1
Multimedia Features and Newspaper Websites: A 2007 Content Analysis of Daily Newspapers
The buzzword in journalism today–actually, for the past decade–is "convergence,” the merging of print and broadcast through the medium of the internet. For most newspapers, the first phase of newspapers' transition to the web in the 1990s consisted largely of dumping their print content online, but that model is changing. While a decade ago only high-traffic newspaper sites such as USAToday.com were posting occasional video on their websites, more and more publications are now embracing convergence, incorporating sound and video from outlets such as the Associated Press, from their own reporters and multimedia staff, and even from users. This foray into multimedia, in addition to creating archives and interactive features not possible before in either broadcast or print, has dramatically changed print journalism, broadcast journalism and journalism education.
Journalism scholars have accordingly jumped on this bandwagon, studying ways convergence is being implemented, the impact of convergence on the journalism produced, the effect of convergence on business models and the curricular implications involved in preparing future journalists for work in a convergence landscape. However, in spite of this relatively new scholarly attention to the field, there are still many areas which have not been examined or not examined in great depth. One of the biggest gaps is a lack of broad-based studies of newspaper websites. To help meet that need for such studies, we conducted a detailed analysis of more than 350 daily and 350 weekly newspaper websites in the summer of 2007. Using a matrix of 24 different features–separated into the categories of multimedia, interactivity and distribution methods–our group attempted to take a snapshot of newspaper convergence practices. Having already presented our findings on interactivity at a conference late last year, we will focus this paper on multimedia, perhaps the most interesting part of that 2007 convergence snapshot. Our goal is to give both educators and practitioners a better understanding of what is being done, and with that information, teachers and journalists can better know what can, could or should be done in their own classrooms and newsrooms.
Literature Review
While our study addresses the “who” and “what” of convergence, it does not answer the “how,” how these newspapers are going about the process of creating multimedia for their websites. That question has been answered, to some degree, by studies that have taken a more ethnographic approach to subject matter, such as those by Dibean and Garrison[1] whose 2001 study took an in-depth look at how six newspapers were incorporating new web capabilities on their sites. Some of the most noted case studies have looked at Tampa (Florida) Tribune and WFLA merger, in which print and broadcast reporters were literally placed side by side in the same office.[2] Huang and his group of researchers produced a report in 2004 about how the convergence model in Tampa did not seem to negatively impact the quality of journalism there[3], while Garrison and Dupagne focused their case study on job skills and the newsroom culture.[4] Earlier studies focused on newspaper transitions to the web in the mid and late 1990s, as newspapers quickly gravitated to the web and began were experimenting with sound and video files on their sites. Some early examinations of this process included studies by Martin[5] and Singer[6].
While very few studies have looked at more than a handful of newspapers at a time, one of the first more widely-based studies was Peng, Tham and Xiaoming’s 1997 survey of 247 publishers and content analysis of 80 newspaper websites.[7] Also of note is Greer and Mensing’s 1997-2003 study, which not only looked a large number of newspapers (83), but also involved an analysis of those websites over several years.[8] There have been a few other notable exceptions, most recently Murley and Carroll, who conducted a survey of college publication advisers about multimedia and interactivity on college newspaper websites.[9] But, while there has been some other research–most notably, the Bivings Group’s 2006 report on the features of the websites of the 100 newspapers with the largest circulations[10]–there has not been a comprehensive look at the entire gamut of daily newspaper websites. Much of the literature has focused on single issues, such as internet coverage of an issue such as war or protest (Dimitrova and Neznanski in 2006[11], Nah, Veenstra and Shah in 2006[12]), single convergent elements such as nonlinear storytelling[13] or, also using case study approaches, graphic design and the web[14] and the impact of institutional history and the organization of the newsroom on the ways in which multimedia and interactivity are incorporated into newspaper websites.[15] Some other studies have focused on the user experience and how or why users select web news coverage[16], [17] and the impact of multimedia on the user[18].
Much other scholarship has focused on how journalism educators to need to rethink the journalism curriculum in light of all of the technological changes. Edgar Huang and his research colleagues perhaps put it best when they wrote, “Media convergence, as a trend that is gradually shaping the landscape of the media industry in the new century, has called into question the conventional journalism school practice of having separate tracks–print, broadcast, etc. Journalism educators around the country also are trying to figure out what they should do, if anything, to better prepare students for the converged media.”[19] Part of the problem, of course, is that no one–including the practitioners in the field–knows for sure what direction convergence will take. As Quill editor Jeff Mohl points out, “How can teachers predict where the profession is headed if the profession hasn’t figured it out yet?”[20]
There is little doubt among the faculty, however, that something must be done. In one survey, the vast majority of faculty agreed that graduates should know how to do web searches (100 percent agreed), how to prepare information for the Web (92 percent), that faculty be able to teach Web-based courses such as Web page design (77 percent) and that journalism and mass communication programs should keep up with technology changes in the media industry (99 percent).[21]
Just how these technologies should be incorporated into the curriculum, though, is a matter of debate. Kraeplin and Criado differentiate between the two main styles of incorporation: multidisciplinary and interdisciplinary. The former combines aspects of different disciplines, but much preferred is interdisciplinary, in which the components of the different fields are truly integrated.[22] Huesca similarly identifies two different ways that journalism education has responded to the problem and opportunity of convergence. One approach involves a complete reinventing of the entire curriculum in response to the radical nature of hypertext, which has thrown traditional definitions of readers and authors out the window and mandates presentation styles which “allow readers to construct their own versions of reality, rather than simply reading a reporter’s version of reality.” The second approach takes a more conservative tack, accommodating some changes in course content to address the technological changes but largely preserving the core values stressed in journalism programs for decades.[23] For Quinn, two of the key core values that become even more important in this technological age are ethics and information retrieval.[24]
Methodology
After combing through this literature about multimedia and interactive features of convergent media websites, in addition to reviewing numerous newspaper websites, our group created and refined a list of 24 items. These items fit into three main categories: distribution capabilities (such as mobile cell phone alerts of breaking news events), interactive features (such as reader polls) and multimedia (such as audio clips).
Random Sample
After creating this 24-item matrix on Excel spreadsheets, we then selected the newspapers to analyze. To provide a reliable and thorough picture of the current state of daily newspaper websites, we created a random sample of daily newspapers in the United States. We began with the 1,437 daily newspapers listed in the 2007 Editor and Publisher Yearbook and, using a random number generator to select the first number, chose every fourth newspaper, yielding a total of 360 U.S. newspapers to study. This sample size was enough to achieve a 95% confidence rating an error margin of +/- 4.5% (while a +/-3% error margin might have been more ideal, such a study would have entailed analyzing over 600 newspapers, too many for our available resources and time).
Inter-rater reliability
Because we were a four-person team with each person analyzing his or her portion of the websites individually, inter-rater reliability was critical. To ensure we were all finding and tabulating the features consistently, our team conducted a norming session, coding a handful of sites as a group and discussing any differences or omissions in how we each coded the websites. After achieving consensus, we each individually coded the data on the 24-item matrix for the first 10 newspaper websites of our random sample. A comparison of the codings revealed an inter-rater reliability of 94%. The most typical coding error was the omission of a feature buried deep inside of the website on an article that one or more other coders missed. After more discussion, the team members analyzed the remaining 351 websites in the sample group, between July 5 and July 31, 2007.
Size breakdowns
Our hypothesis was that size would play an important factor in the presence of multimedia features on newspaper websites. Consequently, when working with the Editor and Publisher list, we entered the circulation figures of each newspaper. After looking at the total numbers and the size breakdowns used by E&P, we decided to create five different demographic categories:
Less than 10,000 circulation: 144 newspapers
10,001-25,000: 105 newspapers
25,001-50,000: 49 newspapers
50,001-100,000: 32 newspapers
Over 100,000: 31 newspapers
While having a more even distribution in each demographic group might have been desirable, we did not feel the need to break down the smaller circulation categories further, as most newspapers under 10,000 circulation likely face the same problems with lack of personnel, technology, and financial resources to invest in adding multimedia features to their websites.
Results
For the purposes of this paper, we define multimedia features as those that are not text-based or cannot be replicated in print. For example, comments fields, forums and reader blogs are clearly items that cannot be found in the newspaper delivered to the doorstep, but still are very much text-based. On the other hand, a video clip (including a video blog) was categorized as multimedia because it is not text-based and involves different technologies/software for creation, distribution and reception. Using this distinction between multimedia and interactivity, seven of the 24 features studied fell into the category of multimedia: Video, Audio, Photo Gallery, Audio Slideshow, Podcast, Games, and Interactive Graphics. An interactive graphic, because it is both highly interactive and unable to be truly reproduced in print, was the one element that fit into both the multimedia and interactive categories.
Video
After analyzing the first several websites, we realized we would need to clearly distinguish between sites that had video created in-house and video which was created by the Associated Press or by a TV station partnering with the newspaper. A vast majority of the smaller papers had video, but it was typically AP video. Still, 51% of the newspapers under 10,000 produced their own video interviews, video blogs or video packages with voiceovers. Nearly all of the newspapers over 100,000 (97%) had their own video somewhere on their websites. Collectively, 65% of daily newspapers had some form of self-produced video on their websites.
Audio
Audio included any audio blogs, sound clips from interviews or downloaded music examples. Almost one third of newspapers under 10,000 were using sound media; nearly twice as many newspapers over 100,000 used audio. Overall, 52% of newspapers had some form of audio on their websites, a smaller percentage than video.
Podcasts
Podcasts seem to have not caught on yet among many newspapers. At smaller newspapers, the use of podcasts was virtually nonexistent, according to our study. Even at the larger papers, podcasts were by far the exception and not the rule, which is a bit unexpected, given how ubiquitous the use of audio clips was.
Interactive Graphics
Newspapers such as USA Today have set the bar for the usefulness, functionality and attractiveness of interactive graphics. These graphics, almost always created in Flash, have the ability to explain complex information in a visual, easy-to-understand manner. Often employing timelines and maps and illustrations with rollover/popups and animation and sometimes other media such as sound and video, these pieces often require teams of graphic artists and web programmers. Such resources are naturally not found at the smallest newspapers, although a few of the papers under 25,000 still experimented with interactive graphics. Even at the largest papers, though, only about 7% had interactive graphics on their websites when those websites were analyzed for this study.
Photo Galleries
By far the most common multimedia element was the photo gallery. As a whole, 70% of daily newspaper websites had photo galleries. These photo galleries usually took one of two different forms: a set of thumbnails with hyperlinks to larger individual images (on another page or in a popup window), or a more interactive slideshow which allowed users to hit forward or back buttons to go through a collection of photos. Because the former technology is not complex--often found even on first-generation websites--at even the smallest two categories of newspapers, two-thirds had photo galleries. At the larger newspapers, that fraction rose to approximately six-sevenths.
Audio Slideshows
Similar to the Photo Galleries, Audio Slideshows typically include a set of photographs. While the photo galleries are usually more interactive, allowing the user to move back and forth among the images, the audio slideshows are typically linear, with the images progressing by themselves, often linked with the sound. In some cases, the audio is simply background music that plays as the browser window moves from photo to photo. But, in numerous cases these slideshows also had interview clips, ambient noise, or, in rare cases, voiceovers. As one would expect, the largest papers had the highest percentage using slideshows, at 74%. But, a very significant amount, 40%, of those papers at the other end of the spectrum also had audio slideshows. The majority of these smaller papers with audio slideshows were using the Soundslides program, which was freeware (now low-cost software) which makes producing audio slideshows very easy, and this low cost and ease of use no doubt helped editors, reporters and photographers at these smaller papers still produce this form of multimedia.
Games
A late addition to our coding scheme was Games. These games took many varieties, but the most typical were interactive Flash games with card, word or number games. By and large, these were clearly produced/contracted externally, the same way that Sodoku and crossword puzzles are provided to print newspapers by services or syndicates. In some cases, at a couple of the larger papers, there were interactive quizzes and games that were produced in-house. Because the technology to include these games on websites through the services sometimes is as simple as providing a hyperlink, the size of the newspaper did not make a great deal of difference in the frequency of the instance of games.
Conclusion
Clearly, size made a very distinct difference in the presence of multimedia elements on the newspaper websites. The larger the circulation of the newspaper was, the greater the chance that it contained multimedia. Almost across the board, the largest newspapers had the most multimedia elements. And, the more difficult, complex or expensive producing the multimedia element was, the greater the impact of size was. For elements such as audio, video, audio slideshows and interactive graphics, the largest newspapers were often about twice as likely or more to have the multimedia element than the smallest newspapers.
While the influence of size did not surprise us, the low frequency of podcasts and interactive graphics among the largest papers did. We especially expected podcasts to be more popular; it would be interesting to revisit the use of podcasts by these newspapers in two or three years to see if they catch on more. As for interactive graphics, it was not surprising for the smaller newspapers to not have the expertise or money to support Flash interactives, but for only one out of 15 of the largest newspapers to have interactive graphics--that was unexpected.
Limitations of study
One of the limitations of the study is that it was not a complete analysis of all 1,437 websites. Although a scientific random sampling method was used, not having any margin of error would have been better, and a complete study would have been very useful in reexamining the same newspapers in a few years to gauge how much the field continues to change. Of course, such a study would have been very, very time-consuming, and the data obtained is still very reliable and useful.
Perhaps the biggest limitation is the study is, as mentioned in the introduction, only a snapshot of each newspaper on the day its website was examined. So, while a newspaper may create wonderful interactive Flash graphics on a weekly basis to accompany a lead story, if the interactive graphics were not placed in a special section and were either removed or the links to them removed by the day the site was studied, the newspaper was coded as not having interactive graphics, even though the newspaper had dedicated resources to original interactives on a consistent basis. The same is true for other multimedia as well, and so we would not be surprised if the actual instances of these features were higher than the data might indicate. Again, it was simply not practical to review the 700+ daily and weekly newspapers examined in our project on multiple days. While employing a survey methodology might have prevented such omissions, that approach would have been fraught with other problems, not the least of which could be a skewing of data because those responding to the survey would be the most likely to be interested in and showcasing multimedia. Using a content analysis ensured that 100% of the newspapers sampled provided data for the study.