Good Fonts for Dyslexia

1 Good Fonts for DyslexiaLuz RelloNLP & Web Research GroupsUniversitat Pompeu FabraBarcelona, Baeza-YatesYahoo! Labs &Web Research Group, UPFB arcelona, 10% of the people have Dyslexia , a neurological dis-ability that impairs a person s ability to read and is evidence that the presentation of the text has asignificant effect on a text s accessibility for people withdyslexia. However, to the best of our knowledge, there areno experiments that objectively measure the impact of thefont type on reading performance. In this paper, we presentthe first experiment that uses eye-tracking to measure theeffect of font type on reading speed. Using a within-subjectdesign, 48 subjects with Dyslexia read 12 texts with 12 dif-ferent serif,monospacedandromanfont stylessignificantly improved the reading performance overserif,proportionalanditalicfonts. On the basis of our results,we present a set of more accessible Fonts for people , font types, typography, readability, legibility, textlayout, text presentation, INTRODUCTIONW orldwide, around 15-20% of the population has a languagebased learning disability [17].

2 Likely, 70-80% of them havedyslexia [17], a neurological disability which impairs a per-son s ability to read and write. Previous research has shownthat text presentation can be an important factor regardingthe reading performance of people with Dyslexia [11, 25].On the other hand, any digital text has to be written usingone or several certain font types. Although the selectionof font types is crucial in the text design process, empiricalanalyses of reading performance of people with Dyslexia hasfocused more on font size [23, 26] rather than on font this paper we present the first study that measures theimpact of the font type on the reading performance of 48people with Dyslexia using eye-tracking, as well as askingthem their personal 2013 Bellevue, Washington, USAThe main contributions of this study are: Font types have a significant impact on readability ofpeople with Dyslexia . Good Fonts for people with Dyslexia areHelvetica,Courier,Arial,VerdanaandCom puter Modern Uni-code, taking into consideration reading performanceand subjective preferences.

3 On the contrary,Arial be avoided since it decreases readability. Sans serif,romanandmonospacedfont types increasedthe reading performance of our participants, whileitalicfonts did the section focuses on Dyslexia , while Section 3 reviews re-lated work. Section 4 explains the experimental methodol-ogy and Section 5 presents the results, which are discussedin Section 6. In Section 7 we derive recommendations fordyslexic-friendly font types and we mention future lines DYSLEXIAD yslexia is ahiddendisability. A person with Dyslexia can-not perceive if they are reading or writing correctly. Dyslexiais characterized by difficulties with accurate word recogni-tion and by poor spelling and decoding abilities [16]. Thisimplies that people with Dyslexia have more difficulty access-ing written information and, as side effect, this impedes thegrowth of vocabulary and background knowledge [16]. Pop-ularly, Dyslexia is identified with its superficial consequences,such as writing problems like letter reversals; but Dyslexia isa reading disability with a neurological origin.

4 Brain struc-ture, brain function, and genetics studies confirm the bio-logical foundations of Dyslexia [31].1 Although Dyslexia isalso popularly identified with brilliant famous people, suchas Steve Jobs or Steven Spielberg, the most frequent way todetect a child with Dyslexia is by low-performance in school[4]. Moreover, Dyslexia is frequent. From 10 to ofthe population in the [15] and from to 11% of theSpanish speaking population [18] have this cognitive disabil-ity. The frequency and the universal neuro-cognitive basisof Dyslexia are the main motivations of this its universal neuro-cognitive basis, Dyslexia mani-festations are variable and culture-specific [31].3. RELATED WORKThe relationship between Fonts and Dyslexia has drawn theattention of many fields, such as psychology, arts, and acces-sibility. We divide related work in: (1) Fonts recommendedfor people with Dyslexia , (2) Fonts designed for this targetgroup, and (3) related user RecommendationsMost of the recommendations come from associations forpeople with Dyslexia and they agree in usingsans-seriffonts.

5 The British Dyslexia Association recommends to useArial,Comic Sansor, as alternatives to these,Verdana,Tahoma,Century Gothic, andTrebuchet[2]. However, thewebsite does not disclose on the basis of which evidence theserecommendations are made. In [10] recommendations forreaders with low vision as well as readers with Dyslexia areput in comparison, giving as a result the recommendation ofusing alsoArialandComic Sans. In [22] is recommended toavoid italics and fancy Fonts , which are particularly difficultfor a reader with Dyslexia , and also point toArialas pre-ferred font. Another font recommended in 2010 wasSassoonPrimarybut not anymore [9].The only recommendation forseriffonts has been doneby the International Dyslexia Centre [13] and that was forTimes New Roman. According to [1],Courieris easier toread by people with Dyslexia because it is theWeb Content Accessibility Guidelines(WCAG) [3], Dyslexia is treated as part of a diverse group of cognitivedisabilities and they do not propose any specific guidelinesabout font types for people with , none of the typefaces recommended by thedyslexia organizations mentioned above were ever designedspecifically for readers with Fonts Designed for People with DyslexiaWe found four Fonts designed for people with Dyslexia :Sylex-iad[12],Dyslexie[21],Read Regular, four Fonts have in common that the letters are moredifferentiated compared to regular Fonts .

6 For example, theshape of the letter b is not a mirror image of d . Fromthese Fonts , we choose to studyOpen Dyslexic(both romanand italic styles), because it is the only open sourced andhence free. This font has been already integrated in User StudiesThere are several uses studies on text presentation and peo-ple with Dyslexia regarding font and background colors [25],font [23, 26] or letter spacing [33].The closest work to ours is a study with people with Dyslexia [21] that comparedArialandDyslexie. They conducteda word-reading test with 21 students with Dyslexia (DutchOne Minute Test).Dyslexiedid not lead to faster reading,but could help with some dyslexic-related errors in [29], text design for people with Dyslexia is explored witha qualitative study with just eleven students. In some tasks,2 participants needed to choose the font they prefer, butno analyses of the chosen Fonts is What is Missing?What is missing is an objective investigation into the effect ofthe most frequent Fonts on reading performance.

7 Our exper-iment advances previous work by providing this evidence viaquantitative data from eye-tracking measurements. In ad-dition, with testing 12 different Fonts with 48 participants,we compare a greater number of font types with a largernumber of participants than previous studies. We selectedthe Fonts on the basis of their popularity and frequency ofuse in the METHODOLOGYTo study the effect of font type on readability and com-prehensibility of texts on the screen, we conducted an ex-periment where 48 participants with Dyslexia had to read12 comparable texts with varying font types. Readabilityand comprehensibility were analyzed via eye-tracking andcomprehension tests, respectively, using the latter as a con-trol variable. The participants preference was gathered DesignIn our experimental design ,Font Typeserved as aninde-pendent variablewith 12 levels:Arial,Arial Italic,Com-puter Modern Unicode (CMU),Courier,Garamond,Hel-vetica,Myriad ,OpenDyslexic,OpenDyslexic Italic,Times,Times Italic, andVerdana(See Figure 1).

8 We use for brevityOpenDysfor the corresponding Fonts in the rest of the is ArialThis is Arial is Computer ModernThis is CourierThis is GaramondThis is HelveticaThis is MyriadThis is OpenDyslexicThis is OpenDyslexic is TimesThis is Times is VerdanaFigure 1: Fonts used in the chose to studyArialandTimesbecause they are themost common Fonts used on screen and printed texts, re-spectively [5].OpenDyslexicwas selected because is a freefont type designed specifically for people with Dyslexia andVerdanabecause is the recommended font for this targetgroup [2]. We chooseCourierbecause is the most commonexample of monospaced font [5].HelveticaandMyriadwerechosen for being broadly used in graphic design and for beingthe typeface of choice of Microsoft and Apple, choseGaramondbecause is claimed to have strong legi-bility for printed materials [5] and we selectedCMUbecauseis widely used in scientific publishing, as is the default of thetypesetting program TeX, as well as a free font supportingmany languages [20].

9 We also made sure that the Fonts cover variations of essentialfont characteristics: Italicsserved as independent variable with two values:italicdenotes the condition where the text was pre-sented using an italic type, that is a cursive typeface,andromandenotes the condition when the text waspresented in a roman type. We study the italic typesofArial,OpenDyslexic, andTimes. Serifserved as independent variable with two values:serifdenotes the condition where the text was pre-sented with typefaces with serifs, small lines trailingfrom the edges of letters and symbols, andsans serifdenotes the condition when the text used typefaceswithout serifs. In our set of Fonts there are threeseriffonts CMU,Garamond, andTimes and foursansseriffonts Arial,Helvetica,Myriad, andVerdana . Monospaceserved as independent variable with twovalues:monospaceddenotes the condition where thetext was presented using a monospaced type, that is, afont whose letters and characters each occupy the sameamount of horizontal space, andproportional, wherethe text was presented using proportional Fonts .

10 Wechose the most commonly used monospaced font, theroman seriffontCourier, and we compare it with therest of theromanandseriffonts that areproportional:CMU, quantifying readability, we used twodependent mea-sures:Reading TimeandFixation duration, both extractedfrom the eye-tracking data. To control text comprehensionof the texts we use onecomprehensionquestion as acon-trol variable. To collect the participant preferences, we usedsubjectivePreference Ratingsthrough Time:Shorter reading durations are preferred tolonger ones since faster reading is related to more readabletexts [32]. Therefore, we useReading Time, the time ittakes a participant to completely read one text, as a measureof readability, in addition toFixation Duration:We used fixation duration as an ob-jective approximation of readability. When reading a text,the eye does not move contiguously over the text, but alter-nates saccades and visual fixations, that is, jumps in shortsteps and rests on parts of the durationde-notes how long the eye rests still on a single place of the textand we use the mean of the fixation durations obtained bythe eye-tracker.