How Much Time Should Be Allowed to Read Text

In 2010, researchers at the Academy of Alberta found that reading comprehension was impaired when content was presented on a mobile-size screen versus a larger computer screen. A simple explanation for this event was that, with a pocket-sized screen, users saw less of the text at whatsoever given time, so they had to rely more on their memory to admission contextual data needed during reading. In other words, the smaller screen resulted in a college working-retentivity load. People could not sustain that college load, so their comprehension suffered.

In our enquiry, conducted six years afterwards, we found a surprisingly different event. We asked 276 participants to read a variety of articles on various topics on either a mobile phone or a personal computer. Some of the articles were easy and some were difficult. After each article, we asked participants to answer a few questions to measure their level of comprehension of the content. Nosotros institute no practical differences in the comprehension scores of the participants, whether they were reading on a mobile device or a computer.

Despite this finding, nosotros all the same recommend prioritizing brevity and reducing unnecessary content when writing for mobile.

Methodology and Analysis

Nosotros began our written report with the expectation that our findings would support the original 2010 conclusions. Our two hypotheses were:

1. Reading comprehension is lower when manufactures are read on mobile phones vs computers.
2. Articles that are difficult to read affect mobile comprehension more than desktop comprehension.

Our participants were a broad sample of general web users. In all phases of the written report, participants were asked to read a diverseness of articles on different topics and levels of difficulty.

The difficulty of the articles ("easy" or "hard") was determined by the length of the article (word count) and the difficulty of the linguistic communication used (co-ordinate to the Flesch-Kincaid reading-level formula). All of the articles were presented as HTML pages created from the same simple pattern template.

The divergence between the easy and difficult manufactures is summarized in the following table (averaged across the articles used in the last two rounds of our inquiry):

	Piece of cake articles	Hard articles
Average length	404 words	988 words
Average reading level	8th form	12th form

Past way of comparison, the article you're reading correct now has 2,072 words and is written at the 13th grade readability level.

Participants read one-half of the manufactures on a computer and half of the articles on a phone, alternating between the computer and the phone (we randomized the device used for each person'southward first article). Later reading each article, they answered multiple selection questions to appraise how well they had comprehended and retained the information they had just read.

We began our study with a small pilot examination (something we highly recommend for all UX research studies). When our results contradicted the previous inquiry by suggesting that there was no divergence between comprehension for mobile and computers, we had to consider the possibility that our methodology was faulty, and so nosotros proceeded through a serial of studies with different stimuli and test conditions. In the finish, our enquiry combined iv measurement studies:

• 10-participant online pilot
• 30-participant online report
• 40-participant in-person report
• 206-participant online study

For the pilot, we used content pulled from live websites; for the other studies, nosotros used articles that we wrote ourselves to have more than control over the content. Half the articles in our studies were piece of cake and half were difficult, and each participant saw both hard and easy passages on each device. (In other words, both commodity difficulty and presentation device were within-subjects independent variables. In our final analysis, the type of study — online or in-person — was the third between-subjects independent variable.)

For all studies, we used comprehension scores as the main dependent variable. These scores are percentages from 0 to 100% that took into business relationship right responses, but also penalized incorrect responses (meet the attached materials for a precise definition). For the in-person studies we also measured article reading times. Note that our comprehension metric was unlike than that used by Singh and colleagues in their original study (they used a cloze test).

In each stage of the study, we tweaked our methodology or stimuli, but found the aforementioned surprising result — no perceptible difference in reading comprehension between the devices. To supplement the quantitative testing, we besides ran a set of focus groups with the participants in our in-person study, asking them to discuss how they read web content and how they perceive reading on mobile devices vs. reading on computers.

(If you lot'd like to replicate our inquiry — or conduct new reading research, maybe with additional devices — you tin can download our concluding stimuli and associated comprehension quiz questions from the link at the bottom of this commodity.)

To maximize the statistical ability of our study, we performed a mixed ANOVA on the comprehension scores from all four phases of the study, controlling for the unlike articles and study procedures used. This assay included one,629 cases where a user read an article and completed the comprehension quiz for that article.

For the in-person data, nosotros also ran a repeated-measure ANOVA on the reading speed (defined equally the time taken to read one word).

Comprehension Scores: Slightly College on Mobile

We establish that, on boilerplate, comprehension scores were slightly higher when users read the articles on mobile devices. Although the effect of device was statistically significant (at p = 0.0006), the difference in comprehension scores was non practically significant: comprehension on mobile was about 3 per centum points higher than on a computer, with a 95% confidence interval of one% to 5%.

Unsurprisingly, comprehension scores were lower for difficult articles compared with easy ones (this main effect of difficulty was pregnant at p=0.0001).

Average comprehension scores for like shooting fish in a barrel and hard articles, past device.

Very Difficult Content May Cause Lower Comprehension on Mobile

Our information analysis of comprehension scores too found a marginally significant interaction (p=.10) between content difficulty (easy vs. hard articles) and reading device (mobile vs. computer), indicating that the (already very pocket-size) comprehension-score advantage for mobile is reduced when reading difficult articles.

More research is needed to know if this effect is real, but if it is, and if it continues to exist true for progressively more difficult content (beyond the difficulty levels included in our study), then it may be the case that very difficult content is harder to read on a telephone than on a reckoner.

Reading Speeds: Readers Slow Downward for Difficult Articles on Mobile

For the in-person data, we as well captured the time each user spent reading each article. Considering articles varied in length, instead of analyzing the overall reading fourth dimension, we looked at the reading speed, defined as the article reading time divided past the article length (in words).

Our repeated-measures ANOVA yielded a significant interaction of device and difficulty (p =0.01). Like shooting fish in a barrel passages were read near every bit fast on both devices, but hard passages actually took longer on mobile versus estimator. (On boilerplate, participants spent about 30 milliseconds more than on each word when reading on mobile than on a figurer.)

Average reading speeds for easy and hard articles, by device.

Speed-Accuracy Tradeoff on Mobile

Why did we get no comprehension-score difference between the devices? Does this effect contradict our theory that text presented on a small screen incurs a higher cognitive load than text presented on a larger screen?

We can detect the answer by considering the reading-speed deviation in mobile vs. computer. Call back, when participants were reading like shooting fish in a barrel articles, their reading speeds were about the aforementioned on mobile as on a figurer. All the same, when the participants read difficult articles (long word counts, challenging topics and language), their reading speeds slowed down.

In other words, they could not sustain the college working-retentiveness load, and, to achieve the same level of comprehension, they had to either:

1. read more carefully and try to recollect potentially relevant data, or
2. go back and re-read certain passages.

In psychology, this phenomenon is referred to equally a speed–accuracy tradeoff — users had to ho-hum down to accomplish the same level of comprehension for difficult articles on a phone as they did on a computer.

This suggests that, while reading comprehension may be comparable on a phone and a computer for easy articles, reading on mobile becomes more difficult as the complexity of the content increases. (The marginally pregnant interaction on comprehension scores points in that direction, too.)

The speed-accuracy tradeoff also offers a potential explanation for why we obtained very different results than Singh and his colleagues: their report used very difficult content (privacy policies) as stimuli. It's possible that, at that level of complexity, participants simply ended up sacrificing some of their comprehension to maintain a decent completion speed in the experiment. With extremely complex content, we may notwithstanding see substantial decreases in comprehension scores on mobile. But then, we'd never recommend that any web content be as complex as privacy policies tend to be.

There are other several possible reasons that may have contributed to the departure in the results:

• The prior study used a different comprehension metric (cloze examination) than nosotros did. It's possible that our test tapped into different cerebral processes.
• Text presentation on mobile devices has significantly improved since Singh et al.'s study. Smartphones screens are bigger, and their resolutions are crisper: a typical phone screen today (iPhone seven) has six.v times more pixels than a typical phone screen at the fourth dimension of the original research (iPhone 3).
• Some participants reported frequently reading manufactures on their phones, and feeling comfortable doing so. They commented that thumb scrolling felt easier than acquiring the scrollbar and dragging it — which some users nonetheless do, not beingness accepted to a roll wheel on a mouse.
• Some participants reported that they liked the lack of "distractions" on the mobile device. For sequential reading, like articles, mobile may have the advantage hither. Though the smaller window limits the amount of information that tin can exist viewed at one time, it likewise can filter out competing information.

Central Takeaways

For linear content like articles, especially easy-to-read content, comprehension on mobile appears to be on-par with larger devices.

Does this result mean that mobile devices are now but as easy to utilise as desktops or laptops? Unfortunately, no.

First, we know that, in general, task performance on mobile devices is nevertheless lower than desktops or laptops. We measured reading comprehension in this report, simply most web tasks involve much more than than reading. Articles are linear content — they don't reflect all web content or online tasks. Well-nigh online activities involve some degree of navigation and interaction. Ecommerce and other web tasks require substantial navigation and comparisons between multiple pieces of content.

Second, even if the comprehension scores were comparable on mobile devices and computers, nosotros saw that mobile readers paid a toll in reading speeds: when the articles were more difficult, they were slower to achieve the same level of comprehension as on the reckoner. Thus, for difficult passages, mobile readers had to work harder than figurer readers. Comprehension scores are simply one aspect of chore performance; reading speed is another i, and to get a full picture, they must be considered together.

Recommendations for Mobile Content

We've long advocated brevity in mobile content, and that dominion even so stands. Short, easy passages were faster to read, regardless of the device. That said, the strict requirement for ultra-short content in mobile may be relaxed somewhat if the content:

• Is appropriately written for full general web audience (no challenging topics or linguistic communication)
• Serves an entertainment, time-killing, or informative purpose

However, sure sites do offer extremely challenging content, including many organizations within the fiscal, medical, and scientific sectors; certain government agencies; and B2B sites that target Information technology or engineering customers. If yous're one of these sites, nosotros highly recommend that you run your own usability studies of any high-complexity material you desire your readers to access on mobile devices.

Even though mobile reading comprehension for easy articles seems to be comparable with estimator comprehension, it doesn't hateful nosotros tin can ignore the still nowadays limitations of mobile.

Near writing on the web is not in a linear format—it requires some degree of interactive or comparative effort, which adds to the reader's cognitive load. As demonstrated past the speed-accuracy tradeoff, readers volition probably need to exert more effort to comprehend hard subjects on mobile. Many mobile activities are also performed on-the-become, which means environmental conditions will often fragment user attending and focus.

For the majority of mobile content scenarios, the need for brevity and prioritization is however critical.

Reference

R.I. Singh, Yard. Sumeeth, and J. Miller: "Evaluating the Readability of Privacy Policies in Mobile Environments," International Journal of Mobile Human Computer Interaction, vol. 3, no. 1 (January–March 2011), pp. 55–78.

lyonuponed.blogspot.com

Source: https://www.nngroup.com/articles/mobile-content/