Research Findings

To gain new knowledge on how the perceptual system functions when presented with graphic communication, we need to isolate individual visual variables for investigation. This is a difficult task as different elements in design tend to influence each other. None the less, the main gold of the Centre is to identify the effect that each element has on perception.

Knowing the function of the parts, informs us of the function of the whole.

Below we show various examples of our findings.

The optimal letter boldness for fonts at small sizes or great reading distances 

We have shown that at both small and large point sizes, light and ultra-black fonts were inferior to all the fonts in the middle of the weight scale. Further, the bolder weights in the middle of the scale enhanced recognition in the small font size, while failing to do so in large font size. 

These findings provide evidence that under reading situations involving small font sizes and signs read form great reading distances, bolder weights will improve letter recognition. They also showed that both light and ultra-black font weights should be avoided in any case where letter recognition is a priority. 

For more see:

Beier, S. & Oderkerk, C.A.T. (2019). ‘Smaller Visual Angles Show Greater Benefit of Letter Boldness than Larger Visual Angles’. Acta Psychologica 199 (August): 102904. 1–8.

The legibility of numerals

By varying the skeleton of the digits we identified versions of the digits ‘1’, ‘3’, ‘7’ and ‘9’ that are more legible than other versions. 

To enhance the peripheral legibility of numerals we designed three versions of the digits from 1 through 9 by modifying their skeletons while controlling for variations of other physical parameters. Observers identified the different versions of the digits in random three-digit strings, presented within their peripheral visual field. 

Our results showed that the digit ‘1’ should have a narrow design without a crossbar at the bottom, the digits ‘3’ and ‘9’ benefit from open apertures, and the digit ‘7’ should have a straight leg and no serif at the horizontal bar. 

For more see:

Beier, S., Bernard, J.B. & Castet, E. (2018) ‘Numeral legibility and visual complexity’, Proceedings of DRS2018, Design Research Society, vol. 5, 1841-1854, Limerick, 25th-28th June.

Character complexity

The simpler the character skeleton, the more legible the character. This rule of thumb only applies when the shape of the character does not increase the rate of misreadings. 

In an investigation of the legibility of digits in a short exposure of three-digit strings at the peripheral visual field, we confirmed previous findings (Bernard & Chung 2011) that digits and letters benefit from having the most simple shapes. In other words, the shorter the morphological skeleton, the more legible the character. 

These results somewhat contradict the approach applied by several renowned type designers whose focus on ensuring differentiation between characters may result in added features, such as cross bars and tails. For the London Underground typeface, for example, Edward Johnston created a loop in the lowercase ‘l’ to differentiate it from the capital ‘I’. This resulted in a more complex letter skeleton, which in theory would lower legibility. Our own studies showed that at greater reading distances, a tail on the ‘l’ results in fewer errors, that a cross bar on the letter ‘i’ can improve legibility, and that the letter ‘a’ should have a two-storey design. Following this, we conclude that the advantage of the shorter morphological skeleton only applies in situations where the simple character skeleton will not result in a greater number of misreadings. 

For more, see: 

Beier, S., Bernard, J.B. & Castet, E. (2018) ‘Numeral legibility and visual complexity’, Proceedings of DRS2018, Design Research Society, vol. 5, 1841-1854, Limerick, 25th-28th June. 

Beier, S. & Larson, K. (2010) ‘Design Improvements for Frequently Misrecognized Letters’, Information Design Journal, 18(2), 118-137.

The role of serifs at a distance

At great reading distances, serifs at the vertical extremes of letters can improve the legibility of the individual character. 

The experiment looked into the distance identification of the lowercase letters ‘j’, ‘i’, ‘l’, ‘b’, ‘h’,

‘n’, ‘u’ and ‘a’ in isolation. All the letters originate in the same typeface and are presented in one version with serifs and one version without serifs. 

The investigation found no difference in the distance legibility of sans versus serif typeface styles when we studied the collective results of the individual letters. 

However, looking at a group of letters with serifs at the vertical extremes, the experiment showed higher distance legibility when these letters have serifs on the stems; this demonstrates that having serifs on the stem plays a central role in facilitating legibility at great distances. 

The data further indicates that serifs on the counter of ‘h’ and on both ends of the stem of ‘i’ cause a higher rate of misreadings, where these letters are misread as ‘b’ and ‘l’, respectively. 

The findings as a whole suggest that serifs should not always be applied in the conventional fashion, as in traditional Old Style and Didone typefaces; instead they can facilitate higher distance legibility if they are placed in a semi-serif fashion, that is, at relevant stroke endings. 

For more see:

Beier. S., & Dyson, M. (2014) ‘The influence of serifs on 'h' and 'i': useful knowledge from design-led scientific research’, Visible Language, 47(3), 74-95. 

Typeface familiarity

After a short practice session, readers will read an uncommon typeface at the same rate as a very common typeface. However, they do not like doing so.

 Some typographers have proposed that typeface familiarity is defined by the amount of time that a reader has been exposed to a typeface, as famously stated by Zuzana Licko: ‘People read best what they read most.’

 Other typographers have proposed that familiarity is defined by commonalities in letter shapes. One example of this is Adrian Frutiger’s letterform matrix model. Frutiger argued that every character has a basic skeleton that is based on a collective memory of all the different character variations a person has ever encountered. The skeleton emerges when widely read typefaces are superimposed, so that the parts of the letterform that are shared across all typefaces appear.

We tested these two hypotheses by measuring the reading speed and preferences of participants. Participants were tested twice with common and uncommon letter shapes, once before and once after spending 20 minutes reading a story presented in the given font.

The results indicate that exposure time does impact reading speed, so that the uncommon typefaces were read almost as fast as the common typefaces. The revelation that unusual letterforms do not slow down reading after a 20-minute exposure period surprisingly tells us that the presence of common letterforms in typefaces is not important to reading performance. The outcome of the reading speed test supports the view of proactive designers advocating improvements in letter shapes. 

Readers’ opinions, on the other hand, support the argument of traditionalist designers by demonstrating that readers were noticeably more critical of the fonts that had uncommon letterforms, compared to the fonts that had common letterforms. Readers simply did not enjoy reading the uncommon type. When asked to rate statements such as ‘I will enjoy reading this typeface in the future’, both before and after the practise sessions, the subjects’ answers were negative for the uncommon fonts, even when their reading speed proved unencumbered.

 Based on these findings, we conclude that the reason letter skeletons have changed so little over time has to with readers’ subjective opinions, not with legibility. In a normal reading situation outside of the laboratory setting, readers simply stop reading if the situation is uncomfortable, and thus their opinions of the reading situation will end up overruling any possible advantages in reading speed. 

For more see:

Beier, S. & Larson, K. (2013) ‘How does typeface familiarity affect reading performance and reader preferences?’, Information Design Journal, 20(1), 16-31.

Italics are subtler than bold for emphasis

We have confirmed the functions of the widespread approach of highlighting text with italics and using bold in headlines. 

This study looked at the effect of weight, width, contrast and italics in combination with the regular style.

Previous studies have found a ‘regularity effect’, in that readers have a faster reaction time if the stimulus is presented in the same font throughout than when it is presented in a mix of different styles. 

The objective was to determine which stylistic features might produce a regularity effect and, further, to explore whether the incidence and extent of the effect are dependent on the nature of the visual and perceptual differences between the fonts. 

While the previously described experiments with font tuning and the regularity effect compared different typefaces or highly unusual typefaces, this investigation dealt with fonts that can be described as a typeface family. This is similar to a layout situation, where designers mix regular styles, bold styles and italics. 

In three experiments we investigated whether there was a regularity effect in words using a set of fonts that vary systematically from a neutral test font. We used Rapid Serial Visual Presentation (RSVP), where a string of letters are displayed on a screen, presented sequentially in the same position. The task in this mode was to determine whether the stimulus was a word or non-word.

The findings suggested the following and thus lent support to what is a common approach among graphic designers. 

  • Words presented in a regular style are a good basis for efficient letter recognition, while expanded fonts hamper legibility. 
  • Setting words in italics serves to distinguish text elements without significantly disrupting reading and is therefore suitable for use in continuous text. 
  • Bold is more appropriate than italics for headings by making words stand out. 

For more see:

Dyson, M. & Beier, S. (2016) ’Investigating typographic differentiation: Italics are more subtle than bold for emphasis’, Information Design Journal, 22(1), 3-18. 

One-storey or two-storey ‘a’

Our findings show that adult readers often misread the one-storey ‘a’ as ‘o’ or ‘q’. For maximum legibility, we recommend applying a two-storey ‘a’. 

An often-voiced hypothesis among designers is that a one-storey ‘a’ is less legible than a two-storey ‘a’. This assumption has been borne out by our research. The findings demonstrated that at great reading distances and at short exposures in the peripheral visual field, a one-storey ‘a’ frequently results in misreading, where the ‘a’ is confused with the characters ‘o’ and ‘q’. We recommend against the one-storey ‘a’ when setting text for adult normal-vision readers. 

For more, see:

Beier, S. & Larson, K. (2010) ‘Design Improvements for Frequently Misrecognized Letters’, Information Design Journal, 18(2), 118-137.


This website uses cookies

KADK uses cookies to create a better user experience, to interact with social platforms and for anonymised statistics of traffic on our website.

Social media cookies enable us to interact with well-known social media platforms and content. This may be for statistical or marketing reasons.
Neccesary to display YouTube videos
Statistics cookies collect anonymised information that helps KADK to understand how visitors use our website and to provide users with an optimal experience.
Preference cookies enable a website to remember information that changes the way the website behaves or looks, like your preferred language or the region that you are in.
Is used for UI states