Reading Comprehension: Paper or Screen?
In a recent post, I critiqued the claims in Ferris Jabr's Scientific American article, "Why the Brain Prefers Paper" that addressed the differences in comprehension between reading from paper and reading from screens.
Where Jabr argues that screen reading limits comprehension, I show how his analysis focused only on certain kinds of screen reading, ignoring features of digital texts that challenge his claims. I also looked at some of the research Jabr cites, showing how the results of that research didn't support his particular arguments. In this post, I want to take another look at that research.
Although I believe Jabr misinterprets the research he cites, I think that research paints a complicated and interesting picture of the state of electronic reading that is worth further discussion. Although it is wrong, at least on the basis of these studies, to say that electronic reading inhibits comprehension, the question of how digital technologies affect reading remains important for educators to understand. The two studies I focused on from Jabr's piece — one that compared reading from paper and PDFs by Norwegian 10th graders and one of university students under similar conditions — show differences in comprehension between groups who read from paper and those who read from PDFs. How, then, can we account for these differences? What is it about the reading technologies — in print and on screen — that affected the reader comprehension?
In the study of the 10th graders, the electronic reading interface did not allow them to view multiple pages of the document at a time and when it came time to answer the comprehension questions, they had to use the same screen to read the text and answer questions, tabbing from the quiz window to the PDF document and back. Conversely, students reading on paper could spread out their documents so that they could see all of the information at once, and they were able to access these documents in this way while taking the quiz, glancing from paper to screen.
These points might seem like quibbles, except that the study's authors indicate that these differences could have caused the lower comprehension scores for the PDF group. In that light, it is crucial to note that the restrictions facing the PDF group are not inherent to screens or digital text; rather they were limitations of the technology they used. Students could just as easily have looked from a secondary screen displaying the PDF to the screen with the quiz as those in the print group looked from their handouts to the quiz. Further, one of the primary affordances of digital text — text searching — appears to either have been disabled for students in the PDF group or not used.
Low resolution monitors could have affected the comprehension levels of the PDF group in the second study by Wästlund et al., in which students read documents on a screen with an 800x600 resolution. At this resolution, it is difficult to legibly view even a single page of text at one time. For example, this is how a PDF of that study would appear at 800x600 resolution when viewing the document at 100% of its size:
Researchers have known for decades that the resolution of display technology and the technical features of reading software affect screen reading, yet this point is not given much attention in either study. This could be because the display technologies in both studies were ones that were common at that time. However, given the advancement of this technology, it is difficult to indict current screen reading based on these results.
I argue that the differences revealed in these studies are not between paper and screen, as Jabr suggests, or even between print and digital. Rather, these differences in reading comprehension seem to be a product of how readers interface with these print and electronic technologies and the affordances that each provides to these readers.
Media theorists use the term affordance to refer to the types of interactions that a technology allows for. One of the best features of Jabr's article is how he draws attention to the physicality of reading and how reading technologies can aid in navigating and comprehending a text. The spatiality of a text is important to comprehension — media theorist Walter Ong has argued at length that the spatial fixity of writing and print, as opposed to the transience of oral communication, was a major factor in enabling analytic thought — and Jabr's description of the effect of the physicality of books on the spatial navigation of a text illustrates a number of the important affordances of book technology.
Unfortunately, despite the attention he pays to the physicality of print, Jabr consistently ignores or dismisses the physicality of digital texts and the screens on which we read them. For example, he argues that screens prevent readers from seeing what they are reading in the context of the whole, a process that he likens to "zooming out" on a map to a particular location in a wider context. Of course, books only allow readers to do this in the sense of knowing a specific spot or location on a page. Although he may have been simply referring to the ability to locate the position of a passage within the whole by noting the physicality of the book, Jabr's map example suggests something different.
When you zoom out on a map, you don't simply locate the latitude and longitude of a location, but you see all of the surrounding details as well. This is not an affordance of the books that Jabr uses as examples. Although some print technologies allow readers to zoom out and see an entire text — one can spread out loose-leaf pages only printed on one side so that they all are in one's field of vision; screenfold books like the Codex Zouche-Nuttall allowed readers to see the entire contents of the text or juxtapose non-contiguous pages — the books we are familiar with, books that are bound on one side and only allow the reader to see two pages at a time, do not.
In fact, comparing the affordances of various reading technologies — paper and screen — in this method works to undermine Jabr's argument, suggesting ways in which electronic reading provides affordances that print reading cannot. Most PDF reading software allows readers to zoom out to see the entirety of a text, placing multiple pages within a single field of vision. The image below shows every page of Wästlund et al.'s article displayed on a high-resolution monitor:
While the body text is too small to read at this resolution, it is possible to read headings and to visually determine sections within the text, as well as to identify features like tables and the references list. Readers can quickly navigate between this view and individual pages by clicking on the desired page.
One might object that while digital texts make such views possible, readers cannot easily read any single page as they might if they had a similar number of pages spread on the floor or some other surface. I would counter that even in the paper situation, readers must engage in a process of physically zooming in and out to read pages that are designed to be viewed at most at arm's length.
Jabr also criticizes the scroll bars in digital texts that indicate a reader's position in the text and its overall length, dismissing them as being "more vague" than the positioning and locative features of print books. It is not clear how such a measure is more vague than determining the position of a passage in a book via the weight and relative thickness of the book's two halves. Rather, books present this information via tactile means while digital texts do so visually. Tellingly, Jabr ignores the features of digital books that are entirely less vague — the indication of the current page along with total pages given by nearly every PDF reader or the location position and total provided in Amazon's Kindle format. Such measures are in no sense vague, and they are surprisingly less vague than the sense of position offered by one's location in a print book or via a scroll bar.
There is another piece of this puzzle. In the study of university students, the researchers not only tested comprehension, they also asked participants a series of questions about their reading experience. One of their results was that students in the PDF group reported higher levels of fatigue and stress than those who read on paper. Mental effort sapped by stress and fatigue can affect comprehension, so determining the source of these effects is an important one to note.
First, it is possible that the design of the reading technologies contributed to this reaction. As noted earlier, screen resolution and projection technologies like the CRT screens used in the university study, seem to be harder to read than higher resolution text available via newer technologies, like the electronic paper in ereaders. Further, researchers have known for decades that one of the primary benefits of print is our familiarity with it. As Dillon notes, "there is a bewildering range of interfaces" for accessing electronic texts and "mastery of manipulation in one application is no guarantee of an ability to use another."
Sheets of paper and books, on the other hand, benefit from their familiarity as reading technologies. While a 10th grader or university student unfamiliar with a new computer or its software may have to devote precious cognitive resources to understanding and navigating the reader interface, their counterparts can simply read printed documents. Dillon acknowledges that these differences can affect reading practices, and this is understandable. However, this acknowledgement does not say anything about intrinsic differences between reading on paper versus reading on screens. Rather, it speaks to our culture's long acquaintance with print. Indeed, the decidedly mixed results of more recent comprehension studies speak to the plasticity of our minds and our ability to adapt to new media.
In short, there are a number of ways in which the screen resolution and/or software limitations of the PDF reading in these studies could have affected comprehension, and it isn't clear how comprehension would suffer when readers have access to paper-level text resolution and tools for zooming out of texts. It is possible that higher screen resolutions, combined with the superior affordances of digital texts researching and zooming, that screen reading will eclipse paper reading for comprehension tasks, for the very reasons that Jabr criticizes it. Further, it is likely that as readers become more familiar with the interfaces for electronic reading technologies that they will be able to devote more cognitive energy to reading and less to navigation, thus further improving comprehension in electronic reading.
What makes for effective reading, then? Before trying to answer this question, one has to acknowledge that the answer is complicated and depends on individual users' familiarity with the reading tools that they are using and the purpose for which they are reading. With that in mind, these studies suggest at least a few features. Readers should be able to clearly see the text, and the more text that can be placed within their field of vision at a time, the better. They should have a means for accessing their current place within the text, as well as the ability to quickly move through the text to find information. Finally, they should be familiar with the reading technology, so they can ignore it and simply read.
No single technology for reading from print meets all of these criteria. Loose sheets of paper are not an ideal medium for reading long texts, and books limit readers' access to just two pages at a time. However, it is possible to imagine an electronic reading interface that meets all of these criteria, even if current, popular reading technologies do not. This suggests that it will not be long before electronic reading might be a better choice than paper reading for reading for comprehension. Such a claim, of course, depends on the familiarity of readers with the interface of that reading technology and its ability to allow for the reading features listed here.