News Photographer, April 2003 v58 i4 pS4(8)

Professionalizing online news photo presentations

(VCQ/Spring 2003)

Edgar Shaohua Huang

Online dailies can win over both photo-loving readers and text-oriented readers by controlling number, size, and location of photos, using linkable photos and photo galleries, and maintaining consistency of photo presentations.

ABSTRACT

With the advent of online dailies, the role of news photography is undergoing fundamental changes. Many online dailies use small photos, use as few photos as possible, or use no photos at all. Can news photos attract reader attention to online newspapers as they do in print newspapers? Do readers expect photos online? This study provides guidance on how to present photos with reader expectations taken into consideration. A content analysis of Yahoo's most popular American online dailies indicated that close to 90% carried at least one photo on the site, though the average diagonal of a home-page lead photo on a news site was less than three inches. A survey of more than 450 serious readers determined that most say they pay attention to news photos when they read print dailies and that they would like to see photos on the home pages of their most frequently read online dailies. The study shows that online dailies need visuals, most readers expect photos to be eye-catchers, and an online daily can win over both photo-loving readers and text-oriented readers by appropriately controlling the number, size, and, especially, location of lead photos.

**********

TRADITIONALLY, NEWS PHOTOS HAVE BEEN USED AS EYE-CATCHERS on the front page of a newspaper. Harrower (1998) teaches journalism students in his popular textbook The Newspaper Designer's Handbook: "If you want to hook passing readers, photos are even more valuable than text" (p. 93). He repeatedly emphasizes the importance of placing the four basic elements in newspaper design in the correct order: photo, cutline, headline and body text. Garcia and Stark (1991), as well as Bain and Weaver (1979), point out in their studies that photos are the most-looked-at items in newspapers; stories with pictures command better readership and hold the reader's attention longer than stories without pictures.

A cursory look at online dailies, however, reveals that photos--if there are any-are much smaller and are sometimes placed deep inside a story. Some online dailies do not include photos. Media industry studies suggest a shaky future for news photography in online dailies. According to a survey by cPulse, "one in five visitors to news websites leaves unhappy" and "photos and graphics on news sites have no impact on user satisfaction" (Adweek.com, 2000). Another survey also found that online daily readers either ignore photos and graphics or read them after they have read text (Wired.com, 2000).

Can lead photos (1) catch readers' eyes online the way they do in print? Do readers expect online news photos to be eye-catching? Do readers want to see photos in online dailies at all? What can media do to serve readers better?

After roughly a decade of development, online newspapers have become the top choice for local news and information for Internet users in the United States (Newspaper Association of America, 2002). "It just seems inevitable that someday digital delivery of indepth, personalized information - including text, audio and video -- to electronic devices will supplant the trucking of heavy physical loads door-to-door" (Gates, 2002). With such an increase in the number of online readers, online daily editors need to understand what readers want and do not want. Vargo (1999) argued: "Newspapers need to work with readers to find out what they really want in an electronic paper. Only then will they know how to make the best use of this new medium." Mings (1997) also said: "Claims about the advantages of moving newspapers online are grounded in assumptions about the online audience. However, for such claims to be valid - and any related hopes for newspapers' economic viability to be realized - an understanding of the audi ence is necessary."

Online newspapers still are developing, and much is expected to occur in the years to come. The goal of this study was to provide guidance to online daily editors on how to present photos, taking reader expectations into consideration.

The study had two stages. First, a content analysis was conducted to find out how online news photos were presented. Second, a survey was taken to find out what readers' expectations were for such presentations. Actual practice and reader expectations were then compared to find out how online dailies can better serve the readers.

LITERATURE REVIEW

Few researchers have focused on online news readers, and even fewer on online news photography. Nevertheless, existing research on online journalism was useful to this study in suggesting both research questions and methods to investigate them.

Online journalism research is new compared to research about traditional media. Sampling American online dailies for content analysis purposes has been particularly problematic. For example, the number of online newspapers cited by different authoritative Web sites can vary drastically. (2) Determining a reliable sample frame requires caution and careful attention to the research purpose. Researchers have used methods based on Editor & Publisher Web site's online newspaper section (Gubman & Greer, 1997), the Yahoo search engine (Kamerer & Bressers (1998), U.s. journalism Web sites "that contained original material on general news topics, updated daily, and which were national in scope" (Tremayne, 1999), "full-service, general circulation U.S. daily online newspapers that were listed in American Journalism Review's NewsLink" (Tankard, 1998); and on geographical region (Alexander, 1997).

Kamerer & Bressers' 1998 study created a profile of sampled online newspapers in terms of photo uses. They found that in November of 1997, 63% of online newspapers carried local photos, while 25% used national photos. Only 6% had local photo links. Their study found that 77% of online newspapers used photos on their sites, but fewer than 10% percent of the sites offered animated graphics, audio clips, video clips or tables or graphs.

Mings' 1997 study on readers' preferences for use of graphics in online newspaper sites helped form some of the research questions in this study. Ming found that the experiment participants spent the majority of their online time viewing textual, rather than audio-visual or graphic materials. However, her study did not examine why that was the case.

To some extent, the eyetracking study by Lewenstein, DeVigal, & Jacobson (2000) confirmed Mings' finding. By using sophisticated camera equipment to capture readers' first three fixation-clusters, a term that describes usually overlapping glances that include the eye's precise point of focus and the immediately surrounding area, Lewenstein et al. (2000) found that people reading news on Internet sites tend to focus on the article text first, looking at photos and graphics afterwards. Those who read newspapers and magazine do the opposite: look at illustrations first then read the text. Lewenstein et al. (2000) acknowledged: "Because our research did not evaluate the quality of graphics or text displayed on the pages studied, it's impossible to determine what prompted the subjects to prefer text as starting points."

According to Jacobson (2000), one of the researchers on the team, all results in the Lewenstein study were based on tests of only 67 people drawn from the limited universe of "frequent Internet news readers" rather than the Internet-using public at large. That is, they would limit their conclusions only to that narrowly focused group. Of those 67 people, only 14 were examined for their behavior in home page attention. Of those 14, fewer than seven pages viewed by these subjects contained a single photo. Again, this number seems stunningly small to make the assumption that text is preferred to photos, considering how few photos were available. Subjects viewed sites familiar to them. But virtually all the graphics at those sites were static and never changing. Jacobson further explained, of the three sample pages used in the study, only two had photos and each of these pages was limited to a single photo. Neither of these photos contained an image of a person. Again, Jacobson reasoned that images of people are more likely to draw attention than images of things, and that multiple images are more likely to attract attention than a single image.

DeVigal (2000), another researcher on the team, wrote in an article reviewing the findings: "Does this mean that photos or graphics do not have a place on the homepage? Hardly." He continued: With 64 percent of all available photos having fixations, photos clearly are essential. In fact, many online news organizations' log files show that photo galleries are among their most popular destinations. In the Page Elements sections of the study, page elements for all pages were ranked by volume of fixations. Photos are third to article texts and briefs, which were looked at 92% and 82% of the time, respectively. The conclusions of these early findings are not meant to encourage text-only presentation of websites, nor do the findings reveal that online readers prefer text to images to get their news. (3)

Most studies found on online journalism readers involved students drawn as convenience samples (Schierhorn, 1998; Vargo, 1998; Chamberlin, 1998; Mings, 1997) and the researchers primarily used experiment methods. The findings from those studies give us a glimpse into a particular group of readers' ways of consuming online news. We cannot assume that college students' online news reading preferences are generalizable to a larger population.

Research Questions

The general research question was, "What can media do to better serve readers in terms of photo presentation?" It is represented by the following specific research questions:

* Do online dailies try to catch readers' attention with lead photos by size, placement and other design elements?

* Do readers pay attention to online news photos or expect lead photos to be eye-catchers in online dailies? If yes, what are their preferences for viewing online photos? If not, what are the major concerns?

* What can media do better by taking into consideration readers' expectations?

METHODS

Content analysis

Borrowing sampling ideas from Kamerer and Bressers' 1998 study, the author focused the content analysis of photo presentations on the 56 most popular American online dailies listed under the category of Most Popular Sites at the Yahoo Newspaper site (http://dir.yahoo.com/News_and_Media/Newspapers/). To some extent, these mainstream online dailies represent the development of new technology in news media and they set up standards for numerous less popular local online dailies with local interests and with a smaller circulation. Because of the potential influence these popular online dailies have on other online dailies, the author believes that the conclusions from this non-probability sample can be applied suggestively to a broader universe of online dailies. Because of the digital nature of online news sites, the suggestions from this study can also be applied to those online dailies run by TV stations and broadcast stations. However, the focus of the content analysis based on this non-probability sample was not to extend the conclusions to a bigger population, but rather to compare results from the content analysis with readers' expectations.

The data were collected five times from each of the 56 online dailies over a period of three months in order to create a constructed week sample. Each daily had the chance of being observed at different times from Monday to Friday. A daily observation journal was written. The observation of "photo presentations" included the following six indices:

* number of photos

* sizes of photos (4)

* time for loading pages with or without photos (5)

* locations of photos

* photo linkability, and photo gallery (6)

* consistency of photo presentations over time.

A photo was defined as a still image generated by a camera. Photo illustrations with combined graphics or text(s) from different sources were not counted as photos.

A pilot study was done before the actual coding took place. Four online daily sites from the Yahoo site mentioned above were selected for pre-coding practice. Two coders coded all pages. Intercoder reliability, computed by using Scott's Pi, was 0.93. All coding was done in a home environment by using an internal 56K modem installed on the author's computer with a Pentium 3 processor regardless of day or night. This effort was made to simulate the situations that a reader could encounter when reading online dailies at home. Each time Web sites were to be observed, both memory cache and disk cache were cleared beforehand so as to avoid possible imprecision in counting photo-loading speed.

Numerical data in each category such as photo loading speed and photo size collected from the five times of observation over three months, were averaged. The mode of each of the categorical data was chosen to represent a typical situation of each category. However, variations were also recorded for consistency analysis. The unit of analysis was each online daily.

Survey

In order to let readers get a sense of comparative photo size, location, page-loading speed, etc., this study used an online survey method. The author designed and constructed the survey questionnaire and tested it through a pilot study. Examples used in the study were taken from actual online dailies. The author removed all names of the online dailies and advertisements on each real-life web page to minimize any distractions and potential bias caused by readers' recognition with certain dailies or products. Although questions were close-ended, a text box was included under many questions for readers to explain their answers if they wished. Such open-ended answers, which served as quasi-interviews, helped the author better understand the readers' choices and better explain the quantitative data. The questionnaire was placed on a university Web site. The data were collected by dk3.com and directly sent to the author's email account.

The author recruited online daily readers from the Message Boards, Bulletin Boards, Discussion Boards, Forums, Talk, or Opinions links carried in those dailies. It could be reasonably presumed that the opinion-givers who appeared in such sites read online dailies; their reading activities were current; they cared about what was presented inside the dailies; and they knew how to use the Internet. Most important of all, the author could access many of these readers via email addresses published in the discussion.

Thirty of the 50 dailies in the content analysis included discussion links. Participants with an email address from each discussion board link were counted. A discussion board usually carried discussions from multiple days. If a participant appeared multiple times, s/he was counted as one entry. Thirty readers were randomly recruited from each of those 30 online dailies. If an email address was later found invalid, a substitute would be added from the rest of the discussion participants. In total, 900 individual survey invitation emails were sent successfully. The unit of analysis was each survey respondent. It is important to note that the findings from this study could be biased toward those readers who participated in online discussions and volunteered their email addresses. The trade-off is that opinions were obtained from readers who cared enough about online photo presentations to participate in an online discussion. The survey was conducted in mid-2001.

FINDINGS AND DISCUSSION

In total, 466 (52%) completed survey questionnaires were received. Demographically, a typical online daily reader was an adult male under the age of 50 with a college degree. Among all the readers, 32% were females and 68% were males. The majority of the readers (88%) surfed the web every day. Female readers surfed the web as frequently as male readers did. Of all the readers, 56% read online dailies every day, and 40% of those online readers read print dailies every day. Male readers and female readers indicated no difference in terms of traditional newspaper reading frequency, but male readers read online dailies more frequently than female readers did in general. Most readers (75%) read more than one online daily.

Findings and discussion from both the content analysis and survey follow.

1. Do online dailies try to catch readers' attention with lead photos by size, placement and other design elements?

In the content analysis, 50 (89%) out of the 56 online dailies surveyed carried at least one photo somewhere on the site. (7) In other words, most online daily editors were aware of the importance of presenting news visually and did not abandon photos completely in online dailies.

At the time of study, a typical online news daily had these characteristics: Forty-two percent of the stories inside the daily carried at least one photo. The home page carried one linkable photo of 2.7 inches long diagonally (8) raw-wrapped on the right of the headline and took 40 seconds to load with a 56K internal model on a Pentium 3 processor. Story pages that carried a photo used one 3.1-inches-long unlinkable photo using the armpit design (though design varied from story to story and from day to day). Although the daily rarely used photo galleries, if it did, the photo was 5.6 inches long diagonally. Few online dailies used photo galleries every day. If a daily carried a photo gallery, photo size averaged 5.6 inches diagonally. The daily performed consistently over time.

Here are further details.

Number

The number of photos used in a newspaper reflects, from one perspective, to what degree that newspaper attempts to show news visually. The Detroit News (80%), The Oakland Tribune (83%), and San Diego Union-Tribune (83%) carried most stories illustrated with photo(s). The vast majority (92%) of the online dailies had at least one photo on their home pages. Usually (48%), only one photo was used.

In story pages, 78% of the 50 online dailies used photos. (9) The vast majority (94%) of the stories included either one or two photos. Typically (54%), there was only one. St. Petersburg Times and San Francisco Chronicle used the most photos story-wide, running an average of three photos in each story that carried photos.

Most online dailies (70%) carried photos on both their home pages and story pages.

Size

The average diagonal of a home-page lead photo was 2.7 inches. The biggest home-page lead photos were observed in The Salt Lake Tribune (5.5 inches) and Detroit Free Press (5.5 inches). The smallest home-pale lead photos were observed in The Providence Journal (1.1 inches).

The average diagonal of a story-page lead photo was 3.1 inches. The biggest story-page photos were found in The Detroit News, The Virginian-Pilot of Norfolk and St. Petersburg Times. All of them were 5.1 inches diagonally. The smallest story-page lead photos were found in San Francisco Chronicle (one inch).

About one-third of the online dailies (32%) had a photo gallery (or called "slide show" under the title such as "Today's Photos") that carried news photos from current day or current week, (11) but only three online dailies carried a photo gallery every day. The sizes of gallery photos were much bigger and were usually displayed in a separate window. Gallery photos were at least 3.1 inches long (seen in USA Today). The average diagonal of a gallery photo was 5.6 inches. The biggest gallery photo (9.5 inches) was found in the New York Post.

Time for loading pages with or without photos

The average time for loading the home page of a daily either with or without photo(s) was 35 seconds. (12) The average time for loading a home page with one photo was 40 seconds, and the counterpart for a home page with no photo was 17 seconds. The smaller the lead photo (R=0.7, p<0.01) and the fewer the photos (R=0.27, p<0.05) on a home page, the less time it takes to load the home page. It also was found that a two-to-three-inch-diagonal photo could often be loaded at the same time when the text is loaded. A five-inch-diagonal photo was usually completely loaded three to four seconds after the text was seen with a 56K modem.

The slow speed of loading large photos is still a major technological bottleneck for online dailies. The content analysis showed that many online dailies tried different ways to diminish the effect of slow photo loading speed. Some dailies used no photo either on the home page or on the story pages. (13) If photos were used on home pages or stories pages, sizes were small, photos were few, and sometimes photos were buried inside stories. (14) Some dailies used photos for the second and third headlines on the home page so that readers could see text first and photos later. (15) Under these conditions, photos, if any, in those online dailies could hardly be expected to catch readers', especially photo-loving readers', eyes. Readers had to rely heavily on the headline and deck information to get a sense of the story.

Location

In print newspapers, lead photos have often been placed on top of their headlines. Readers decide whether they will continue to read the accompanying story depending on how interesting the lead photo is to them. Apart from the "above-headline" design, a lead photo often is placed on the right side of the headline, with part of the photo inside the text (called "raw-wrap" design), or directly below a headline and side by side with text (called "armpit" design). (16) Designers rarely place a lead photo between the headline and text, or totally sink a lead photo to the middle of the text or below the text. Because the homepages for online dailies differ from story pages and page downloading speed governs online daily reading, designs for photo location in online dailies have had to differ from those in print newspapers.

This study indicates that the dominant (28%) lead photo mode on homes pages was raw-wrapping on the right of a headline. The other two most popular home-page designs were armpit design (24%) and above-headline design (22%). These three models (74%) were mainstream home-page photo location designs in the online dailies under study. Noticeably, 11% of the online dailies placed lead photos between the headline and text, which, according to the standard of print journalism, is not a good practice.

On story pages, the most popular location design was armpit design (46%). Obviously, those editors wanted readers to see the headline first. A small portion of the online dailies (18%) placed lead in the middle of the story. That is, readers read the story first, and after a while, they saw the lead photo. Raw-wrap designs (10%) were occasionally seen, and below-headline design (5%) was rare. Extremely few online dailies used above-headline design and if they did, they did not use it every day. During observation for the study, only two above-headline designs in story pages were seen. Some online dailies (21%) varied photo location design from day to day on story pages.

Photo linkability

Fifty-nine percent of the online dailies allowed readers to click on a home-page photo to view different items such as the accompanying story, a bigger photo, more photos or a group of headlines while only 23% of the online dailies that used any photos in story pages made story-page photos linkable to a larger version of the same photo or to a photo gallery. One-third of the online dailies (33%) had linkable home-page photos but unlinkable story-page photos; 35% had no linkable photos at all; and 15% had linkable photos in both cases.

Several online dailies did not design photo linkability consistently. Clicking on some photos would lead to the corresponding stories while clicking on other photos would lead to a cluster of headlines or a larger version of the photo. Or yesterday's photos were linkable, but today's might not be. Readers of these online dailies could easily get confused as to what they could expect by clicking on a photo.

Consistency of photo presentations over time

Most online dailies used the same number of photos on the home page (70%), placed the home-page lead photos (90%) and the story-page lead photos (88%) at the same location, made the home-page photos (90%) and the story-page lead photos (94%) either always linkable or always not linkable, and either always carried one or always carried no photo gallery (80%). It is understandable that editors would adjust the number, size and even location of photographs from day to day and from story to story based on needs. However, some basic features such as photo linkability and photo gallery should be consistently used so that, when readers come back to a daily, they can rely on their prior knowledge about that daily to see what they expect to see.

2. Do readers pay attention to online news photos or expect lead photos to be eye-catchers in online dailies? If they do, what are their preferences for viewing online photos? If they do not, what are the major concerns?

The great majority of the online daily readers (88%) claimed that they paid attention to news photos when they read print dailies. In comparison, 48% of them would like to view news photos in online dailies while a large portion of them (36%) maintained a neutral attitude. A middle-road mentality in using photos online was pervasive in their answers to many of the questions, and is best represented by a comment made by a respondent: "I like to see photographs but I have no patience with slow-loading files." The readers' struggles with photo uses revolved around the speed of image loading.

A little more than half of the readers agreed or strongly agreed that the contents of photos in their favorite online dailies were generally interesting (54%) and that online photos were attractively designed (59%). If a reader thought that the news photo contents were generally interesting and that the photos were well designed in online dailies, chances are s/he would like to view the photos (Pearson [chi square]=24.45, df=16, p<0.000). In other words, the quality of photo content and design decided their willingness to view online photos.

Most (83%) readers would like to see photos on the home pages of their most frequently read online dailies. For those 17% of the readers who did not like to see home-page news photos, the biggest complaints were slow photo loading speed (52%); boring news content (18%) and poor designs of photos (9%). More than 20 readers wrote in their voluntary verbal statements that photos should not be used unless they are relevant to the story and interesting to see and contribute to an understanding of the story.

Many readers liked to see a very visual home page. Half of them preferred to see multiple small photos on the home page of an online daily. A little more than a quarter of them (28%) preferred to see only one but big lead photo on a home page. This study supports Jacobson's assumption that multiple images are more likely to attract attention than a single image.

Almost half of the readers (49%) wished to see a bigger photo by clicking on a photo on the home page of an online daily. Another 12% of the readers wished even to see more photos after clicking. Thirty percent of the readers wished to see the accompanying story by clicking on a photo. Some readers believed it makes sense to click on a photo for a larger or more photos, and click on a headline for the accompanying story.

More than half of the readers (57%) liked to view photos in a photo gallery in an online daily. This study supports the findings about the popularity of photo galleries in many online dailies revealed in DeVigal's article (2000). Only 18% of the readers did not like the photo gallery. The other 25% were neutral.

The readers were asked, if a photo and a text coexist, which one they normally would like to see first in an online daily. Readers had almost no preference between photo (27%) and text (28%). Instead, 40% of them preferred to see them together at the same time. One reader offered this statement about the relationship between text and photo which was representative of many readers:

Photos make stories interesting and add depth to their content. But the text is the heart of the story, so the photo shouldn't obscure or overpower the text, and it is undesirable when large photos make the reader scroll too much further to get to the substance of the article.

The findings from this study did not support the findings from the Lewenstein et al. 2000 eye-tracking experiment, which suggested that readers preferred seeing text to seeing photos.

When asked on which side of the headline they wanted to see the photo on a story page, more readers chose left (21%) than right (8%). Raw-wrapping left was one of the three favorite designs chosen by the readers. The other two favorite designs were photo-below-headline design (24%) and armpit design (21%). Only 16% of the readers favored the photo-above-headline design as the most common design in print dailies. Considering the slow photo loading speed and the possibility of blocking the view of the headline if the photo is big, this finding is not surprising. Most readers urged that photos should not dominate the headline. "It should be of equal or less weight than the headline, never more weight than the headline," wrote a reader. On the other hand, few readers (10%) liked to see a photo completely buried inside a story.

Readers liked to see photos of medium size. The readers were shown three web pages that contained a story with a photo of three different sizes. (17) They were asked to choose the most preferred size with the photo loading speed taken into consideration. As a result, 54% of the readers preferred the medium-size photo, 24% preferred the large photo and only 8% preferred the small photo. The rest of the 14% were not sure. It was found that photo-size preference was not age bound. Older readers also preferred the medium-size photo to the large one. As some readers suggested in their written comments, too small a photo could make them believe that the accompanying story is not important, and they tended to ignore the story. One reader made a comment in this regard: "Photos for online newspapers should be large enough that one can make out details, especially faces and signs, but not so big that they unduly slow download times."

Finally, most readers (71%) cared about the consistencies of using photos on home pages and story pages over days. They hoped that photos would be placed in the same spot on a page and the sizes would be kept the same, but they did not care much about how many photos were used either in a home page or a story page.

3. What can media do better by taking into consideration readers expectations?

Compared to the data from the content analysis, the reader survey data suggest that online dailies can do better in the following areas.

First, readers did want home pages to be visual. Most readers preferred home pages with multiple photos to those with only one and some others preferred to see a large home-page lead photo. Nevertheless, the more photos there are, and the bigger the photos are on a page, the slower it is for a page to load, as shown in the content analysis. To compromise the photo loading speed and the readers' needs, the author suggests that an approximately 3-inch-diagonal dominant photo be used for the first top story, and smaller photos be used for other top stories on the home page. Not using photos in an online daily simply to cater only to the preference of text-oriented readers does not seem like a good strategy. Photos can be used in online dailies to attract photo-oriented readers as has always been done in print dailies, and used as an alterative to text-oriented readers when they wish to see photos.

Second, photo sizes should be appropriately controlled. Readers preferred medium-size photos, which are approximately 3 inches diagonally. The content analysis showed that a two-to-three-inch-diagonal photo is small enough to be completely loaded often at the time when the text is loaded, but also big enough for readers to see clearly the content so long as the photo is not an environmental shot. (18) The loading speed was not too slow even for a four-inch-diagonal photo. For instance, Oakland Tribune often used a four-to-five-inch-diagonal photo raw-wrapped on the left. The photo began to be loaded at the time when the text was shown and the loading was completed often within four seconds on the author's Pentium 3 computer with a 56K modem. The average diagonal of the homepage lead photos was 2.7 inches, while the story-page counterpart was 3.1 inches. These dimensions suggest that many online dailies were aware of the competing needs of loading photos quickly and presenting visual information clearly. If a photo of such a size located appropriately around the headline, it could well catch readers' attention without delay or without much delay. In light of the print newspaper standard, a two-to-three-inch-diagonal photo is small. To better utilize such a small space, the author suggests that online dailies use more close-ups and medium-distance photos and use less environmental shots; and that, if an environmental shot has to be used, the size has to be larger.

Third, photo location is probably the biggest trick in attracting readers. Currently, armpit design prevailed both on home pages and on story pages because it contextualizes the photo while still attracting readers' attention. Raw-wrapping a photo on the right of a headline, a model

carried over from print newspapers, is also popular on home pages. However, readers' other two favorite designs were raw-wrapping photo on the left of the headline (6% found on home pages and 8% on story pages) and below-headline design (10% found on home pages and 4% on story pages) were rarely seen in online dailies.

Raw-wrapping a photo on the left side of a headline is not a standard practice in print dailies, but it could make good sense in online dailies and is routinely used by a few online dailies such as Oakland Tribune and Fort Lauderdale's Sun Sentinel. A photo raw-wrapped by a headline can't be too big and there is no difference between the speeds for loading a photo on the left or on the right of a headline. Raw-wrapping a photo on the left of the headline could well attract the attention of those photo-loving readers, but does not block the view of those text-oriented readers. The photo and the headline can attract both kinds of readers at the same time. "The photo must grab your attention as much as the headline," commented a reader. Another reader made a more specific comment: "Being on the left is helpful because the photo remains in my view for the beginning of the article, providing good motivational interest to continue with the story." Considering the fact that the largest portion of the readers (40%) p referred to see photo and text on the same height, raw-wrapping left is a good choice.

Also, a 3-inch-diagonal-photo below a headline is small enough to be seen on one screen together with the headline, is contextualized by the headline and does not block readers' view of the headline. Therefore, using the photo-below-headline design online also makes very good sense. (19) Based on the survey findings, the author suggests that online dailies use more raw-wrapping photo left design and photo-below-headline design both for story pages and home pages.

Fourth, online dailies should standardize photo linkability and the use of photo gallery. Most readers would like to click on a homepage photo to see a larger photo or a photo gallery if they feel interested in the photo or the story. Most readers (57%) loved photo galleries. Some readers suggested that by clicking on a headline they could read the accompanying story. The status quo is that only 13% of the online dailies allowed readers to click on a home-page photo to see a larger photo. The rest either did not allow photo linkability or allowed clicking a photo to see the accompanying story. Only 32% of the online dailies carried photo galleries. The author suggests that online dailies make photo linkability and photo gallery basic features of an online daily, design photo linking function consistently over time, and carry photo gallery frequently. Readers should be allowed to click on a headline to see the accompanying story, and click on a photo either on the home page or on a story page to see the larger version of the same photo or more photos in a photo gallery.

Finally, photo content and design quality should be strictly controlled. Using photos of inferior content quality or design quality could seriously harm the reputation of an online daily. It is important that online dailies make the same effort to serve online readers as their print counterparts do.

One example of a substandard photo found during observation was one in which the subject's face was smaller than a dime, the subject's back was to the camera, and a lamp-pole appears to stick out of her head.

Another example of substandard use of a photo online involved bad layout. The space for photo and text was not coordinated. There is supposed to be one pica (20) (a newspaper standard) of space between the photo and the body text, but the text almost touched the photo in this example. In addition, the caption was sloppily designed. A large area of empty space was left over and wasted below the story summary. No print newspaper editors would regard such trapped white space as professional and acceptable. In the three months of data collection for content analysis, the author came across numerous examples of poor uses of photos in many of these most popular online dailies. The author suggests that online editors be trained to adopt the same quality control protocols used by their print daily counterparts. The issues of photo-text relationship design, photo quality and presentation consistency are all a matter of professionalism, and professionalism needs to prevail in online dailies.

CONCLUSIONS

American culture has become overwhelmingly visual over the last half century as TV and digital imaging technology became increasingly popular. However, the comparatively primitive Internet technology has made the use of news photography in online dailies problematic. The slow Internet speed still is the top disincentive, both for some online news dailies' willingness to use photos and for readers' patience for viewing photos online. In response, some online dailies have remained photo free or text heavy rather than making news more visual. The role of news photography is being marginalized in those online dailies.

This study shows, however; that there are ways for online dailies to catch readers' eyes with photos. Slow Internet speed should not become an excuse for online dailies to forgo professional communication through effective word-picture combinations. The author has noticed that some online newspapers mentioned in this article have been redesigned since this study was done two years ago. Valid critique about those online dailies made two years ago may no longer apply to them now.

It also is fair to conclude that most online dailies under study do show efforts to attract readers' attention with photos, but the efforts are limited and sometimes can be more appropriate. Most readers do like to see online news photos and many expect lead photos to be eye-catchers. Comparing the online daily performances and readers' expectations, the author believes that, to catch readers' eyes with photos, online dailies should first make visual as many stories as needed. Home pages can use a roughly 3-inch-diagonal linkable dominant photo and several smaller linkable photos. The dominant photo can use above-headline design, raw-wrap left design, armpit design or below-headline design. Story pages can use a slightly bigger unlinkable lead photo with the same location designs except for the above-headline design. Online dailies are strongly encouraged to use more photo galleries with bigger photos to better visualize news stories. Both quality photos and quality designs should be shown consistently across stories and over time in online dailies. An online daily can win over both photo-loving readers and text-oriented readers by appropriately controlling the number, size, and location of photos, appropriately using linkable photos and photo galleries and maintaining the consistency of photo presentations.

ENDNOTES

(1.) A lead photo is a photo around a headline. Both home pages and story pages could contain a lead photo.

(2.) By November 7, 2000, 341 U.S. online newspapers were listed in Newsweb.net. On the same day, 1120 U.S. daily online newspapers were listed in NewsDirectory.com and 1255 listed in Editor and Publisher's Media Links site. See details at http://emedia1.mediainfo.com/emedia/.

(3.) See http://www.poynter.org/centerpiece/071200andrew.htm

(4.) The author understands that image size can be affected by screen setting, browser and even version of browser used by readers' computers. Readers can use very different settings and browsers from the one used by the coders. For coding purpose, however, it is important for coders to maintain consistency so that comparisons among dailies can be made. The author consistently used a screen setting of 1024 by 768 pixels and Netscape 6 when measuring all photos. The author also understands that image size appearing in a browser can be defined by the page producer via HTML coding regardless of the image's actual pixel size. Fortunately, examples of photos randomly sampled from each online daily in the pilot study showed that almost all producers/webmasters understood how to optimize images.

(5.) The time for loading pages was measured only on the author's computer with the same Web browser and the same ISP provider so that comparisons could be made among online dailies. The actual page loading speed on different computers could be affected by many factors such as time of the day, the ISP provider and the robustness of the computer processor. It was found later that most online dailies performed comparatively consistently in the time of loading home pages. The amount of such time collected from the five times of data collecting over three months was averaged for each daily.

(6.) Photo linkability refers to whether a reader can click on a photo to see a bigger version of the photo or the accompanying story. Photo gallery refers to an online slide show.

(7.) In the following finding reporting, those six photo-free online dailies are excluded. All the statistical calculations are based on those 50 online dailies that did use photos.

(8.) All the following photo sizes refer to diagonal length. Diagonal length was used because it is a more accurate index than just width to show how much space a photo takes.

(9.) Stories were not observed on the home pages. Home pages normally contained headline links. Most online dailies offered decks (news summaries), too, on the home pages. A page that carries a story is defined as a story page.

(10.) Mug shots were counted in terms of placement and number of photos used in an online daily, but not counted in terms of size because they are supposed to be small and the inclusion of mug shots in photo size calculation could skew the data.

(11.) Not all photo galleries were used to present news. Photo galleries in some online dailies (i.e. Detroit Free Press) served as a showcase of individual staff photographers' work. The purpose was more to show achievements than to present news. Some other online dailies (i.e. The Miami Herald) used their photo galleries to sell photos taken by their staff photographers. Also, some other online dailies (i.e. The Arizona Republic) used a photo gallery to review important events that happened in current month or current year. The purpose was more to show history than to show news. None of such photo galleries were counted.

(12.) The time was counted from the time when the bookmark of a daily site was clicked on until the home-page lead photo was completely loaded. If an online daily did not have a home-page lead photo, the time counted was the time from when the bookmark was clicked on until the time when the page was seen. A stopwatch with lap and log functions was used to guarantee the precision of counting.

(13.) USA Today, a newspaper famous for its innovative use of photos, put a photo caption instead of a photo on top of a story in its online version if a photo was used in its print version.

(14.) San Francisco Chronicle is a case in point. It did not carry any photos any day on the home page during observation. Inside story pages, most stories did not have photos, and if photos were seen inside a story, they were the miniature version. Readers needed to click on one of them to see the accompanying photo gallery.

(15.) For instance, San Diego Union-Tribune often had five to six photos, but the first story often got no photo.

(16.) Armpitting a photo in the left columns is rarely seen in print newspapers.

(17.) With a standard screen resolution of 1024 by 768 pixels, the large photo was 7 inches long; the medium-size photo was 3 inches long; and the small photo was 1.5 inches long.

(18.) An environmental shot shows the overview of a scene and usually a human subject looks small in it.

(19.) Probably influenced by online newspapers, Chicago Tribune began to use the photo-below-headline design in late 2000.

(20.) 1 inch = 6 picas.

REFERENCES

Adweek.com (2000, Sep 21). Web users unhappy with news sites. Retrieved December 5, 2000, from http://www.nuaknowledgenews.com/?f=VS&art_id=506&rel=true.

Alexander, M. J. (1997, July). Net gain? Online newspapers take time from their cyber-stampede to assess benefits and drawbacks of electronic editions. Paper presented at Association for Education in Journalism and Mass Communication convention, Chicago, IL.

Bain, C. and Weaver, D. (1979). Newspaper design and newspaper readership. Paper presented to the Graphics Division, Association for Education in Journalism, Houston, TX.

Chamberlin, R. J. (1998, August). Perceptions of graphics versus no graphics on web sites. Paper presented at Association for Education in Journalism and Mass Communication convention, Baltimore, MD.

DeVigal A. Putting the Eyetrack Study to Good Use (2000). Retrieved December 5, 2000, from http://www.poynter.org/centerpiece/071200andrew.htm

Garcia, M. and Stark, P. (1991). Eyes on the News. St. Petersburg, Fla.: Poynter Institute for Media Studies.

Gates, D. (2002, May 2). The future of news: Newspapers in the digital age. Retrieved September 27, 2002, from http://www.ojr.org/ojr/future/1020298748.php

Gubman, J. and Greer, J. (1997, July). An analysis of online sites produced by U.S. newspapers: Are the critics right? Paper presented at Association for Education in Journalism and Mass Communication convention, Chicago, IL.

Harrower, T. (1998). The newspaper designer's handbook, 4th Edition. Boston: The McGraw-Hill Companies, Inc.

Jacobson A. (2000). Projects too much from too little. Retrieved December 5, 2000, from http://www.poynter.org/oenterpiece/071200alan.htm.

Kamerer, D. and Bressers, B. (1998, August). Online newspapers: A trend study of news content and technical features. Paper presented at Association for Education in Journalism and Mass Communication convention, Baltimore, MD.

Lewenstein, M. (2000, August 18). Study snapshot suggests serendipity lives online. In Poynter.org. Retrieved December 5, 2000, from http://www.poynter.org/centerpiece/081800.htm.

Lewenstein, M., DeVigal, A. and Jacobson, A. (2000, July 12). Eyetracking: A closer look. In Poynter.org. Retrieved December 5, 2000, from http://www.poynter.org/centerpiece/071200.htm

Mings, S. M. (1997). Uses and gratifications of online newspapers: A preliminary study. In Electronic Journal of Communication, vol. 7, No. 3. Retrieved December 5, 2000, from http://www.cios.org/getfile\MINGS_V7N397.

NAA. (2002, May 1) Online newspaper sites are top for news in US. Retrieved September 27, 2002, from http://www.nua.com/surveys/index.cgi?f=VS&art_id=905357910&rel=true

Schierhorn, C., Wearden, S. T., Schierhorn, A., Tabar P. S. and Andrews S.C. (1998, August). Digital formats for the future: The web vs. paper vs. a vertical-screen, page-based design. Paper presented at Association for Education in Journalism and Mass Communication convention, Baltimore, MD.

Tankard, J. (1998, August). Online newspapers: Living up to their potential? Paper presented at Association for Education in Journalism and Mass Communication convention, Baltimore, MD.

Tremayne, M. (1999, August). Media Convergence on the Internet. Paper presented at Association for Education in Journalism and Mass Communication convention, New Orleans, LA.

Vargo, K. (1998, August). Readers Response to Digital News Stories Presented in Layers and Links. Paper presented at Association for Education in Journalism and Mass Communication convention, Baltimore, MD.

Wired.com (2000, May 05). Eye (Don't) Spy. Retrieved December 5, 2000, from http://www.nuaknowledgenews.com/?f=VS&art_id=211&rel=true.

RELATED ARTICLE: Separate study indicates online readers will respond best when report-related images accompany news headlines

EDITOR'S NOTE: VzBib editor Kimberly Bissell included the following entry in the Spring edition of the annotated bibliography. We thought it important to call your attention to this research because of its resonance with Huang's research article on using photographs in online news sites.

Imagery effects on the selective reading of Internet newsmagazine. Silvia Knobloch, Mattias Hastall, Dolf Zillmann, and Coy Callision, Communication Research, 30(1) February 2003, pages 3-29.

This study examines the effects of imagery on selective exposure to news reports presented on the Internet. The authors predict that the incorporation of report-related images in the online display of news headlines increases the probability that news consumers will both select and read the associated articles, and the effect will be stronger for the incorporation of threatening images than of innocuous images. The manipulated articles were presented either without images, with text-related innocuous images, or with text-related threatening images in both their headline displays and their text bodies. Their results point to the display of headlines as the most effective vehicle for such incorporation. It is suggested that news editors should recognize that the competition for selective attention to text favors imagery as a vital accompaniment of text, that image-rich media environments render imageless narration ineffective, and that the employment of pertinent images is therefore imperative.

Edgar Huang Is an assistant professor in the Department of Journalism and Media Studies, University of South Florida St. Petersburg. He teaches visual communication, Web publishing, media convergence, research methods, media writing and other courses.