When Rosetta’s User Research team presented their usability test findings to a client, one recommendation in particular met with opposition. The team recommended that not all titles and text on the site be capitalized, because it makes reading the text more difficult and time consuming. The client argued that this was a real tough-guy brand and text that is all caps sends a real tough-guy message. In trying to support their point, the User Research team read as much research as possible on the topic but found that only a shockingly small amount of research has been published on the subject. Not wanting to base their recommendations on gut feelings, the User Research team took the only logical step: they did the research themselves.
EXPERIMENT, PHASE 1: TEXT PASSAGES
Sixty participants were instructed to read two product descriptions typed in 14 pt, black Calibri (Body) text. One of the product descriptions was short; the other was longer and more complex. The experiment was randomized so that all participants viewed one passage that was sentence case and one that was entirely uppercase, with each of these passages being one of the two product descriptions (one short and one long).
EXPERIMENT, PHASE 2: WEBSITE IMAGES
Participants were then shown two versions of the same web page, back-to-back, for five seconds each. One version contained sentence-case text within the hero image and the three e-Spot titles, and the other was all capitalized. Half of the participants viewed the sentence-case version first, and the other half saw the all-capitalized version first.
- In Phase 1, reading time was about equal for the shorter product description, regardless of letter case. However, for the description with longer, more complicated chunks of information, reading time was significantly longer when it was in all-capitalized text.
- In addition, participants (aggregated) focused much more intently on the sentence-case text, as indicated by the areas of darker red shown in the heat maps below.
- In Phase 2, participants who viewed the capitalized-text image first spent a significantly longer amount of time looking at this large text within the hero image than did participants who viewed the sentence-case-text image first.
Shown above are visualizations of how two representative participants looked at the text within the hero image. The circles represent fixations, with larger circles indicating longer fixations. Note that in the capitalized version (bottom), there are several more fixations spanning the hero text.
- In the three e-Spots below the hero image, participants (on average) who viewed the capitalized-text image first spent a greater amount of time fixating on the title text within the e-Spots than did participants who viewed the sentence-case image first.
- However, the participants who viewed the capitalized text image first (on average) paid far less attention to the text beneath these titles than did the participants who viewed the sentence-case image first. Shown below are focus maps for the two versions of e-Spots, depicting the aggregated focus of all participants in the study who viewed them first.
Shown at top is the heat map for the image with capitalized e-Spot titles. Shown at bottom is the heat map for the image with sentence-case e-Spot titles. Note that in the capitalized version (top), the focus is largely placed on the titles, whereas in the sentence-case version (bottom), much greater focus is placed on the underlying text.
THE BOTTOM LINE
Based on our study, capitalized text is helpful in small doses.
- Capitalized text can be read just as quickly as sentence-case text when it appears in manageable chunks. However, longer lengths of capitalized text slow down reading time and attract less attention than sentence-case text. Therefore, capitalized text should be kept to a minimum on websites.
- Large-font display of short chunks of capitalized text attracts a greater amount of attention than sentence-case text: the capitalized hero-image text garnered far more attention than its sentence-case counterpart, and as indicated by the previous point, this is not necessarily because reading time took longer.
- However, capitalized text can interrupt psychological flow: this explains why despite the fact that participants focused more on the capitalized e-Spot titles than the sentence-case e-Spot titles, participants were far more likely to continue reading the text beneath sentence-case headlines than they were to continue reading the text beneath capitalized headlines. These findings indicate that short chunks of capitalized text should be used to draw immediate attention to a prominent area, but sentence-case text should be used in smaller areas when the goal is to keep the consumer reading.
There are a million different variables that can affect attention and the reading time of capitalized versus sentence-case text: spacing, leading, font, color, and size are just a few. But from what our study can tell, capitalized text is something to be capitalized on.