suitable groups.

78

This is a screenshot of the coin allocation exercise for which respondents were asked to drag coins onto boxes containing text descriptions of website components according to how the respondent viewed the components relative importance.

79

BIBLIOGRAPHY
2008. 10 Principles Of Effective Web Design. Smashing Magazine. http://www.smashingmagazine.com/2008/01/31/10-principles-of-effective-webdesign/. Armstrong, Jerome, and Markos Moulitsas . 2006. Crashing the Gate: Netroots, Grassroots, and the Rise of People-Powered Politics . White River Junction, Vermont: Chelsea Green Publishing Company. Bassik, Michael. 2008. Online Advertising Strategy for Advocacy Campaigns. http://www.americanprogressaction.org/events/IAR.html. Benoit, Pamela, and William Benoit. 2005. Criteria for evaluating political campaign webpages. Southern Communication Journal 70, no. 3:230-247. Bimber, Bruce, and Richard Davis. 2003. Campaigning Online: The Internet in U.S. Elections. New York: Oxford University Press. Blanchfield, Theodora. 2007. The Online Race. Campaigns & Elections, November. Campaigns & Elections. 2007. And the Dot Goes To... Campaigns & Elections, November. Cohen, Diana Tracy. 2005. My Page is Bigger than your Page: A Comparative Examination of E-Campaign Functionality and Strategy of 2004 Senate Candidate and State Party Sites. Annual Meeting of the American Political Science Association, September 1-4, 2005. CompleteCampaigns.com. 2008. Affordable Political Campaign Websites. CompleteCampaigns.com. http://www.completecampaigns.com/sitebuilder.asp?gclid=CMvdub7w2JECFQ MKPAod-Vxxbw (Accessed February 22, 2008). Cornfield, Michael. 2004. Bush vs. Kerry on the Web. Campaigns & Elections http://findarticles.com/p/articles/mi_m2519/is_5_25/ai_n6078654/pg_1 (Accessed February 16, 2008). Cornfield, Michael, and Lee Rainie. 2006. The Web Era Isn't as New as You Think. The Washington Post, November 5.

80

D'Alessio, Dave. 2000. Adoption of the World Wide Web by American Political Candidates, 1996-1998. Journal of Broadcasting & Electronic Media 44, no. 4:556-569. Daley Print & Web Solutions. 2008. 5 Critical Pieces of Text Needed in a Campaign Website. OnlineCandidate.com. http://onlinecandidate.com/articles/5-criticalpieces-of-text-needed-in-a-campaign-website (Accessed February 22, 2008). Dalton, Aaron. 2007. The Digital Road to the White House. PC Magazine, April 24. Davis, Steve. 2005. Presidential Campaigns Fine-tune Online Strategies. Journalism Studies 6, no. 2:241-244. Delaney, Colin. 2008. 10 (+1) Ways to Build Traffic to a Website. e.politics: online advocacy tools & tactics. http://www.epolitics.com/2008/04/15/10-1-ways-tobuild-traffic-to-a-website/ (Accessed April 19, 2008). Doshi, Ravi. 1916. Presidential candidates log on to check up with younger voters. Daily Bruin, July 12 http://www.dailybruin.ucla.edu/news/2007/dec/06/presidential-candidates-logcheck-younger-voters/ (Accessed February 17, 2008). Druckman, James, Martin Kifer, and Michael Parkin. 2007. The Technological Development of Congressional Candidate Web Sites: How and Why Candidates Use Web Innovations. Social Science Computer Review 25, no. 4:425-442. Endres, Danielle, and Barbara Warnick. 2004. Text-based Interactivity in Candidate Campaign Web Sites: A Case Study from the 2002 Elections. Western Journal of Communication 68, no. 3:322-342. Finkelpearl, Susan. 2008. Twelve Tips for Sprucing Up Your Website. Free Range Studios. http://www.freerangestudios.com/whitepapers/ws_simple_tips.pdf (Access April 12, 2008). Foot, Kristen, and Steven Schneider. 2006. Web Campaigning. Cambridge, Massachusetts: The MIT Press. Gibson, Rachel. 2004. Web Campaigning from a Global Perspective. Asia-Pacific Review 11, no. 1:95-126. Gibson, Rachel, Wainer Lusoli, and Stephen Ward. 2005. Online Participation in the UK: Testing a Contextualised Model of Internet Effects. British Journal of Politics & International Relations 7, no. 4:561-583. 81

Gibson, Rachel, and Stephen Ward. 2000. A Proposed Methodology for Studying the Function and Effectiveness of Party and Candidate Web Sites. Social Science Computer Review 18, no. 3:301-319. Gillmor, Dan. 2007. Some candidates cope better than others with new campaign reality of the Web. Washington Examiner, February 13 http://www.examiner.com/printa562400~Dan_Gillmor:_Some_candidates_cope_better_than_others_with_new_c ampaign_reality_of_the_Web.html (Accessed February 18, 2008). Gleason Blue, Scott, and Jackie Lewis. 2006. Survey says: Campaigns succeed on the Internet. E-Voter Institute. http://evoterinstitute.com/pdfs/voterimpact.pdf (Accessed February 24, 2008). Graff, Garrett. 2007. The First Campaign: Globalization, the Web, and the Race for the White House. New York: Farrar, Strauss and Giroux. Grafton, Carl. 2007. Book Review: Internet Politics. Social Science Computer Review 25:129. Gulati, Girish, and Christine Williams. 2007. Closing the Gap, Raising the Bar: Candidate Web Site Communication in the 2006 Campaigns for Congress. Social Science Computer Review 25, no. 4:443-465. Haridakis, Paul, and Evonne Whitmore. 2006. Understanding Electronic Media Audiences: The Pioneering Research of Alan M. Rubin. Journal of Broadcasting & Electronic Media 50, no. 4:776-774. Hopkins, Mark. 2007. Social Media and American Presidential Candidates. Mashable. http://mashable.com/2007/12/05/social-media-and-american-presidentialcandidates/ (Accessed February 17, 2008). Howard, Philip. 2006. New Media Campaigns and the Managed Citizen. New York: Cambridge University Press. Ireland, Emilienne, and Phil Tajitsu Nash. 2001. Winning Campaigns Online: Strategies for Candidates and Causes. 2nd ed. Bethesda, Maryland: Science Writers Press. Jackson, Nigel. 2007. Political parties, the Internet and the 2005 General Election: third time lucky? Internet Research 17, no. 3:249-271.

82

Kaye, Barbara, and Thomas Johnson. 2002. Online and in the Know: Uses and Gratifications of the Web for Political Information. Journal of Broadcasting & Electronic Media 46, no. 1:54-72. Kippen, Grant, and Gordon Jenkins. 2004. The Challenge for E-Democracy for Political Parties. In Democracy Online: The Prospects for Political Renewal Through the Internet, Ed. Peter Shane, 253-265, New York: Routledge. Klotz, Robert. 2007. Internet Campaigning for Grassroots and Astroturf Support. Social Science Computer Review 25, no. 1:3-12. Klotz, Robert. 2004. The Politics of Internet Communication. New York: Rowman & Littlefield Publishers, Inc. Kohut, Andrew. 2008. The Internet Gains in Politics. Pew Internet & American Life Project http://www.pewinternet.org/PPF/r/234/report_display.asp (Accessed February 23, 2008). Krueger, Brian. 2006. A Comparison of Conventional and Internet Political Mobilization. American Politics Research 34, no. 6:759-776. Leadbeater, Charles. 2008. We Think. http://www.youtube.com/watch?v=qiP79vYsfbo (Accessed March 1, 2008). McGill, Lawrence, Andras Szanto, and Marianne Johnston. 1997. The Voters Speak. First Amendment Center. http://web.archive.org/web/20010222233723/http://www.fac.org/publicat/campa ign/lth96votespk.htm (Accessed February 18, 2008). McGregor, Michael. 2006. When the "Public Interest" is Not What Interests the Public. Communication Law & Policy 11, no. 2:207-224. Myrna the Minx . 2008. Two Best Campaign Websites: Huckabee and Obama | Reno and Its Discontents. Reno and Its Discontents. http://www.renodiscontent.com/2008/02/18/two-best-campaign-websiteshuckabee-and-obama/ (Accessed February 22, 2008). Nickerson, David. 2007. The Ineffectiveness of E-Vites to Democracy: Field Experiments Testing the Role of E-Mail on Voter Turnout. Social Science Computer Review 25, no. 4:494-503. Owen, Diana. 2004. Internet use and the 2000 presidential election. Electoral Studies 23, no. 3:415-429. 83

Owen, Diana, Richard Davis, and Vincent James Strickler. 1999. Congress and the Internet. Harvard International Journal of Press/Politics 4, no. 2:10-29. Palser, Barb. 2007. Politics 2.008: How will the Internet influence the presidential election? Entrepreneur.com. http://www.entrepreneur.com/tradejournals/article/165807335.html (Accessed February 16, 2008). Panagopoulos, Costas. 2007. Technology and the Transformation of Political Campaign Communications. Social Science Computer Review 25, no. 4:423-424. Panagopoulos, Costas, and Daniel Bergan. 2007. Online Fund-Raising and Contributors in the 2004 Presidential Campaign. Social Science Computer Review 25, no. 4:484-493. Papacharissi, Zizi, and Alan Rubin. 2000. Predictors of Internet Use. Journal of Broadcasting & Electronic Media 44, no. 2:175-197. PoliticalWebNames.com. 2007. Political Web Names - We have the domain name for your campaign. NightOwlPublishing.com . http://www.hicksforsenate.com/website_basics.php (Accessed February 22, 2008). Rackaway, Chapman. 2007. Trickle-Down Technology? The Use of Computing and Network Technology in State Legislative Campaigns. Social Science Computer Review 25, no. 4:466-483. Reich, Brian, and Dan Solomon. 2007. 2006: Finally, Campaigns Plug In -- For Real. Campaigns & Elections, February. Ruggiero, Thomas. 2000. Uses and Gratifications Theory in the 21st Century. Mass Communication & Society 3, no. 1:3-37. Schossow, Clay. 2008a. Building Successful Political Campaign Websites - New Media Campaigns. New Media Technology Corporation. http://www.newmediacampaigns.com/content/page/title/Building_Successful_P olitical_Campaign_Websites (Accessed February 22, 2008). Schossow, Clay. 2008b. Campaigns in a New Media World - New Media Campaigns. New Media Technology Corporation. http://www.newmediacampaigns.com/content/page/title/Campaigns_in_a_New_ Media_World (Accessed February 22, 2008). 84

Schossow, Clay. 2008c. It's All About Functionality - New Media Campaigns. New Media Technology Corporation. http://www.newmediacampaigns.com/content/page/title/It's_All_About_Functio nality (Accessed February 22, 2008). Smashing Magazine. 2008. 10 Principles of Effective Web Design. http://www.smashingmagazine.com/2008/01/31/10-principles-of-effective-webdesign/ (Accessed March 29, 2008). Stirland, Sarah Lai. 2008. The Tech of Obamamania: Online Phone Banks, Mass Texting and Blogs. Wired. http://www.wired.com/politics/law/news/2008/02/potomac_primaries. Stromer-Galley, Jennifer. 2000. On-Line Interaction and Why Candidates Avoid It. Journal of Communications 50, no. 4:111-132. Sutton, John. 2007. Strategy Tips for Presidential Campaign Websites. Navigation Arts. http://www.navigationarts.com/insight/campaign-sites/ (Accessed April 19, 2008). Teachout, Zephyr, and Steeter. 2007. Mousepads, Shoe Leather, and Hope: Lessons from the Howard Dean Campaign for the Future of Internet Politics. Paradigm Publishers . Teeling, Erin. 2006. The Internets Role in Political Campaigns The Bivings Report. The Bivings Report. http://www.bivingsreport.com/2006/the-internets-role-inpolitical-campaigns/ (Accessed February 22, 2008). Tisinger, Russell, Natalie Stroud, Kimberly Meltzer, Brett Mueller, and Rachel Gans. 2005. Creating Political Websites: Balancing Complexity & Usability. Knowledge, Technology & Policy 18, no. 2:41-51. Trammell, Kaye. 2006. Blog offensive: An exploratory analysis of attacks published on campaign blog posts from a political public relations perspective. Public Relations Review 32, no. 4:402-406. Trammell, Kaye. 2007. Candidate Campaign Blogs: Directly Reaching Out to the Youth Vote. American Behavioral Scientist 50, no. 9:1255-1263. Trammell, Kaye, Andrew Paul Williams, Monica Postelnicu, and Kristen Landreville. 2006. Evolution of Online Campaigning: Increasing Interactivity in Candidate

85

Web Sites and Blogs Through Text and Technical Features. Mass Communication & Society 9, no. 1:21-44. Turk, Michael. 2008. The (Mostly) New BarackObama.com. techPresident. http://www.techpresident.com/blog/entry/20660/the_mostly_new_barackobama _com (Accessed January 29, 2008). Vargas, Jose Antonio. 2008. The Trail: Online, Obama Starts to Beat Ron Paul . Washingtonpost.com, January 17 http://blog.washingtonpost.com/thetrail/2008/01/17/online_obama_starts_to_beat_ro_1.html (Accessed February 16, 2008). Ward, Stephen, and Rachel Gibson. 2003. On-line and on message? Candidate websites in the 2001 General Election. British Journal of Politics & International Relations 5, no. 2:188-205. Ward, Stephen, Rachel Gibson, and Wainer Lusoli. 2003. Online Participation and Mobilisation in Britain: Hype, Hope and Reality. Parliamentary Affairs 56, no. 4:654-668. Ward, Stephen, Wainer Lusoli, and Rachel Gibson. 2002. Virtually participating: A survey of online party members. Information Polity: The International Journal of Government & Democracy in the Information Age 7, no. 4:199-215. Wheaton, Ken. 2008. How the web will-or won't-change elections. Advertising Age, February 4. Williams, Andrew Paul, and Kaye Trammell. 2005. Candidate Campaign E-Mail Messages in the Presidential Election. American Behavioral Scientist 49, no. 4:560-574. Williams, Andrew Paul, Kaye Trammell, Monica Postelnicu, Kristen Landreville, and Justin Martin. 2005. Blogging and Hyperlinking; use of the Web to enhance viability during the 2004 campaign. Journalism Studies 6, no. 2:177-186. Williams, Christine. 2003. An Assessment of Candidate Web Site Performance and Effectiveness . 2003 Annual Meeting of the Midwest Political Science Association, Chicago, Illinois (Accessed February 17, 2008). Zeigler, Tidd. 2006. How to Build a Better Political Campaign Website The Bivings Report. The Bivings Report. http://www.bivingsreport.com/2006/howto-build-a-better-political-campaign-website/ (Accessed February 22, 2008).

86

Zeigler, Todd. 2006. How to Build a Better Political Campaign Website. The Bivings Report. http://www.bivingsreport.com/2006/how-to-build-a-better-politicalcampaign-website/ (Accessed April 19, 2006).

87

Вам также может понравиться