1 Key Findings from Part II

Once again, as was mentioned in Sect. 1.4, Part II of this book has reviewed diverse studies using diverse methods of data collection, and this precludes any formal meta-analysis to integrate the findings. One must instead focus on the most common finding—the modal finding—regarding the legibility of serif and sans serif typefaces: superiority of serif typefaces; superiority of sans serif typefaces; or no difference. Part II has been concerned with the question of whether there are differences in the legibility of serif and sans serif typefaces when they are used to produce material to be read on computer monitors or other screens.

Studies of the legibility of letters and words have proved inconclusive (Sect. 11.1). The studies that employed authentic typefaces showed at most that some sans serif typefaces (Consolas, Letter Gothic, and Verdana) are more legible than some serif typefaces (Times New Roman and the slab serif typeface Courier). In addition, researchers who employed artificial typefaces confounded the presence or absence of serifs with variations in the width of the letters and variations in the spacing among successive letters. Early studies using the tachistoscopic presentation of printed letters and words showed that errors in their identification were often the result of confusions among visually similar letters (see Sect. 4.2), and this idea has been confirmed using screen-based presentation (Sect. 11.2). However, such confusions seem to be due to the design of individual characters in specific typefaces rather than to the presence or absence of serifs. Indeed, visual confusions are not more likely with sans serif typefaces than with serif typefaces, which contradicts the old hypothesis that serifs make letters easier to discriminate and identify (Legros, 1922, p. 11).

Early research using print-based presentation suggested that visual confusions were much less important when reading connected sentences (Vernon, 1929), but this notion does not seem to have been tested using presentation on computer screens. Five studies compared the speed with which participants read sentences displayed in serif and sans serif typefaces (Sect. 12.1): one found that serif typefaces were read more quickly, two found that sans serif typefaces were read more quickly, while two found no significant difference. Three of these studies also compared the participants’ accuracy: one found that sans serif typefaces were read more accurately, while two found no significant difference. Research into readers’ eye movements has not yielded consistent findings with regard to the presence or absence of serifs.

Five studies compared readers’ comprehension of meaningful text when presented on computer screens in serif and sans serif typefaces (Sect. 12.2). Four of the studies found no significant difference in the time taken to read the material in question. The fifth (Hojjati & Muniandy, 2014) found that material displayed in a sans serif typeface was read more quickly, but this study suffered from serious methodological problems. Three of the studies reported measures of comprehension: one found an advantage for a serif typeface, one found an advantage for a sans serif typeface, and the third found no significant difference.

The rapid serial visual presentation (RSVP) procedure constitutes another means of presenting meaningful material, and five studies have compared readers’ performance with serif and sans serif typefaces in terms of their comprehension, speed, or accuracy (Sect. 12.3). All five studies found no significant difference between serif and sans serif typefaces, except that serif styles were less legible with very small type or under conditions of low luminance (when, of course, serifs are likely to have been faint or even completely invisible). Two Korean studies compared the legibility of serif typefaces and sans serif typefaces when viewed on the screen of a personal digital assistant or a smartphone, but neither study found any significant difference in the participants’ reading performance (Sect. 12.4).

Differences in the legibility of serif and sans serif typefaces are not apparent even in readers whose visual systems are challenged as the result of disablement. Different studies have examined a heterogeneous sample of patients with visual impairment (Sect. 13.1), a large sample of readers with dyslexia (Sect. 13.2), and people with and without age-related macular degeneration (Sect. 13.3). In each case, there was no difference in the participants’ reading speed between serif typefaces and sans serif typefaces. Any differences in the participants’ preferences or eye movements could be attributed to the researchers’ failure to control the width or inter-letter spacing of different typefaces.

Finally, when reading material in internet browsers, by far the most common finding is that there is no significant difference between serif typefaces and sans serif typefaces in terms of the users’ reading comprehension, reading speed, or reading accuracy (Chap. 14). This applies regardless of whether or not researchers have used measures of reading “efficiency” to control for the possibility that readers employ some kind of trade-off between their speed and their accuracy. There is also no consistent evidence that readers have a preference between serif typefaces and sans serif typefaces when reading material in internet browsers. Once again, both serif and sans serif typefaces are regarded as being broadly appropriate for internet sites, whereas display and cursive typefaces are regarded as being generally inappropriate for serious use.

2 Preferences and Connotations

Some studies, though not all, have found that readers express a preference for sans serif typefaces when reading on computer screens, but in general both serif and sans serif typefaces are regarded by users as appropriate for online purposes (Sect. 12.5).

There is some evidence that serif and sans serif typefaces differ in their connotations or “personality”. This seems to reflect variations in reader’ expectations, which in turn depend on their prior experience and familiarity with different typefaces. When reading on computer screens, these differences are small in magnitude and are often not statistically significant. One study (Kaspar et al., 2015) found that scientific abstracts were rated more positively in a serif typeface than in a sans serif typeface when artificial typefaces were used, but the reverse was true when authentic typefaces were used. This suggests that other features can override any effect of the presence or absence of serifs.

One limitation of the latter study is that the ratings were provided by students and not by experienced teachers or researchers. Nevertheless, it raises the possibility that evaluations of academic writing may be influenced by readers’ preferences and the connotations of serif and sans serif typefaces. Section 9.2 discussed the possible implications of this notion in the context of reading from paper, but similar points can be made about reading on-screen:

  • In the context of academic publication, authors will want to be assured that their work is evaluated in terms of its content rather than in terms of its typographical appearance. The findings of Kaspar et al. (2015) indicate that on-screen evaluations of scientific abstracts can be influenced by the typeface in which they are presented. It is reasonable to assume that the same applies to on-screen evaluations of entire articles or books (although there seems to be no empirical evidence on this matter). One solution to this problem is for academic journals and publishers to require that manuscripts should be submitted for publication in a standard typeface so that they can be compared on a like-for-like basis.

  • In the context of academic assessment, it is possible that teachers and other assessors will give more positive on-screen evaluations of students’ online assignments if the teachers and their students share the same typographical preferences than if they differ in those preferences (although once again there seems to be no empirical evidence on this matter). It would be both useful and fairer in the interests of ensuring valid assessment if teachers responsible for particular course units (and, ideally, for entire degree programmes) could agree on their typographical preferences and make these known to their students.

Of course, in both contexts these variations in readers’ expectations and preferences might depend on their prior experience and familiarity with different typefaces rather than on any intrinsic properties of the typefaces themselves (and there is empirical evidence on this point in the case of reading from paper, if not in the case of reading from screens: see Sect. 9.2). Even so, there is a need for research on the extent to which reviewers’ on-screen evaluations of academic manuscripts and teachers’ on-screen evaluations of their students’ assignments are affected by the reviewers’ and teachers’ preferences and expectations.

3 Implications for Previous Assumptions

Where does this leave the recommendations of designers and design educators who have traditionally advocated the use of sans serif typefaces for material presented on-screen? Poncelet and Proctor (1993, p. 101) simply stated without qualification or supporting evidence: “Often san-serif fonts work better on the computer screen than serif fonts.” This assertion is clearly not supported by the research evidence reviewed in Part II. Similarly, Schriver (1997) claimed without qualification or supporting evidence that sans serif was “the preferred style of type for online because of its simple, highly legible, modern appearance” (p. 508). Her claims about the appearance of sans serif typefaces are partially supported by the results of research on the connotations of different typefaces, but they do not appear to translate into objective differences in their legibility.

Universal design is an approach to the development of educational websites that aims to ensure accessibility for all students instead of implementing ad hoc adjustments for those with particular disabilities. Proponents of universal design have advocated: “Use Sans Serif fonts for text. Letters with serifs are difficult to read on-screen and can create visual fatigue when large amounts of text are included on web sites” (“Universal Design”, 1999, p. 6). In general, there is no support for the idea that serif typefaces are harder to read on-screen than sans serif typefaces. One might expect that such effects would be more evident in readers with visual impairment under the high demands of the RSVP procedure, yet two studies failed to find any significant difference in performance between serif typefaces and sans serif typefaces in such readers. The quoted text refers specifically to material included on web sites, yet there is no evidence for differences between serif typefaces and sans serif typefaces when reading from internet browsers.

In fact, the notion of universal design has had its critics. For instance, Raymaker et al. (2019, p. 148) argued that it was

ultimately impractical due to the fact that access needs can conflict with each other. For example, some guidelines intended for people with intellectual disability . . . recommend simplifying vocabulary, which—if implemented without retaining the precision afforded by more complex wording—can make language pragmatics more difficult for autistic users to understand. . . . Likewise, high-contrast color schemes suitable for people with low vision may be painful or unreadable to autistic users with hypersensitive vision

Even so, Raymaker et al. concurred that websites designed for people with autism should “use a plain accessible sans-serif font (e.g., Arial) for ease of readability” (p. 147). Once again, they provided no empirical evidence to support this recommendation, and so we are back in the world of “everybody knows”.

4 Conclusions

This chapter concludes Part II by summarising and discussing the key findings. Studies of the legibility of letters and words when presented on computer monitors or other screens have failed to yield a consistent pattern of results, as have studies of the legibility of connected sentences, even when using the RSVP procedure. With regard to readers’ comprehension of meaningful text when presented on computer screens, the modal finding is that of no significant difference in reading speed between serif and sans serif typefaces. The studies that reported measures of accuracy in this situation failed to show a consistent pattern of results. Some studies have found that readers express a preference for sans serif typefaces, and there is some evidence that serif and sans serif typefaces differ in their connotations; however, these findings can be attributed to readers’ previous experience of reading text on-screen. Previously stated assumptions about the legibility of serif and sans serif typefaces when used to present material on computer screens are not supported by empirical research findings. In short, despite various assertions and recommendations, the conclusion of Part II is that there is no difference in the legibility of serif typefaces and sans serif typefaces when they are used to produce material that is presented on computer monitors or other screens.