1 THE EFFECTS OF SOURCE DOCUMENTS ON RECALL AND CREDIBILITY ON NEWS By MEGAN DUNCAN A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER O F ARTS IN MASS COMMUNICATION UNIVERSITY OF FLORIDA 2012
2 2012 Megan Duncan
3 To fun and adventure
4 ACKNOWLEDGMENTS Gratitude is due to my committee members Dr. Wayne Wanta and Dr. Amy Zerba for the time and care they put into helping me throu gh this process. Dr. Cory Armstrong, my chair, spent weekends reading my drafts, office hours explaining the notes, and even more hours lending her expertise. It is because of her that this is finished. I thank her. And, I give a tip of the hat to my paren ts, who never stop believing in me.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 10 Recent Technology Changes to Newspaper Websites ................................ ........... 10 Past Newspaper Technology Changes ................................ ................................ ... 11 Purpose of Study ................................ ................................ ................................ .... 12 Implications of Research ................................ ................................ ......................... 13 2 LITERATURE REVIEW ................................ ................................ .......................... 16 Defining the Role of Sources in Journali sm ................................ ............................ 16 Role of Journalism in Interpreting Information ................................ ........................ 18 Inverted pyramid ................................ ................................ ................................ ..... 20 Narrative ................................ ................................ ................................ ................. 21 Theoretical Framework ................................ ................................ ........................... 22 Factors Involved in Recall ................................ ................................ ....................... 25 Factors Involved in Credibility of News ................................ ................................ ... 27 3 METHOD ................................ ................................ ................................ ................ 31 Experimental Design ................................ ................................ ............................... 31 The News Story and the Source Document ................................ ............................ 32 Participants ................................ ................................ ................................ ............. 34 Procedure ................................ ................................ ................................ ............... 34 Measures ................................ ................................ ................................ ................ 36 4 RESULTS ................................ ................................ ................................ ............... 38 Open ended Summaries ................................ ................................ ......................... 38 Variable Construction ................................ ................................ .............................. 40 Time ................................ ................................ ................................ ................. 40 News interest ................................ ................................ ................................ .... 40 Story interest ................................ ................................ ................................ .... 41 Recall Questions ................................ ................................ ................................ ..... 41
6 Factors Influencing Recall ................................ ................................ ....................... 43 Credibility ................................ ................................ ................................ ................ 44 Factors Influencing Credibility ................................ ................................ ................. 45 5 DISCUSSION ................................ ................................ ................................ ......... 61 Recall ................................ ................................ ................................ ...................... 62 Credibility ................................ ................................ ................................ ................ 68 Study Limitations ................................ ................................ ................................ .... 71 Futu re Research ................................ ................................ ................................ ..... 72 Conclusion ................................ ................................ ................................ .............. 74 APPENDIX A EXPERIMENT GRAPHICS ................................ ................................ ..................... 78 B POST TEST QUESTIONS ................................ ................................ ...................... 82 C INFORMED CONSENT ................................ ................................ .......................... 85 LIST OF REFERENCES ................................ ................................ ............................... 87 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 93
7 LIST OF TABLES Table page 3 1 Items included in the News Interest variable from all participants ...................... 48 3 2 Items included in the Story Interest variable for all participants .......................... 49 3 3 Independent Sample T test of Proportion of Recall Questions Correct by C ondition ................................ ................................ ................................ ............ 50 3 4 Independent T test of proportion of recall questions correct with those responses that answered all incorrectly removed ................................ ............... 51 3 5 Independent sample T test of specific recall questions by condition .................. 52 3 6 Multivariate analysis of variance examining proportion of recall questions correctly by influencing fa ctor ................................ ................................ ............. 53 3 7 Items included in credibility variable for all participants ................................ ...... 56 3 8 Independent sample T test of perception of credib ility by condition ................... 57 3 9 Significance of individual credibility questions ................................ .................... 58 3 10 Multivariate analysis of variance examining p erception of credibility by influencing factor ................................ ................................ ................................ 59
8 LIST OF FIGURES Figure page 2 1 Proposed model of influences on recall and credibility ................................ ....... 30 3 1 Histogram of frequencies of combined time participants spent reading article and source document ................................ ................................ ......................... 47 3 2 Mean proportion of recall q uestions answered correctly by condition influenced by story interest ................................ ................................ ................. 55 5 1 Factors influencing recall and credibility in this experiment ................................ 77
9 Abstract o f Th esis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Master of Arts in Mass Communication s THE EFFECTS OF SOURCE DOCUMENTS ON RECALL AND CREDIBILITY ON NEWS By Megan Duncan May 2012 Chair: Cory Armstrong Major: Journalism News organizations are using portable document formats of source documents such as criminal complaints, grand jury reports, and vital statistic forms to accompany news stories about those documents on the ir websites. Employing the heuristic systematic processing model, this study examined how those documents affected that included 158 university student participants fou nd little effect on recall or the perception of credibility. However, several factors influencing recall and perception of credibility were found. Implications and future research are discussed.
10 CHAPTER 1 INTRODUCTION Recent T echnology C hanges to N ewspap er W ebsites Web 2.0 has enhanced the newspaper reading experience with videos, soundclips, extended photograph galleries and interactive elements like opinion polls. Newspapers are also using the technology to post original source documents along with the newspaper stories. The BBC changed its Web policy in 2010 to require at least one external link per story posted at its online site (Stray, 2010). This was part of an effort to increase transparency, to let the reader see what information the reporter drew from (Stray, 2010). The website editor specifically mentioned that the reader should have access to the source document whenever possible (Stray, 2010). BBC News is not the first organization to push for providing an electronic copy of source documents t o readers. Local papers like The Arctic Sounder in Alaska provided a copy of the criminal complaint with a story about the arrest of an accused bootlegger (DeMarban, 2011). The Pittsburgh Post Gazette provided copies of court documents online after those d ocuments were used to declare a mistrial in a public corruption case (Majors & McNulty, 2011). One of the largest online databases for source documents, DocumentCloud, saw 10 times its regular Internet traffic when newspapers linked to an electronic copy o form birth certifica te in May 2011 (Hickman, 2011). More recently, the Jerry Sandusky grand jury report that spurred scandal at Penn State University infiltrated the Internet. The 23 page document was available from newspape rs like USA Today and The Wall Street Journal (USA Today, 2011; Miller, Everson & Searcey, 2011 ). The Wall Street Journal provided a link to the report and
11 other supplemental material such as a statement from the state a ttorney general on the same web page as a story about the scandal (Miller, Everson & Searcey, 2011) The Washington Post paired the electronic version of the court document with a n editorial column about the allegations (The Washington Post, 2011 ), which allowed the reader to see what the jour nalist had to say about the report and then compare it with the actual words of the report. Traditionally, citizens would have to search on a computer or at a government office to read the original documents journalists use as sources. In the new era, howe ver, the source documents are provided online by news organizations Though this practice has not been uniformly adopted or studied by researchers, it is anecdotally becoming more common. Past Newspaper Technology Changes Research on past advancements of i nformation presentation by newspapers may help explain this phenomenon. Like the 1982 change to a graphic look at the news with the advent of USA Today newspapers today may be increasing recall of information when they present it in a form other than the traditional inverted pyramid format. Source documents, such as a criminal complaint, are written in narrative format (Biggs, 2004). Traditional news stories are written in inverted pyramid format (Rich, 2010). Reading information about the Iraq election in narrative form led to stronger comprehension than reading it in inverted pyramid form among younger readers (Zerba, 2008). of recall in newspapers looked at the impact of des ign elements like pull quotes and info graphics (Wanta & Gao, 1994; Wanta & Remy 1995). These studies found younger readers could remember specific information better when it was included in pull quotes,
12 graphs, and charts. Other newspaper recall studies s howed readers could correctly answer more questions about content when design elements displayed information than when it was included only in a traditional news story (Griffin & Stevenson, 1992; Smith & Hajash, 1998). Displaying source documents in an ele ctronic format may have the same effects as these early studies, carried out when newspaper graphics were still mostly text (Smith & Hajash, 1998). Purpose of Study The purpose of this study is to examine the effects on the reader of posting electronic co pies of source documents alongside news stories on the websites of and ability to recall information when an electronic copy of an original source document accompanies a traditional news story. First, it review s the research about the roles sources play in journalism as well as the role of the journalist to interpret information. Then, it examine s the current research on factors involved in credibility and recall in print and online journalism. To show the effects of reading a source document, half of the participants in an experiment read the traditional news story and the source document. The other half of participants will read only the traditional news story. Then all participants answer questions about recall and perception of credibility of the news story. The answers are compared between the two groups and analyzed for the effects of reading the original source document when consuming news. The findings and the dire ction of future research are discussed.
13 Implications of Research Motives and effects of posting source documents to accompany news stories on the websites of news organizations has not been examined in research. However, research on past newspaper techniqu es supports that this new practice will improve credibility and recall of the news content because readers have access to multiple story forms, repeated information, and cues to indicate the validity of the information. Sources, named and anonymous, are wh at fuel journalism. While people are the vast majority of sources in newspapers (Kasoma & Maier, 2005), more than a third of newspaper stories use non human sources like government documents or information from the Internet, a content analysis showed (Kaso ma & Maier, 2005). Documents might be the basis of an investigative piece (Berry, 2009), or they might provide the details for a police or crime story (Chermak, 1995). posting source documents to a news website (Stray, 2010). When posting source documents to aid in transparency, newspapers may also be adding to their credibility. Credibility studies are categorized into medium credibility (Kiousis, 2001) and publication credibi lty (Armstrong & Collins, 2009). If a particular message is believed as true and valid, that particular message is credible (Wilson & Sherrell, 1993). Anonymous sources have been shown to hurt credibility (Sternadori & Thorson, 2001), but making source doc uments available for examination may increase credibility. The extent that a reader pays attention to the source of the information rather than the content of the information is explained by the heuristic systematic information processing model. Systemati c thinking is defined as careful scrutiny of the content of the message (Trumbo, 1999). Heuristic thinking is defined as using cues to come to a
14 quick decision about the message (Trumbo, 1999). Systematic thinking is predicted by high involvement in the me ssage, while low involvement predicts heuristic thinking (Trumbo, 1999). When this theoretical framework is applied to this study, it predicts that those who are not interested in the news story will use the source of the information to redibility. Those same people who are not systematically evaluating the content of the message will resort to peripheral cues to judge the credibility of the message. The news story example used in this research carries the generic by line The source document is labeled with the name of federal agencies and stamped by a federal magistrate among other identifying characteristics. The results of this research could be helpful to news organizations that maintain websites and consider posting so urce documents to accompany news stories. Providing original source documents is a convenient and new tool to reinforce the information provided in news stories. If recall is increased by reading source documents, media could institutionalize the practice to help the reader further understand the subject of the report. If credibility is increased by reading source documents, newspapers could set company policies to demonstrate to readers the veracity of the information in a news story. Improved recall and c redibility can enhance the ability of journalism to act as a watchdog. Providing source documents may be one way newspapers can take advantage of emerging technology to deliver the news and the documents that create the news. While this experiment will employ only one particular type of source document a criminal complaint it could spark interest in the effects of other types of source documents, like various court records, vital statics forms, or government agency budgets.
15 Additionally, this resear ch could benefit the scholarship on the effects of source documents on newspaper readers. Because providing source documents to the reader was not as easy or convenient before recent technology advances, no known research has looked at the effects of sourc e documents on the reader. If recall and credibility are increased when readers have access to source documents, this study could expand the knowledge of those variables.
16 CHAPTER 2 L ITERATURE REVIEW Defining the Role of Sources in Journalism Sources driv e journalism. News sources are central to the study of journalism because sources give voice to the flow of ideas and information that preserve capitalism and democracy (Manning, 2001). Sources are central to story selection and also influence story produ ction (Gans, 1979). Human sources are often the starting point for a story because they make the story credible and readable (Rich, 2010). The individual journalist has the most power in deciding who to use as a source, though the news organization may exe rt some pressure in source selection (Powers & Fico, 1994). These people might be public relations professionals, leaders of organizations, or eye witnesses (Rich, 2010). Some journalists use the same source repeatedly (Gans, 1979). A s a journalist and a source develop a relationship t he source will use information to get access to the journalist and the power of the press (Gans, 1979). As the relationship develops, t he source will better understand what the journalist is looking for and how to fashion i nformation in a way that is more likely to make it into the news (Manning, 2001). The journalist will start to lean on reliable sources because of the pressures of deadline (Manning, 2001). I nterviews with Israeli journalists show journalists rate 86 per cent of the sources used in publication as either credible or highly credible (Reich, 2011). On the opposite end of the spectrum, the journalists ranked 2 percent of sources used on a daily basis as not very credible or not at all credible (Reich, 2011).
17 P eople are the primary source of information for news stories, accounting for 90 percent of named news sources (Kasoma & Maier, 2005). A content analysis of 900 articles in nine daily U.S. newspapers found the average number of sources per story was 4.1, an d the average number of human sources was 3.3 (Kasoma & Maier, 2005). Thirty five percent of news stories used non human sources like government documents or the Internet (Kasoma & Maier, 2005). Stories that rely more on documents and public records can t ake longer to investigate and interpret than the stories based on human sources. What ended in a Pulitzer Prize winning series on atrocities committed by an Army unit known as Tiger Force began with five weeks of combing through documents hoping that there was an untold story in the pages (Berry, 2009). Documents, indeed, can be a powerful basis for a story. When journalists uncovered the 2002 Olympic bribery scandal at Salt Lake City, two documents a letter and an IRS form fueled the fire (Campbell, 20 07). Because records and documents can add credibility to a story, journalism students are taught to ask for supporting documents during an interview, and to use courthouses and other government offices as valuable resources (Nash, 1984). In newspaper crim e stories, 4.1 percent of sources are police and court documents (Chermak, 1995). When writing about crime, several police documents can help a reporter including a blotter, incident report, and booking log (Police Sources, 2008). A common source for writi ng about criminal charges is the criminal complaint. Federal Rules of Criminal Procedure define the complaint as a written statement detailing the essential facts in the offenses charged and stipulates that the statements must be sworn to under oath (Rule 3). Though the process varies from state to state, the state courts use much of the
18 the U.S. Constitution. The availability of those records also varies by state (Police So urces, 2008). Because of the variation between states this research will focus on a federal criminal complaint. Role of Journalism in Interpreting Information When online media organization Wikileaks recieved thousands of politically charged internationa l cables in 2010 site founder Julian Assange did not post all the raw, unorganized data online leadership partnered with traditional newspapers in the United Kingdom and United States to allow these profes sionals access to the information. Only after journalists analyzed interpreted and redacted specific cables were those messages published to messages, the data would h ave attracted only those with the time and intense curiosity to sift through information looking for the intriguing bits (Keenan, 2010). The public has shown it loves to comment on what ot hers have unearthed, connected or investigated. But, someone has to do the unearthing, connecting and investigating and often the public leaves that work to the media personnel (Goldsborough, 2010). One of the primary and original roles of the media has been as a government watchdog. The media have the obligation to play ombudsman for the taxpayers, and occasionally is pressed into blowing the whistle on corruption. Sometimes this watchdog role means reporting on a school board budget meeting, and other times it involves examining documents at the state Capitol, trying to interpret numbers and make
19 Both journalists and readers rate interpreting information high on the list of roles the media plays in societ y and democracy (Burgoon, Bernstein & Burgoon, 1983; Burgoon, & Burgoon, 1981; Chung, 2009). In a content analysis of 10 national news organizations, Tremayne (2004) found that a framework for the facts was needed to make the journalism meaningful to the r eader. A content analysis of Web news stories found stories about international issues contained more hyperlinks than stories about local issues (Tremayne, 2004). The author surmised this was because more background information was needed to put internatio nal stories in context than was needed for local stories. Reporters put facts in context so readers can understand the stories and come to their own conclusions. The context and background reporters provide help preserve democracy, he concluded. So vital is the interpretive aspect of journalism that Zelizer (1993) argued from an anthropological perspective that journalism should not be called a profession, but rather an interpretive community. This community, the author said, is united by its shared percep tions of reality. Using both the Watergate stories and the reporting on McCarthyism, the author explained the interpretive process happens twice for journalists: first as they are experiencing the event and second as they write about it. The results of thi s double interpretation become essential about the event (Zelizer, 1993). to put the news into clear and understandable language. Burgoon, B urgoon and Wilkinson (1981) found that fewer words per sentence and a lower reading level of the
20 text both increased trust of several news stories among a group of undergraduate students. Inverted pyramid is not always repl icated exactly to readers. The Myers & Sternadori, 2009). Then, the typical perso n will move onto supplemental materials like videos or sound clips. Wise, Bolls, Myers and Sternadori (2009) found that changing the structure of the news story from inverted pyramid to narrative changed the understanding of an accompanying news video. Whe n the story was written in narrative format, the participants recognized more details given in the video than they did when the story was written in inverted pyramid format. The inverted pyramid format is structure used for writing hard news that places th e most important information and impact at the beginning of the story and then works through the details in descending order of importance It is taught to student journalists as a fundamental to journalistic writing (Rich, 2010). There are three major the ories on the origin of the inverted pyramid: unreliable telegraph lines of the American Civil War, a shift in the educational paradigm, and as a time saving and cost saving measure (Pottker, 2003). A content analysis of two New York newspapers during the 19 th and 20 th centuries showed that both papers had a jump in the number of stories written in inverted pyramid style between 1855 and 1920. During that time period, the percentage of inverted pyramid stories a t the New York Herald increased from 6.1 to 39 percent of stories. At
21 the New York Times the percentage rose from 4.6 percent in 1875 to 14.5 percent in 1895 (Pottker, 2003). Narrative The narrative format describes the details of a story with a beginning, middle and end. It is used in many types of non fiction writing, but not used often in journalism except in feature s. Though the inverted pyramid style has become pervasive, it may not be the best format for a target group of online readers: the youth. Zerba (2008) found a group of college students recalled more details of a story about Iraq and showed more interest in the story when it was presented in narrative format than in inverted pyramid format. Narrative storytelling in the media, the author concluded, could increase the comprehension, inter est and enjoyment among younger readers. Knobloch, Patzig, Mende and Hastall (2004) found a general audience reported more reading enjoyment for story structures typically used in entertainment, regardless the content. The authors concluded the inverted py ramid format fails to maximize pleasure for readers. Taking it a step further, Boyer (1976) attacked the inverted pyramid as irrational to both the reader and the writer. The enforce ment agents. In this section, standard practice of the investigating official is to write a detailed, chronological first person account of the events that led to filling charges (Biggs, 1994). Narrative journalism is written with a different purpose, but the writing styles share similarities. Narrative journalism is told in plot form, which may be chronological (Zerba, 2008). Though it is not usually written in first person, narrative journalism is often written through the perspective of a particular char acter (Zerba, 2008). Also, small details are emphasized in narrative journalism more than in inverted
22 pyramid style (Zerba, 2008). When presented both a traditional news story written in inverted pyramid and a criminal complaint written in narrative format the reader reap s the benefits of both writing styles. While the purpose is different, this paper suggests that the effects may be similar based in part on the writing styles of the narrative and inverted pyramid. Theoretical Framework In addition to the benefits of a different writing style, a criminal complaint may also benefit the reader looking to make a quick judgment about the credibility of the story, according to one model of information processing (Mondak, 1990). The heuristic systematic informati on processing model is one of several models used in persuasive message theory to describe and predict what portions of a message a receiver will pay attention to and how the message will be evaluated. This model predicts that not all online newspaper read ers will scrutinize the news and the source of news on the same level. HSM predicts as the interest of the reader increases, so will the scrutiny the reader gives the news (Mondak, 1990). The importance of source credibility and the interest of the reader to further investigate a content of a news story will likely be influenced by his or her interest or engagement in the story. Interest in reading material is positively correlated with engagement in the reading (Flowerday, Schraw, & Stevens, 2004). Inter est affects the attention paid to the material and the learning strategy used to review the material (Shraw & Lehman, 2001). Interest can be either personal (long term) or situational (short term), but both affect deep learning, like understanding the them es and symbolism of a text (Schraw & which is the term used by the author in reference to the type of learning needed to answer multiple choice questions about facts, is not
23 influenced by interest (Flowerday, Shraw & Stev ens, 2004). There is no evidence that there is a negative relationship between interest in and basic recall (Shraw & Lehman, 2001). There is a positive relationship between interest and engagement. Interest effects the emotion invested in the activity, whi ch is an emotional engagement (Schraw & Lehman, 2001). Chaiken (1980) found engagement or involvement dictated how information is processed. Chaiken studied two methods of processing information heuristic and systematic. A later study found similar resul ts for the types of information processing on the credibility of the message. Chaiken and Maheswaran (1994) studied how participants evaluated messages based on the importance of the task at hand. When the task was categorized as high importance and the me ssage content was specific, participants scrutinized the central message systematically. When the task importance was low and the message content was ambiguous participants used heuristic source credibility cues such as likeability to judge the message. P articipants combined elements of both methods of information processing when the importance of the task was high but the message content was ambiguous (Chaiken & Maheswaran, 1994). In experiments, Chaiken (1980) found message receivers not involved or enga ged in the message are more likely to pay attention to peripheral clues about the reliability of the message. On the other hand, those highly involved in the message were more likely to scrutinize the central message for credibility (Chaiken, 1980). Resear ch observed that high stakes participants were more likely to pay attention to the content of a message while low stakes participants were more likely to pay attention to the demeanor of the person delivering the message (Chaiken, 1980).
24 Mondak (1990) fou nd the HSM model predicted the persuasive ability of messages from the Supreme Court. Though most citizens do not spend time to investigating decisions by the high court, they are likely to base their opinion on the decisions. Mondak (1990) found that deci sions made by the Supreme Court were viewed more favorably than fully explained decision s made by less credible sources. He explained that participants had low involvement in the issue, so they evaluated the message based on the heuristic cue of source cre dibility. More recent research found that source credibility plays multiple roles in high elaboration models. Tormala, Brinol and Petty (2007) determined that perception of the message was changed depending on when in the message the source was revealed. When the source of the message was given after the message, it had a stronger effect on the participant than when the source was given before the message (Tormala, Brinol & Petty, 2007). This model predicts that the when news readers have a high interest o r involvement in a news story, they will use the systematic information process and pay more attention to the content of the message. When news readers have a low level of involvement in the story, however, they will make a quicker judgment about the story based on heuristic cues, particularly source credibility. This model can be extended to make predictions about the effects of providing the original source information of a news story on recall and credibility. Those who rate a high interest in a news sto ry should read more carefully the content of the news story and therefore would have a higher recall of the details than someone who is not interested in the story. Those who indicate a low interest in news story would pay more attention to the source of t he information. If
25 the message source is an official source, the message would have higher perceived credibility among those with a low interest than a high interest. Factors Involved in Recall The factors that influence recall of news stories are varied a nd not always controlled by the journalist. In a survey, Robinson and Levy (1986) found education to be the largest predictor of comprehension. The authors also found interpersonal communication played a large roll. Even so, there are tactics news organiza tions can take to improve the understanding ability to recall messages. Some are in the hands of the reporter. For instance, Callison, Gibson and Zillmann (2009) found that the way math ratios were written was far more influential on recall than the reader Page designers can add more pull quotes to improve recall (Wanta & Gao, 1994). In an experiment, Wolf and Grotta (1985) paired stories with photographs. They found that the photo was cited as the biggest reason for reading the story. They also found that when stories are paired with photographs, the participants had better recall of details in the stories. Pipps, Walter, Endres and Tabatcher (2009) found that the impact of visual elements could differ by age. In their experiment, they foun d students had better recall of news stories after reading text. Conversely, participants aged 21 and older had better recall after they saw photographs and videos, according to their study. Wanta & Gao (1994) found the presentation of print news changed t he perception of the content. Using a variety of newspaper front pages, they found graphics and order made a difference in recall and attention to information in a group of college students. The authors concluded newspapers should include more pull quotes, more photos and out quotes improved recall more than information graphics (Wanta & Remy, 1995). Likewise, Griffin and Stevenson (1992)
26 found when a background informational sidebar with a map was included with a news story on Lithuania, participants could answer correctly a mean of 7.88 of 10 questions on a multiple choice test. With text alone, participants answered correctly a mean of 6.9 questions. Coleman and Thorson (2002) found that in two of three c ases reader knowledge increased when public health information was provided in graphical format rather than text format. While a criminal complaint, written by a police officer in narrative on a standardized form, is not entirely pictorial, many of these e arly studies included text as graphic in their definitions and results. With the appearance of USA Today on front stoops in the 1980s, criticism and research sprung forth focusing on informational graphics. Smith and Hajash (1988) analyzed 30 newspapers in 1986 and found one graphic for every 17.48 pages. Comparatively, USA Today was at the time publishing 1.3 graphics per page. The graphics were most often associated with business and economic stories, though weather was another dominating subject area for informational graphics. In these early days, infographics included charts, tables and diagrams (Smith & Hajash, 1988). Maps were the most prominent type of graphic. In an experiment by Utt and Pasternack (1993), college students were given several story packages from USA Today and asked to look over the pages as they normally would. Then they were asked questions about the content. The authors found the location of graphics was the primary predictor of what story was read first. While the graphic could dr aw the reader, the authors also found a quarter of participants misunderstood the information contained in the graphic. Ramaprasad (1991) also found informational graphics could confuse the
27 reader. In an experiment with a news story and a diagram explainin g earthquakes, the author found in general the graphic did not increase recall and led to misinformation. The author concluded with mixed feelings on infographics. Because they draw the ppen after attention is gained, informational graphics can be beneficial. But, poorly written graphics could also add to misunderstanding. attention, present the informatio n in a narrative that repeats much of the information contained in the traditional news story, and display the information in a graphic on a standardized form, recall of information will increase when the reader reads both t he traditional news story and th e source document. H1: Exposure to a copy of a court document and a traditional news story about exposed to a news story alone. RQ1: What factors influence recall? Factors Involved in Credibility of News While the credibility of print journalism has long been the subject of research, more recently those principles are applied to journalism presented online. Credibility research is separated into specific news story credibil ity (Meyer, 1988; Armstrong & Collins, 2009), source credibility (Whitehouse, White & Andsager, 1991), and message source (Sundar & Nass, 2001). This study will draw on the benefits of the source credibility, a sworn statement from a police officer, as wel l as message credibility created on a news
28 A message is credible if the receiver believes it to be true and valid (Wilson & Sherrell, 1993). Expertise, goodwill and reliability make up the majority of source credibility, while likea bility and dynamism are also factors (Wilson & Sherrell, 1993). In their quantitative meta analysis of previous studies, Wilson and Sherrell (1993) found expertise to be the most effective factor in source credibility. Other factors include being fair, unb iased, and accurate (Meyer, 1988). Because of the expectation that experts will have more valid points than lay people, use of experts in a news story can bias a Perry, 200 7). Additionally, Sundar and Nass (2001) posited that the technology that delivers the news (i.e. the Internet) is considered a source itself by some readers. Readers may perceive source documents as more credible because the documents are authored by expe rts in the field instead of a journalist. While anonymous sources hurt the credibility of the news story (Sternadori & Thorson, 2001), including the source in its entirety is at the other end of the transparency scale. Reading the source document after th e news story could have a larger effect on credibility than reading it before. Tormala, Brinol and Perry (2007) found providing source information after the message was more effective than providing it before the message. This study examines whether inclu ding a criminal affidavit in narrative form as an accompaniment to an inverted pyramid news story could increase the readability and improve credibility.
29 H2: Exposure to a copy of a court document and a traditional news story about that document will incr compared to being exposed to a news story alone. RQ2: What factors influence credibility?
30 Figure 2 1. Proposed model of influences on recall and credibility
31 CHAPTER 3 METHOD Experimental D esign A post test with control experimental design was used to test these hypotheses and research questions. Participants in the treatment condition read a traditional news story about a crime and then the original source document, a criminal complaint. Pa rticipants in the control condition read only the traditional news story. At the end of the reading period, both groups answered questions to assess their recall of the news story and their perception of credibility of the news story. The answers to the qu estions were analyzed to measure the differences between the control and treatment conditions. Based on the experiment design e xposure to the source document was the primary independent variable. Other independent variables measured included time spent re ading the material, story interest and news interest. Recall and perception of credibility are the primary dependent variables. Variable construction will be discussed later in Chapter 4 This post test only design is limited because it cannot control for all outside factors. While the questionnaire inquired about interest in the story and self reported prior knowledge about the case, it did not measure prior knowledge about the case or similar cases. While the procedure measures the time the reader spends with the news story and the source document, it cannot account for individual learning differences. perceived credibility of the news story.
32 The News Story and the Source Document The news article used in this experiment was published October 6, 2010, on the website of The Boston Globe after American citizen and technology company employee Elliot Doxer was charged by the FBI with wire fraud. The 360 word story is written in traditional inverted pyramid news style. See Appendix A. It begins with a hard news lead, delayed identification, and then gives details in time sequential order. Information and quotations are taken directly from the federal complaint and are attributed accordingly. It was published in the Business Briefs section of website and does not contain any notations that it was changed since its original publication. This story was chosen because the date (October 2010) and the location (Massachusetts) reduced the likelihood that the participants will have previous knowledge of the case and its details. The byline and masthead were not included. The text of the story was placed on an electronic document designed to look like the webpage of an aggregate news website. The only change to the content of the story was to indicate the cities mentioned were in Massachusetts. Documents can be converted to an electronic format and made available to online readers. Adobe Systems created portable document file, or PDF, as a file format that captures and replicates the formatting of a document (Noack, 2001). The software needed to read PDFs is free, and newspapers began using this file format as a way to make an online copy of newspapers available with the same look as the print version (Noack, 2001). Some newspaper s use this technology to sell subscriptions of the newspaper online instead of charging on a pay per view basis (Beleyen & Van Hove,
33 2010). Other newspapers have made PDFs of non newspaper documents available to supplement online news coverage. The origina l PDF of the complaint is a 22 page file available directly from the U.S. Courts through Public Access to Court Electronic Records (PACER). It was also published in its entirety by many news organizations. The document has numerous clues to its origin, fro m the legal headings to the stamp by the federal magistrate judge who signed the arrest warrant. The document begins with an affidavit by the FBI special agent who investigated the case, James A. Cromer that details his credentials and jurisdiction. It co ntinues with a narrative of the evidence collected and investigation into the defendant. This narrative shares characteristics with narrative journalism because it is told through the perspective of the investigating agent and has a beginning, middle and c onclusion. The final pages include a criminal case cover sheet that lists vital statistics and basic information about the defendant and the location of the alleged crime, a list of charges (one count of wire fraud), and a signed arrest warrant. The warra nt does not indicate it was returned, meaning the warrant does not record that the accused was apprehended even though the news story does indicate arrested and held without bail. The PDF of the criminal complaint was reformatted and condensed to three pag es. This was done by removing the double spaced formatting, removing extraneous pages like the cover page, arrest warrant and physical description page. The narrative written by the investigating officer was not significantly altered in content because pri or recall (Zerba, 2008) Also left in the PDF were the stamps by courts and signatures by
34 the investigating agent because those elements may have been clues used in the he uristic systematic information processing to determine the credibility of the document. See Appendix A. Participants P articipants for this experiment were recruited from a general elective course ( TV and American Society ) offered within the Coll ege of Journalism and Communications at t he University of Florida. This class was chosen because it provides a large pool of stu dents from a variety of majors. Volunteers were invited to participate in the experiment by email in exchange for extra credit i n the course. The experiment was available from Nov. 29 to Dec. 9, 2011. At that time, 164 participants had completed the experiment. The results of three participants were removed because of missing data. An additional three surveys were removed because the participant indicated he or she had prior knowledge or was unsure if he or she had prior knowledge. The answers of 158 participants were analyzed in this study Participants ranged in age from 17 to 26, and the median age was 20. Of those participants, 71 percent were female. The majority of the participants were white (64 percent). Ten percent of the participants were black, and 17 percent indicated they were of Hispanic ethnicity Almost all participants (96 percent) indicated they were American citiz ens. represented. With the exception of one graduate student, the participants were 10 percent freshmen, 21 percent sophomores, 42 percent juniors, and 26 percent seniors. Procedure P articipants were s ent an email that included a link to the experiment, which they could complete on any computer with Internet access. The link led participants to the
35 informed consent agreement, which participants could read and print. Once participants agreed to the infor med consent, they were randomly assigned to either the treatment or Next, all participants were presented with a screen modeled to look like a generic news aggregations website and instructed to read a traditional news story about an employee of a technology company accused of wire fraud. On the left side of the story link did not lea d the participant anywhere, but the click was recorded. This was available to all participants. See Appendix A. Participants were instructed to read through the news story as they would normally read through an article on a newspaper website. The time the participant spent on this page of the survey was recorded by the experiment software Participants in the treatment condition were additionally presented the document used as a source for the wire fraud story, a federal criminal complaint filed against the accused. Participants were instructed to read though the article and documents as if they were reading it on a news website, and time spent on each of the three pages was recorded by the experiment software When a participant finished with the reading, h e or she was asked to complete a questionnaire that captures demographic information, and tests recall, perceived credibility, and interest in the news article. The experiment software prevented participants from using Internet navigation to see prior scre ens once the participant advanced to the next question. On parts of the experiment that had questions, the experiment software also required participants to give an answer before advancing to the next part of the experiment.
36 Of the 158 participants used in analysis, 81 were assigned the treatment condition and 77 were assigned to the control condition. All participants spent an average of 177 seconds ( 2 minutes 57 seconds ) reading the one page news story. Participants in the treatment condition spent an ave rage of 394 seconds ( 6 minutes 34 seconds ) reading the three page complaint. Measures The questionnaire administered to all participants includes four sections: recall, credibility, interest and demographics. See Appendix B Participants were first asked t o summarize the news story in a few sentences. Next, recall questions were administered to reduce the elapsed time between reading the news story and answering the questions. The seven multiple choice questions asked the participant to recall details from the content of the news story. The answers to all the recall questions are all in the news story. Reading the original source document was not necessary to answer any of the questions T he questionnaire then asked eight questions on credibility on a scale of 1 to 7. These questions measure elements of credibility of the news story (Armstrong & Collins, 2009; Arpan, 2009; Kiousis, 2001; Gaziano & McGrath, 1986). The results of the questions were compared between the two groups to determine if those who read the source document perceived the news story to be more credible than those who read the news story alone. The participants were asked nine questions about their interest in news, news stories involving the themes of the news story, and this particular ne ws story. These questions, answered on a scale of 1 to 7, were used to determine if recall and perceived
37 credibility are related to interest in the news story. Wanta & Gao (1994) found that interest in the story increased recall in younger newspaper reader s. This set of questions was used to determine how interest affected recall and observe any instance of heuristic systematic processing Finally, the questionnaire asked eight demographic questions. Demographics collected included age, gender, major, race /ethnicity, and status as an American citizen. Participants were asked to provide their university identification username and number for the purposes of awarding extra credit. However, this information was separated from the responses so there was no link between individual responses and identifying information. The measures were pre tested on a mixture of graduate students at the university and adults not affiliated with the university. The results of the pre test participants were not included in the ana lysis, and none of those who took the pre test also participated in the experiment.
38 CHAPTER 4 RESULTS Using a post test experimental design, this study examined the effects of a source document on recall and credibility of a news story. Participants were randomly assigned to either the control condition and read only the news story or the treatment condition and read the news story and the source document. A link in the source document indicating it led to a PDF of the source document was available to all participants, though that link was not functioning during this experiment. While participants were not instructed to click the link, 7 percent of participants did click that text Open ended Summaries All participants were asked to summarize what they ha d read in an open ended question. This question was mainly used as a manipulation check. The control group was asked this question immediately after it finished reading the news story. The treatment group was asked this question immediately after it finish ed reading the final page of the criminal complaint. Responses from both groups indicate participants read the story. The answers from the control group ranged from a vague single sentence to detailed summaries. Many included specific names and numbers. F or example: the information contai ning contact information for financial clients, saying he was doing to help his homeland, he is a Jewish American. He does not have a lawyer for his court case and he is being charged up to $250,000 and up to 20 y (Participant 30, control condition.)
39 Some participants were blunt with their responses: I only skimmed over it because it did not pique my interest and the directions said to read it like you normally would. I do remember that there was sensitive information that was being given to other individuals over and extensive period of time. The suspect or person visit ed the dead drop over 60 times. (Participant 100, control condition.) Participants in the treatment condition showed a similar range of generalized to specific summaries of the stories. While many of their descriptions continued to focus on the details included in the news story, several summaries included details that were only contained in the criminal complaint. Nine of the 81 responses from those in the treat ment group referenced a specific portion of the complaint talking about the mother a terrible human being and caused me a tremendous suffering. Not eno ugh bad things can happen t (Criminal complaint). This detail, which was not included in the news s tory, was repeated almost as an aside in several summaries. A man named Doxer was accused of committing a crime. He claims he did nothing wrong but the authorities, and the criminal report, say differently. His moth (Participant 66, treatment condition). Generally, the summaries of both those in the control and treatment condition reflect that the p articipants read the information presented to them. The summaries include details of the readings that participants were able to recall without the prompting of multiple choice options.
40 Variable Construction Time The news story was shown on one screen. Th e time participant was instructed to read the news story and continue to the next screen when finished. The time the participant spent on the story with the news story was recorded. The mean length of time spent on the news story by all participants was 2 minutes 57 seconds (S.D.= 270 seconds). The criminal complaint was distributed across three pages. Participants in the treatment condition saw only one page of the complaint at a time. They were instructed to continue to the next page when they finished. T he time spent on each page was recorded. The time reading each page was summed to get the cumulative time spent reading the criminal complaint. Of those who were in the treatment group, the cumulative mean time spent reading the complaint was 394 seconds ( 6 minutes 34 seconds ) (S D =647 seconds) The range was from 7 seconds to 3,958 seconds (1 hour, 5 minutes, 58 seconds). A variable was created that combined the time spent on the news story and the time spent on the criminal complaint for all participants The mean combined time for all participants was 6 minutes 19 seconds (S.D.= 620 seconds ) News interest All participants were asked to self report their opinions on and interest in general news and this specific story. For these questions, participants w ere asked to rate their answers on a scale of 1 to 7, with 1 being low and 7 being high. These questions were created with the topics and categories of news related to this story.
41 There were two batteries of questions about interest in news. The news inte rest battery included six questions about news and types of news. Participants were asked to rate their answers on a scale of 1 to 7, with 1 being not at all and 7 being a great deal. Those questions included: The C for these six items is 0. 91. This indicates high internal consistency (George & Mallery, 2003) The cumulative mean was 26.53 of a possible 42 (S.D.=7.7) Story interest The story interest set of questions focused on this particular news story. Participants were asked to rate th eir answers on a scale of 1 to 7, with 1 being not a t all and 7 being a great deal. The C for these six items is 0.936. This indicates high internal consistency (George & Mallery, 2003) The cumulative mean was 23.68 of a possible 42 (S.D.= 8.86 ) Recall Questions Recall was examined using seven multiple choice questions that could be answered using the news story alone While some questions were more basic and others were more detailed, each correct answer was weighted equally. The mean num ber of questions answered correctly was compared between the two groups. T he information included in the news story was repeated in the complaint. The most frequent answer for each of the recall questions was the correct answer. The mean number of recall q uestions answered correctly was 5.28 (S.D.=1.73). Twenty seven percent of participants answered all seven questions correctly. Almost all participants answered at
42 least half questions correctly. Fewer than 8 percent of participants answered two or fewer qu estions correctly. To test the H1 positing that those who were exposed to the news story and the source document would recall the details of the story more than those who saw the news story alone, an independent sample T test of means was used. Participan ts were awarded one point for a correct answer and zero points for an incorrect answer. The average number of questions answered correctly was compared between the group that was exposed to the source document (M=5.51) and the group that saw the only the n ews story (M=5.05). There was no statistical difference betwe en the means of the two groups. The proportion of correct answers was calculated for each participant. There was no statistical difference between the proportion correct in the control group (M=0 .7217, S D =0.26412) and the treatment group (M=0.7866, SD=0.22714). The exposure to a source document did not increase recall of the news story. See Table 3 3. questions was totaled fo r each participant. There was marginal statistical significance (p<.10) to show that the participants in the control group (M=1.1688, SD=1.31922) were more than those in th e treatment group (M=0.8519, SD=1.05013). When those who answered all recall questions incorrectly were removed from the analysis, there was marginal statistical significant difference (p<.10) in the proportion of recall questions answered correctly betwee n the control (N=75) and treatment (N=79) groups. In this analysis, those in the treatment group (M=0.8065, SD=0.19144) answered more questions correctly than the control group (M=0.7410, SD=0.23915).
43 The treatment group was more likely to answer two parti cular recall questions correctly. Participants were asked the name of the agency in charge of the criminal investigation. Eighty four percent of treatment group correctly answered that question, while 58 percent of the control group got it right (p<.00 1 ). Participants in the control group (p<.1 0 ). Similarly, participants were asked were asked what the subject of the story left at drop off location. Eighty five percent of treatment participants answered that correctly, while 73 percent of the control group answered it correctly (p<.1 0 ). See Table 3 5 for a comparison of means of these two questions. In the remaining five recall questions, there was no correlation between ex posure to the source document and a correct answer. Based on these results, t here was not enough evidence to support H1: Exposure to a copy of a court document and a traditional news story about that document will when compared to being exposed to a news story alone. Factors Influencing Recall RQ1: What factors influence news reader recall? A M ANOVA was created to analyze the predictors of recall. To create this model, variables were split at the mean for all part icipants and assigned to high and low groups For news interest, variables were split into the high group (N=71) and low group (N=78). For story interest, participants into a high group (N=82) and low group (N=67). For combined time, participants were spli t into a high group (N=72) and a low group (N=77). For credibility, participants were split into a high group (N=74) and a low group
44 (N=75). Participants were also separated into two groups by race/ethnicity. The first group was those who identified themse lves as white (N=95) and all others (N=54). High news interest was marginally significant with a higher number of recall questions answered correctly (p<.1 0 ). High story interest was a statistically significant influence in answering more recall questions answered correctly (p<.00). Those who spent a longer than average time reading the information also answered more recall questions correctly (p<.00). Using the data collected to create a MANOVA predicting the proportion of recall questions participants an swered correctly, several main effects and one interaction are statistically significant Main effects included story interest and time. Those who had high story interest answered more recall questions correctly (p<.01). Those who spent more than the mean time with the information also answered more recall questions correctly (p<.05). Race emerged as a marginally significant factor in predicting recall. Whites answered more recall questions correctly than non whites (p<.1 0 ). See Table 3 6. Analysis showed a n interaction between story interest and condition in predicting the number of recall questions answered correctly. Those with high story interest answered more questions correctly in both the treatment and control conditions than those with low story inte rest. However, there was less of a difference between the proportion of recall questions answered correctly by those in the high and low story interest groups in the treatment group than in the control group. See Figure 3 1. Credibility The first slate of questions originally included seven questions designed to assess the perceived credibility of the news story. These questions were culled from several
45 studies (Armstrong & Collins, 2009; Arpan, 2009; Kiousis, 2001; Gaziano & McGrath, 1986). Based on factor See Table 3 7 for the items included in this variable. The C This indicates a moderately high level of internal consistency. The cumulative mean of these six items was 29.06 of a possible 42 (S.D.= 6.427 ) To test the H2 positing that those exposed to a news story and source document would rate the news story as more credible than those who saw only the news story, an independent sample T test was used to compare means. Using the six question credibility measure, there was no significant difference between the control (M= 29.1429 S.D.= 5.89539 ) and treatment (M= 28.9012 S.D.= 6.9292 9 ) groups. See Table 3 8 for a comparison of means of these questions There was no significant difference between condition groups for any particular question. See Table 3 9 for individual credibility questions. Based on these results, there was no eviden ce to support H2: Exposure to a copy of a court document and a traditional news story about that document will increase the news story alone. Factors Influencing Credibili ty RQ2: What factors influence the perception of credibility of news readers? Using the data collected to create a model predicting the perception of credibility, there were a several statistically significant factors A MANOVA analyzed the factors measure d in this study including story interest, news interest, time, gender, race, recall
46 and condition. Main effects included time, news interest and story interest. The more time participants spent viewing the information, the more likely they were to rate its credibility higher than the mean (p<.05) The higher participants rated their interest in (p<.05) The most significant predictor of credibility was story interest. The group that rated story interest above the mean was more likely to highly rate the credibility of the story (p<.01) See Table 3 10.
47 Figure 3 1. Histogram of frequencies of combined time participants spent reading article and source document
48 Table 3 1. Items included in the News Interest variable from all participants Question M (of 7) S.D. I am interested in general news. 4.87 1.540 I am interested in reading news online or in a newspaper. 4.63 1.610 I am interested in news about crime. 4.58 1.507 I am interested in news about law enforcement investigations. 4.37 1.469 I am interested in news about international affairs. 4.30 1.496 I am interested in news about courts. 3.78 1.595
49 Table 3 2 Items included in the Story Intere st variable for all participants Question M (of 7) S.D. This news story was interesting to me. 3.94 1.748 I enjoyed reading this news story. 3.71 1.698 I thought this news story was understandable. 4.64 1.617 I though t this news story was e njoyable. 3.71 1.636 I wanted to know more information about this news story. 3.51 1.812 How much attention did you pay to this news story?* 4.09 1.607 *This attention question removed from the credibility questions was included in this analysi s.
50 Table 3 3. Independent Sample T test of Proportion of Recall Questions Correct by Condition Mean SD t p Control (77) .72 2 .264 1.652 .101 Treatment (81) .78 7 .227
51 Table 3 4. Independent T test of proportion of recall questions correct with th ose responses that answered all incorrectly removed Mean S.D t p Control (75) .741 .239 1.872 .063 Treatment (79) .80 7 .191
52 Table 3 5. Independent sample T test of specific recall questions by condition Question Mean SD t p What did the person charged with a crime leave at the Control (77) .73 .448 1.925 .056 Treatment (81) .85 .357 Which agency investigated the crime? Control (77) .58 .496 3.652 .000 Treatment (81) .84 .369
53 Table 3 6 Multivariate a na lysis of v ariance examining p roportion of r ecall q uestions c orrectly by i nfluencing f actor Means Mean Sq DF F value p value Condition Control Treatment .717 .761 .053 1 1.160 .283 Gender Male Female .741 .731 .000 1 .009 .923 Race/ethni city All others White .706 .773 .153 1 3.345 .070 News interest Low High .739 .740 5.464E 5 1 .001 .972 Story interest Low High .667 .882 .588 1 12.869 .000 Time Low High .692 .786 .243 1 5.320 .023 Condition X story int erest Control/low interest Control/high interest Treatment/low interest Treatment/high interest .595 .839 .738 .734 .344 1 7.522 .007
54 Table 3 6. Continued. Means Mean Sq DF F value p value News interest X story interest Low interest/low int erest Low interest/high interest High interest/low interest High interest/high interest .683 .794 .651 .829 .034 1 .743 .390 Error Total .046 13 9 14 9
55 Figure 3 2 Mean proportion of recall questions answered correctly by condition infl uenced by story interest
56 Table 3 7 Items included in credibility variable for all participants Question M S.D. How much do you trust this news story? 4.726 1.247 How accurate do you think the information in this news story is? 4.764 1.432 How credible do you think the information in this news story is? 4.768 1.362 How newsworthy do you think this story is? 4.986 1.499 How fair was this news story? 4.725 1.216 How informational do you think this news story was? 5.044 1.351
57 Tabl e 3 8. Independent sample T test of perception of credibility by condition Mean SD t p Control (77) 29.1429 5.89539 .235 .814 Treatment (81) 28.9012 6.92929
58 Table 3 9. Significance of individual credibility questions Question Condition N Mean (Scal e of 1 to 7) St. Dev. Sig. How much do you trust the information in this news story? Control 77 4.88 1.11 .129 Treatment 81 4.58 1.37 How accurate do you think the information in this news story is? Control 77 4.79 1.21 .803 Treatment 81 4.74 1.36 How credible do you think the information in this news story is? Control 77 4.81 1.29 .724 Treatment 81 4.73 1.42 How newsworthy is this news story Control 77 4.91 1.48 .524 Treatment 81 5.06 1.51 How fair was th is news story Control 77 4.70 1.18 .790 Treatment 81 4.75 1.25 How informational was this story Control 77 5.05 1.24 .945 Treatment 81 5.04 1.45
59 Table 3 1 0 Multivariate a nalysis of v ariance e xamining p erception of c redibility by i nfluen cing f actor Means Mean Sq DF F value p value Condition Control Treatment 29.462 28.422 29.953 1 .812 .369 Gender Male Female 28.602 29.276 12.667 1 .343 .559 Race/ethnicity All others White 28.704 29.180 7.637 1 .207 .650 News interest Low High 27.556 30.328 226.094 1 6.129 .014 Story interest Low High 27.397 30.487 278.838 1 7.558 .007 Time Low High 27.544 30.340 218.711 1 5.928 .016 Condition X story interest Control/low interest Control/high interes t Treatment/low interest Treatment/high interest 27.639 31.284 27.155 29.689 10.849 1 .294 .588
60 Table 3 10. Continued. Means Mean Sq DF F value p value News interest X story interest Low interest/low interest Low interest/high interest High intere st/low interest High interest/high interest 25.462 29.650 29.332 31.323 35.909 1 .973 .326 Error Total 36.892 140 149
61 CHAPTER 5 DISCUSSION This quantitative experiment examined effects of exposure to a source document on the reader. It hy pothesized that recall of the details of a news story was increased by also reading the document used to as a source for the information in the story There was minimal evidence to show that reading the source document increased recall of details from the news story based on these recall questions. Two questions had significance between exposure to the source document and correct answers to the recall questions. On the majority of recall questions, exposure to the source of the document had little effect. T here was no evidence to show that exposure to a source document increased perception of credibility on the news story. Details of the news story were remembered just as well when readers saw a one page summary of the events written as a traditional news st ory as when readers saw that news story and a three page document written in chronological order with many more details. Readers perceived a news story that lacked any indication of its origin or author as more credible than not. When readers saw an offici al document giving a personal account of the events described in the news story, it did not increase the perception of credibility. Other factors, particularly story interest, played a much larger role in predicting recall than providing the source docume nt. The influence of story interest on recall was diminished by exposure to the source document. Though the literature linked heuristic systematic processing model to credibility, the results of this study show it may be linked to recall. Recall was positi vely influenced by time exposed to the news story and
62 source document. Overall, participants spent a higher than anticipated length of time reading the documents. Participants also generally scored well on the recall questions. The perception of credibilit y was generally unaffected by exposure to the source document. Here also, story interest had a much larger influence. The news story was rated as more credible than not, but the results of this experiment did not reveal much of what led readers to that per ception. Emerging research shows that the definition of credibility is changing between generations, and this experiment showed exposure to a source document did not enhance the credibility of a news story among that group. Based on these results, providin g a source document alongside a news story may not be an effective method for a news organization to increase the across the board perception of credibility and recall among its readers. This should not be taken as reason for the news industry to abandon t he practice of posting source documents. Other effects and effects on specific readers are not known yet. The results of this study suggest academic research needs to explore the concepts of credibility and heuristic systematic processing of news further. Recall The results showed that participants did read the news story. The average student answered most of the seven questions correctly. Participants spent almost three minutes reading the story before moving onto the next portion of the experiment. The t ime spent exposed to the source document was also significant enough to read the three page document. At about 2 minutes per page, there would be time for a careful reading. This length of time in fact, seems on the high side of what would be expected of college aged students. There are some possible contributing factors to the average
63 time, including density of the material, unfamiliarity with the jargon and terms used, and distractions. On the first page of the document, the narrative was only the botto m third credentials and then launched into some legal language. This type of sentence construction and vocabulary is not used in journalism. Perhaps participants needed to read slowly or reread particular passages to understand the content. It is also possible that because participants were not in a controlled setting, they were distracted by other information on the Internet or in the environment. In the case of the partici pant who spent more than an hour on the source document, this is highly likely. However, participants could begin this experiment in their homes or at school, which are likely the same locatio ns they would be consuming real news. Participants may face thos e same distractions in non experimental news reading. Acknowledging that the time recorded for exposure may not be the time spent reading, the open ended summaries checked that the readers knew the main themes of the story. These open ended question s refle cted that participants did read the article Robinson and Levy (1986) found education to be the largest predictor of comprehension. Because the population used in this experiment were college students, the group may have a higher overall comprehension leve l than a population more evenly distr ibuted across education levels. While overall there was no statistical difference between the control and treatment groups in recall, there was marginal statistical difference between the groups when the participants wh o answered all recall questions incorrectly (N=4) were removed from the analysis. Those who answered did not answer a single recall question correctly
64 meaning they could not identify correctly whether the main subject of the story was either male or fema le either may not have read the news story or may have a language barrier that prevented them from understanding anything about the story. Comparing the means of the control and treatment groups of only those who answered at least one recall question cor rectly may be a better gauge of recall of those who actually read the story. There was a positive correlation between exposure to the source document and recall on two particular questions. The relationship between recall and one question may be explained because of repeated information. This question asked for the agency been so employe Another question asked what the defendant left was only found in the news story. The criminal comp laint used different language to convey the same concept. It would be interesting to repeat the experiment with more recall questions. While this series of questions did not ask for specific numbers, the open ended summa rizations of the stories from both the control and treatment groups included numbers such as the potential fine and the amount of bail. While neither of those details was included in the source document, numerical recall questions could add a dimension of understanding about the type of information that exposure to source documents increases.
65 The three significant predictors of recall included race, interest in the story, and time. It is difficult to explain why race would influence recall. The race of the subject in this news story was never discussed; however, he did identify himself as a Jewish American. Appiah (2002) found that blacks were more attentive to black characters in television shows and blacks were less attentive to white characters in televis ion stations. These attention levels in his studies led blacks to have less recall about white characters than black characters. While participants may not be self aware of the attention they gave to the story, there was no correlation between race and the reported level of attention paid to the news story. Adams and Cleary (2006) suggested that the model for understanding the relationship between journalism and race may be far more complex than researchers rity paradox increased diversity in newsroom employees did not increase readership of that newspaper. In fact, the authors found that trust of the newspaper by blacks in the community actually decrease as newsroom diversity increased. Alo ng the same pattern of thought, the results of this research should not be taken to mean that non whites do not recall news as well as whites. Rather, it should hint that there is something more complex than is currently understood in the interactions betw een race and this news story. The other two factors in recall interest in the story and time spent reading the story and source document are more readily explained. Those who spent more time reading the story would have more time to carefully read and learn its information.
66 Those who have more interest in the story would be more likely to give attention to the story and its details. The findings of this study are in line with the correlation between interest and recall of young readers that Wanta and G ao (1994) found. They, however, counter the results from Flowerday, Shraw and Stevens ( 2004 ) that concluded the type of learning needed to answer multiple choice questions about facts is not influenced by interest. The analysis in this experiment showed th at recall demonstrated on multiple choice questions about facts was always higher among those with high story interest. However, the effects of story difference were diminished when readers were exposed to the source document. Those results fit with the he uristic systematic processing model. Those who had high interest in the story sought more information from the story when presented it alone. Those who had low interest in the story learned less from the story alone, but when more information was forced on them they absorbed it. Armstong and Nelson (2005) found that when readers used a heuristic method of reading they read less carefully and allow stereotypes to guide judgments. The participants in this study who had low interest in the story may have answe red fewer questions correctly because they read the story less carefully and then guessed the based on what cues they did remember from the story. In non experimental situations, the effects of source documents on recall may not be prevalent among those wi th low story interest. If the reader has low story interest he or she may be unlikely to choose to view the source document to obtain more information. Conversely, the reader with high story interest would be more likely to read
67 the source document even t hough it had a decreased benefit on his or her recall of the material. While some studies showed that narrative journalism led to better recall than the inverted pyramid the narrative criminal complaint in this story did not lead to better recall than the inverted pyramid news story. Narrative journalism has many of the same qualities as the narrative in a criminal complaint, but the two are not the same. A journalist writes a narrative for a different purpose than a complainant writes a narrative. The two professions are taught different standards and vocabulary. It could be that those differences are significant enough that the effects produced by narrative journalism are not replicated by other types of narrative. For this reason, it would be interesting to repeat this experiment using a different type of source document that contains a narrative, such as a grand jury report. T his experiment showed that the inverted pyramid provided information in a way that all of the questions asked could be answered by a majority of readers. In one page, it explained the events in a way that readers understood and remembered. Three pages of more specific explanations written in chronological order and in the first person perspective did not emember the details of the news story. This shows, perhaps, that the journalist who wrote the story did an efficient job by taking complicated ideas, simplifying them, condensing them, and presenting them in an understandable format. Perhaps presenting mor e information to the reader was superfluous to understanding the main points of the story.
68 the news in an understandable way, the author of this news story succeeded in doing that in one third the space as the au thor of the criminal complaint narrative Credibility This experiment measured if exposure to a source document increased the perception of credibility of the news story. This hypothesis was not supported. There was no evidence that those who were shown th e criminal complaint found the news story to be any more credible than those who only saw the news story. A nalysis showed time and story interest were factors positively influencing perception of credibility. Those who had high story interest or spent lon ger than average time reading the material rated the credibility of the news story higher. These results affirm findings of past studie s including Flowerday, Schraw and Stevens (2004) that found interest in stories is positively correlated with engagement in reading. Chaiken (1980) found that readers with high engagement were more likely to scrutinize a message for clues to credibility. Spending a longer time reading the material is a sign of engagement. Those who are interested in reading the story are mor e lik ely to engage with the story. Readers are more likely to believe material that engages them. While the heuristic systematic processing model was used in this study to predict that the perception of credibility would increase in those with low story in terest with exposure to a source document, analysis did not show that result. There was no interaction between story interest and condition on the credibility measur es. These results may indicate that the perception of credibility among these participants is more complicated than past research indicated. There may be more factors involved in judging credibility than simply the content of the message and the sources.
69 On the whole, participants rated the news story as more credible than not. When they were pr esented a story written in inverted pyramid style but lacking any association with a news organization or particular author, they reported that they believed in and trusted in the information there. Providing an official document that carried with it the n ames of federal courts, a federal judge, and a FBI agent did not improve that credibility. Showing the reader exactly where the information contained in the news story came from and providing more details about the source did not i mprove the credibility of the story. While the use of anonymous sources harms the credibility of news, providing extensive details about the source did not improve credibility. It could be that the reader was overwhelmed with details about the source of the news story. While the experiment did not test recall of the source document, the in the open ended summaries. Those summaries revealed that the readers paid more attention to certain details of the motive of the defendant that were not mentioned in the news story. Perhaps it speaks to the way American readers are trained to consume news. It could be that the inverted pyramid format and journalistic writing style is so routinely associated wit h credible information that the association carries over to situations where other credibility cues are absent. Maybe readers need only a certain level of identification of the source of a story to deem its credibility. When given more than that level, sou rce identification ceases to have any further benefit on credibility. This experiment did not reveal what about the news story made it credible. While the credibility questions used h ere were proven in a past studies (Armstrong & Collins, 2009; Arpan, 2009 ; Kiousis, 2001; Gaziano & McGrath, 1986) the results of
70 this study may contribute to a number of studies research suggesting differences in the definition of credibility is different between younger and older readers. Amstrong and Collins (2009) found, for example, that college students did not rate the local daily newspaper produced professionally as any more credible as the campus newspaper produced by students. Chung, Nam and Stefanone (2012) found that the traits college students used to judge credi bility of news are different for print and online news. In a survey, the students reported factors like expertise and trustworthiness did influence their judgment of credibility for the online version of legacy news organizations, such as nytimes.com. Howe ver, the research found that hypertexuality and technological components were the largest influence on the perception of credibility among young readers of aggregate news sites like news.google.com. Armstrong and McAdams (2011) found that among young reade rs the predictors of trust differ between blogs and traditional media. Because the news story presented in this experiment was designed to look like it was on an aggregate news website, the lack of working links and interactivity associated with the story may have influenced the perception of credibility among the college students used as participants in this study. Based on the results of this experiment Figure 2 1 needed revisions. Specifically, heuristic systematic processing was not found to influence credibility but did influence recall. This was inconsistent with other studies (Flowerday, Schraw and Stevens, 2004; Mondak, 1990) While broad human behavior and HSM are not likely to change over time, it is more likely that the way people judge news chan ges as the way news is presented changes. So, while the model of HSM is relevant to other types of messages,
71 it may not be fully understood as it relates to news credibility and recall among American readers who are beginning to see news presented in other formats than traditional newspapers and television broadcasts. See Figure 5 1 for the findings of this experiment of the factors influencing perception of credibility and recall While the arrows indicate the direction of relationships found in this study it is possible some relationships go both ways. For instance while spending more time with the story increases credibility, this study did not examine if having a higher perception of credibility meant that the participant then spent more time reading th e story. Study Limitations While this experiment was designed to test the effects of source documents on recall and credibility of news stories, it is limited by its design, manipulation materials and experimental population. This experiment was a post te st only design. It did not test the change in recall or of credibility. The credibility of the news story graphic used in this experiment was not pretested for cred ibility. Nor was the source document pretested for credibility. While the results indicate what the perception of credibility was of the news story alone, no group was shown the source document alone. Additionally, participants were not in a controlled env ironment when they were exposed to the news story or the source document. While a participant may have left his or her Internet browser open for certain time period to the news story, it is not certain that the participant spent that time reading the infor mation. Distractions from environmental sources could affect the results of this experiment, especially recall because of the level of engagement and attention given to the information. However,
72 this environment may also be a more realistic replication of the conditions where people read news online because they were able to participate from their homes, school or work. A laboratory environment may have eliminated times in the extremes by and reduced the number of distractions to the participants. The popu lation of participants in this study was from a general elective course offered within the College of Journalism and Communications. Students from a variety of disciplines participated, but all the participants were students. They had the same education le vel, were in the same geographic area, and were younger than the average population. The experiment participants were more than two thirds female, though there was no sign that there was any significant difference in the answers between the sexes. An exper iment using a population more diverse in education, location and age may be helpful to generalizing results. The particular news story and source document used in this experiment may be a limitation. While participants were asked to report any prior knowle dge about the story, there was no pre test to determine knowledge of this story or of the criminal court system. The types of news story and source document used in this story were only one example of a news story relying on a document. Other types of news stories, such as investigative stories, historical feature stories, or poll stories, may produce different results. Similarly, different types of source documents, such as government budgets, military reports, or vital statics records may produce differen t results. Future Research Future research is needed to determine if exposure to source documents benefit either news organizations or readers. This study does not answer all questions about the effects of source documents. Nor is it able to answer how be st to present source
73 documents. If the use of source documents continues, more baseline data will be needed for further research. That includes basics about when and why newspapers provide source documents. It also includes when and why readers examine sou rce documents. This study was not designed to measure how many participants choose to read the source document. However, when presented with the news story, 7 percent of readers clicked the link included in the news story. Different research could explain what motivates readers to look at source documents, what types of source documents are most interesting or important to readers, and how frequently readers seek out source documents. While the heuristic systematic information processing model predicted tha t readers would use cues from the source document to judge story credibility, this study found no link between perception of credibility and exposure to source document. More research is needed to understand why there was no effect or what, if any, cues wi thin a source document would improve the perception of credibility. Additional research could establish if any source documents would improve the perception of credibility or that only there was no link between credibility and this particular pairing of ne ws story and source document. This study found a positive correlation between exposure to source documents and recall on particular questions. On other recall questions there was no correlation. More research could establish what types of information read ers recall better when they are exposed to source documents. Overall, the effect of source documents on recall
74 was not highly significant. Future research could repeat this experiment with a different pair of news story and source document. The effects of source documents on readers could also be examined in a study that reviews the online comments section of a news story that provides the source document. Comments could provide clues to what readers are paying attention to in both the news story and the s ource document. The written comments could also indicate how readers scrutinize a news story differently after they read the source document. Finally, future research could study if news stories are written differently when they are accompanied by the sour ce document. If a journalist knows the public will be scrutinizing the same documents used to write the news story, does he or she choose to present the information differently? The effects of providing source documents on the journalist or news story are unknown. While news organizations and academic research could benefit from knowing the effects of source documents, it may also be a benefit to know that journalists are providing information in a way that readers remember the details With the knowledge t hat exposure to source documents do not have substantial benefits to readers, news organizations may choose to dedicate their time and website space to more beneficial features. Conclusion News organizations are innovating ways to use Web 2.0 to deliver ne ws to readers. One current trend in this area is to provide a PDF of source documents related to a news story to the reader. This is a service that news organizations can provide rather simply on the Internet while it was cumbersome or impossible for print ed news
75 publications. It has changed the ease of access the general public has to certain public documents. It also has the potential to change the role of the journalist from reporter to recorder. While there are anecdotal examples of news organizations t hat do this, no empirical evidence shows the frequency of the practice by news organizations, nor anything about the readers who choose to read this information, nor the factors used in the decision about when to provide them, n or the effects on the reader of using them. This quantitative experiment attempted to measure t w o potential effects of source documents on the reader: recall and credibility. It found recall scores on certain questions was positively correlated with exposure to the source document. E xposure to the source document had no effect on recall on other questions. It also found that while story interest and time spent reading the material did influence credibility, exposure to the source document had no effect on the perception of credibility This study is only the beginning of research into the use of source documents by news organizations. While older research on reader effects can be applied to this branch of research to help predict effects, there is no current research that measures the use of source documents. This practice is relatively new in the industry and is likely to develop and spread. Scholarly research on the topic could help news organizations make decisions about what type of content to provide readers. More research in this area could help theorists better understand how the public consumes and uses information. The use of multimedia to provide news on the Web is on the rise, and continued study is necessary as technology develops to ensure that the best practices are employ ed. Continued study is also needed to understand the changing ways the public
76 seeks, consumes, and uses the news available to them. While the public documents used by journalists have always been available to readers, the practice of posting source documen ts to the websites of news organizations makes the process of finding that information far less burdensome for the reader. He or she can access the information with a click on a laptop instead of a drive to a courthouse or municipal authority. Perhaps beca use this era is the first in which it is a realistic possibility that the reader would review source documents, there is no prior research that shows the effects of source documents on the readers. The only past research of aid are studies that look at the factors of effects and research that examined how past technological advances changed the way news was presented. This study provides a beginning to understanding the effects of source documents on recall and the perception of credibility on the reader. O ther mass communication scholars should take this research as supporting a positive effect of exposure to source documents on recall in certain instances and no effect of source documents on perception of credibility. It should be used to understand the ov erall picture of reader recall and perception of credibility and also as a starting point for more research on how
77 Figure 5 1. Factors influencing recall and credibility in this experiment
78 APPENDIX A EXPERIMENT GRAPHICS
79 Source document Page 1
80 Source document Page 2
81 Source Document Page 3
82 APPENDIX B PO S T TEST QUESTIONS Recall 1. Were you aware of this event before reading about it here? Yes No Not sure 2. The main person involved in this news story was An employee of an Internet company A customer of an Internet company A contractor with an Internet company None of the above 3. What is the gender of the person charged with a crime in this story? Male Female None of the ab ove 4. In what area of the country did this news take place? The South The Pacific Northwest New England The Midwest None of the above 5. The person involved in this story was charged with what crime? Stealing a car Wire fraud Money launder ing Murder None of the above 6. Animals Documents Money Computers None of the above 7. Which agency investigated the crime? Boston Police U.S. Marshall Service FBI Massachus etts State Police None of the above 8. Who did the person charged with a crime contact with an offer of information? A foreign consulate A former co worker
83 A cousin A technical support employee None of the above Credibility On a scale o f 1 to 7, rate the following: 1. How much do you trust the information in this news story? 2. How accurate do you think the information in this news story is? 3. How credible do you think the information in this news story is? 4. How believable is this news story? 5. How newsworthy is this news story? 6. How fair was this news story? 7. How informational was this news story? 8. How much attention did you pay to this news story? Interest On a scale of 1 to 7, rate the following: 1. I am interested in news. 2. I am interested in reading news online or in a newspaper. 3. I am interested in news about crime. 4. I am interested in news about law enforcement investigations. 5. I am interested in news about international affairs. 6. I am interested in news about courts. 7. This news story was interesting to me. 8. I enjoyed reading this news story. 9. I wanted to know more information about this news story. Demographics 1. How old were you on your last birthday? 2. What is your gender? Male Female Prefer not to say 3. What year in school are you? Freshman Sophomore Ju nior Senior Graduate Other 4. What is your race? American Indian or Alaskan Native
84 Asian or Asian American (Includes Chinese, Japanese, Korean, Vietnamese, Filipino) Black or African American Hispanic Native Hawaiian or other Pacific Islander White Other ra ce, please specify: 5. In what college at the University of Florida are you enrolled? If you are double majoring, please answer with your primary major in mind. Agricultural and Life Sciences Design Construction and Planning Business Administration Dentistry Education Engineering Fine Arts Health and Human Performance Journalism and Communications Liberal Arts and Sciences Other 6. Are you an American Citizen? YES NO Prefer not to say The following questions are optional. They will be used only to award extra credit. 7. What is your Gatorlink name? 8. What is your Gator ID (numbers only)
85 APPENDIX C INFORMED CONSENT Please read this consent document carefully before you decide to participate in this study. Protocol Title: Examining effects of source documents on rec all and credibility of news #2011 U 1202 Purpose of the research study : The purpose of this study is to examine recall and credibility of news. What you will be asked to do in the study : You will be asked to read a news story and answer questions about t he content. You can take it from any computer with Internet access. Time required : 15 minutes Risks and Benefits : There are no anticipated risks involved with this study. We do not anticipate that you will benefit directly by participating in this study. C ompensation : You may receive extra credit for the class in which you were recruited for your participation in this study. The extra credit will not exceed 2 percent of your final grade in the course and can only be applied toward one course. Confidentialit y : Your identity will be kept confidential to the extent provided by law. You will be asked to provide your Gator ID number and Gatorlink name to provide your information to your instructor for extra credit, but it will not be linked with your answers. Vol untary participation : Your participation in this study is completely voluntary. There is no penalty for not participating. Right to withdraw from the study : You have the right to withdraw from the study at anytime without consequence. Whom to contact if yo u have questions about the study : Megan Duncan, Department of Journalism, University of Florida, P.O. Box 118400, Gainesville, FL 32611 or email@example.com; or Cory Armstrong, Department of Journalism, University of Florida, P.O. Box 118400, Gainesville, FL 32611 or firstname.lastname@example.org Whom to contact about your rights as a research participant in the study : UFIRB Office, Box 112250, University of Florida, Gainesville, FL 32611 2250; ph 392 0433.
86 Agreement: I have read the procedure described above. I voluntarily agree to participate in the procedure. I understand that I may print this page for my own records.
87 LIST OF REFERENCES Adams, T., & Cleary, J. (2006). The Parity Paradox: Reader Response to Minority Newsroom Staffing. Mass Communication & Society 9 (1), 45 61. Armstrong, C. L., & Collins, S. J. (2009). Reaching Out: Newspaper Credibility Among Young Adult Readers. Mass Communication & Society 12(1), 97 114. Armstrong, C.L. and McAdams, M. J. (2011 motivations for blog use, Atlantic Journal of Communication, 19 (2): 113 128. Armstrong, C. L., & Nelson, M. R. (2005). How Newspaper Sources Trigger Gender Stereotypes. Journalism & Mass Communication Quarterly 82 ( 4), 820 837. Arpan, L. M. (2009). The Effects of Exemplification on Perceptions of News Credibility. Mass Communication & Society, 12(3), 249 270. Benkler, Y. (2011). The real significance of WikiLeaks. The American Prospect 22 (5), 31. Berry, S. J. ( 2009). Watchdog journalism: The art of investigative reporting New York: Oxford University Press. Bleyen, V., & Van Hove, L. (2010). To Bundle or Not to Bundle? How Western European Newspapers Package Their Online Content. Journal of Media Economics 23( 3), 117 142. Boczkowski, P. J. (2009). Rethinking Hard and Soft News Production: From Common Ground to Divergent Paths. Journal of Communication 59(1), 98 116. Boyer, J. H. (1976). 'Focus and dialogue' urged as substitute for inverted pyramid. Journali sm Educator 31(2), 57 60. Bray, H. (2010, October 6). Akami employee charged with trying to sell secrets to foreign government. The Boston Globe. Retrieved from: http://www .boston.com/business/ticker/2010/10/akamai_employee.html Burgoon, J. K., Bernstein, J. M., & Burgoon, M. (1983). Public and Journalist Perceptions of Newspaper Functions. Newspaper Research Journal 5(1), 77 89. Burgoon, J. K., & Burgoon, M. (1981). Th e Functions of the Daily Newspaper. Newspaper Research Journal 2(4), 29 39. Burgoon, J. K., Burgoon, M., & Wilkinson, M. (1981). Writing Style as Predictor of Newspaper Readership, Satisfaction and Image. Journalism Quarterly 58(2), 225 231.
88 Callison, C ., Gibson, R., & Zillmann, D. (2009). How to Report Quantitative Information in News Stories. Newspaper Research Journal 30(2), 43 55. Campbell, J. (2007). 8 simple rules to getting the documents you need. Quill 95(2), 38. Chaiken, S. (1980). Heuristic versus systematic information processing and the use of source versus message cues in persuasion Journal of Personality and Social Psychology. 39(5), 752 766. Chaiken, S. & Maheswaran, D. Heuristic processing can bias systematic processing: Effects of s ource credibility, argument ambiguity, and task importance on attitude judgment. Journal of Personality and Social Psychology, 66(3), 460 473. Chermak, S. (1995) "Image control: how police affect the presentation of crime news", American Journal of Polic e Vol. 14 Iss: 2, pp.21 43. Chung, D. S. (2009). How Readers Perceive Journalists' Functions At Online Community Newspapers. Newspaper Research Journal 30(1), 72 80. Chung, C., Nam, Y., & Stefanone, M. A. (2012). Exploring Online News Credibility: Th e Relative Influence of Traditional and Technological Factors. Journal Of Computer Mediated Communication 17 (2), 171 186. Coleman, R., & Thorson, E. (2002). The effects of news stories that put crime and violence into context: Testing the public health m odel of reporting. Journal of Health Communication 7(5), 401 425. DeMarban, A. (2001, May 5). 51 bottles of booze lead to Kobuk arrest. The Arctic Sounder. Retrieved from: http://www.thearcticsounder.com/article/111851_bottles_of_booze_lead_to_kobu k_arrest Flowerday, T., Schraw, G., & Stevens, J. (2004). The role of choice and interest in reader engagement. Journal Of Experimental Education 72(2), 93 11 4. News, Newsweek and Time, New York: Pantheon. Gaziano, C., & McGrath, K. (1986). Measuring the Concept of Credibility Journalism Quarterly 63(3), 451 462. George, D., Ma llery, P. (2003). SPSS for Windows step by step: a simple guide and reference, 11.0 update. 4th ed. Boston: Allyn and Bacon.
89 Griffin, J. L., & Stevenson, R. L. (1992). Influence of Text & Graphics in Increasing Understanding of Foreign News Content. News paper Research Journal 13(1), 84 99. Goldsborough, R. (2010). The New Age of Investigative Journalism?. Teacher Librarian 38(2), 57. DocumentCloud. MediaShift. Retrieved from: http://www.pbs.org/idealab/2011/05/much ado about obamas birth certificate on documentcloud129.html Johnson, K. A., & Wiedenbeck, S. (2009). E nhancing perceived credibility of citizen journalism web sites. Journalism & Mass Communication Quarterly 86(2), 332 348. Kasoma, T., & Maier, S. (2005). Information as Good as its Source: Source Diversity and Accuracy at Nine Daily U.S. Newspapers. Con ference Papers -International Communication Association 1 24. Keenan, J. (2010, August 11). Why raw data sites need journalism. The Guardian. Retrieved from: http://www.guardian.co.uk/commentisfree/libertycentral/2010/aug/11/raw data journalism wikileaks Knobloch, S., Patzig, G., Mende, A., & Hastall, M. (2004). Affective News Effects of Discourse Structure in Narratives on Suspense, Curios ity, and Enjoyment While Reading News and Novels. Communication Research 31(3), 259 287. doi:10.1177/0093650203261517 Kiousis, S. (2001). Public Trust or Mistrust? Perceptions of Media Credibility in the Information Age. Mass Communication & Society 4(4 ), 381 403. Majors, D. & McNulty, T. (2011, March 4). Mistrial frustrates jurors in Orie corruption trial. The Pittsburgh Post Gazette. Retrieved from: http://www.post gazette.com/pg/110 63/1129569 455.stm Manning, P. (2001). News and news sources: A critical introduction London: Sage Publications. Meyer, P. (1988). Defining and Measuring Credibility Of Newspapers: Developing an Index. Journalism Quarterly 65(3), 567 588.
90 Miller, The Wall Street Journal. http://online.wsj.com/article/SB1000142405297020373350457702248174609563 6.html Mondak, Jeffrey J. 1990. "Perceived Legitimacy of Supreme Court Deci sions: Three Functions of Source Credibility." Political Behavior 12 (4): 363 84. Nash, S. (1984). Slides teach class to use documents as news sources. Journalism Educator 39(3), 29 30. Noack, D. (2001). Electronic Newspapers. Link Up 18(5), 30. Pipps, V., Walter, H., Endres, K., & Tabatcher, P. (2009). Information Recall of Internet News: Does Design Make a Difference? A Pilot Study. Journal of Magazine & New Media Research 11(1), 1 20. Police Records. (2008). News Media & the Law 32(1), 1 23 Pottker, H. (2003). News and its communicative quality: the inverted pyramid -when and why did it appear?. Journalism Studies 4(4), 501. Ramaprasad, J. (1991). Informational Graphics in Newspapers. Newspaper Research Journal 12(3), 92 103. Rich, C. (2010). Writing and reporting news: A coaching method Belmont: Thomson Wadsworth. Reich, Z. (2011). Source Credibility and Journalism. Journalism Practice 5(1), 51 67. Robinson, J. P., & Levy, M. R. (1986). Interpersonal Communication and News Comprehe nsion. Public Opinion Quarterly 50(2), 160. Schraw, G. and Lehman, S. (2001). Situational Interest: A Review of the Literature and Directions for Future Research. Educational Psychology Review ,13(1), 23 52. Smith, E. J., & Hajash, D. J. (1988). Informat ional Graphics in 30 Daily Newspapers. Journalism Quarterly 65(3), 714 718. http://www.niemanlab.org/2010/05/why does the bbc want to send its readers away the value of linking/ Sundar, S. S. and Nass, C. (2001), Conceptualizing sources in online news. Journal of Communication 51: 52 72.
91 Tormala, Z. L., Briol, P., & Petty, R. E. (2007). Multiple roles for source credibility Social Cognition 25(4), 536 552. Tremayne, M. (2004). The Web of context: Applying network theory to t he use of hyperlinks in journalism on the Web. Journalism & Mass Communication Quarterly 81(2), 237 253. Trumbo, C. W. (1999), Heuristic Systematic Information Processing and Risk Judgment. Risk Analysis 19: 391 400. Utt, S. H., & Pasternack, S. (1993 ). Infographics today: Using qualitative devices to display quantitative information. Newspaper Research Journal 14(3/4), 146 157. http://www.usatoday.com/sports/college/football/story/2011 11 07/penn state attorney general pdf/51112998/1 Wanta, W., & Gao, D. (1994). Young Readers and the Newspaper: Information Recall and Perceived Enjoyment, Readability and Attractiveness. Journalism Quarterly 71(4), 926 936. Wanta, W., & Remy, J. (1995). Information recall of 4 elements among young newspaper readers. Newspaper Research Journal 16(2), 112 123. Washington Pos t, The. (2011) http://www.washingtonpost.com/wp srv/sports/documents/sandusky grand jury report11052011.html Whitehouse, A. A., Andsager, J., & White, H. (1991). Newspaper column readers' gender bias: Perceived interest and credibility. Journalism Quarterly 68(4), 709 718. Wilson, E. J., & Sherrell, D. L. (1993). Source Effects in Communication and Persuasion Research: A Meta Analysis of Effect Size. Journal of the Academy of Marketing Science 21(2), 101. Wolf, R., & Grotta, G. L. (1985). Images: A Question of Readership. Newspaper Research Journal 6(2), 30 36. Wise, K., Bolls, P., Myers, J., & Sternadori, M. (2009). When Words Collide Online: How Writing Style and Video Intensity Affect Cognitive Processing of Online News. Journal of Broadcasting & Electronic Media 53(4), 532 546. Zelizer, B. (1993). Journalists as Interpretive Communities. Critical Stud ies in Mass Communication 10(3), 219.
92 Zerba, A. (2008). Narrative Storytelling: Putting the Story Back In Hard News to Engage Young Audiences. Newspaper Research Journal 29(3), 94 102.
93 BIOGRAPHICAL SKETCH Megan Duncan received her Master of Arts in Mas s Communication from the University of Florida in s pring 2012 She received her Bachelor of Arts in Journalism from Point Park University in 2005