Easier to read – easier to learn Impact of reader-friendly design on the intake and memorization of information in a classroom setting Herbert Schmidt August 2019 Abstract This study examines wether a well designed text aids pupils or students to read and understand the text and increase or ease the information intake. The study showed that a well designed text led to 18.8 % increase in reading speed, a 13.5 % increase in remembered information as well as an 32.1 % increase in testing time. The biggest difference however was the time needed to retrieve information from the text with a 36.9 % difference between a well designed text and a general teacher designed text. The study was conducted in several classrooms of a vocational school in Austria. Given the age bracket which is centered roughly around 17 and 18 year old students, the results could have a relevance for adult students as well. Motivation This study takes a look at how a reader-friendly-design of texts affects learning. In every learning environment written texts are a central part of the learning routine. As such it is of interest to establish a way in which the students can take the most from a text presented in written form. The aim of this study is to establish if a reader friendly design has an impact on learning as well as the understanding of a given text. This study was conducted at a school environment where subtle typographical nuances are irrelevant due to a lack of understanding on part of the teaching staff as well as lack of software which would make such distinctions possible in everyday work. Because of this some things had to be simplified to accommodate the reality of the workplace of teachers. Aims This study is aimed to offer a guideline on how a text should be designed to enhance the learning effect and to offer the easiest and most efficient design for a student to work with. It will show that a reader-friendly design is enhancing the learning as well as the efficiency of the progress. Setup The study was conducted in 2018 and 2019 at a vocational school in Vorarlberg / Austria. The system of apprenticeship in Austria is a two-folded system. The apprentices work at a company four days a week. There they receive the vocational training in an everyday working environment. Alongside this workplace education the apprentice is obliged to attend a vocational school one (in some cases two) day a week. The vocational school covers the theoretical aspects of the occupation and is usually organized in a two or three part system: Herbert Schmidt | mail@berufsschule.education 2/16 general and mercantile knowledge, theoretical knowledge and practical application. The apprentices receive grades and a yearly school report. The duration of most apprenticeships is three years. An apprenticeship is ended by passing an exam conducted by the Austrian Federal Economic Chamber. The study incorporated three vocations: • Retailers • Wholesalers • Administrative and office clerks Definitions Legibility Legibility is defined as the ease of recognition of each letter of a text. The choice of font, paper colour, size, kerning etc. all are influencing legibility. As legibility is partly responsible for the ease of reading a text, certain elements are incorporated in this study. The term readability thus includes – in this study – also elements of legibility. Readability Readability in this study is defined as the ease of comprehension of a text. As such it is the ease of deciphering, reading and realization of a text in regards to the font, height, spacing and other design elements. Traditionally all this is encompassed by readability and legibility in typography. In this study this term is used in this encompassing sense, including all elements of the design of a text. The content and phrasing is not part of this study. The author is fully aware that the way a text is phrased has a big influence on the ease of understanding a text. Reader-friendly phrasing is a separate field and needs to be studied on its own. The combinatorial effects would certainly be significant. This, however, is not part of this study. There have been many studies about readability. From a purely technical viewpoint Willberg1 notes the following criteria: • clear and unambiguous letter forms which lead to • clear and unambiguous word forms • aiding the eye to remain on the line • kerning (the distance between letters) should be carefully considered • correct use of typographic attributes and/or highlighting • consciously designing the empty parts of the page 11 Willberg 2001, 2005 Herbert Schmidt | mail@berufsschule.education 3/16 • the space between the words has to be smaller than the space between the lines How typography as well as language usage should be adjusted depends on the type of reading that is used. Willberg and Forssman distinguish between the following types (translation by the author) lineares Lesen – linear reading informierendes Lesen – reading to inform, informative reading konsultierendes Lesen – reading to consult, consulting reading selektierendes Lesen – selective reading Their other differentiations are of no relevance to this study. Each of these reading types call for different typographical designs. In this study the text will be read first in linear reading and after that again in a consulting reading. The students will read the text (linear reading) to understand the content. After that they read it again to help answer the questions (reading to consult). Herbert Schmidt | mail@berufsschule.education 4/16 The definition of reader-friendly design There doesn’t seem to be a clear and accepted definition of a reader-friendly design. Different authors argue for different design details. For this study the included design requirements are either agreed upon by most experts or that are falling in line with what is considered to be good typography. These design requirements are as follows: Line Length The length of the individual line should not exceed 60 characters or 130 mm. Distance of Lines The distance between the lines should be greater the longer the line is.2 Usually 1.2 line spacing has been established as a good rule. As most word processing software uses increments of lines, this is what is used here. Font Type There is an ongoing discussion wether serif fonts or grotesque fonts (without serifs) are easier to read. Up until now there seems to be no consensus. This study follows the standard advice that serif fonts are easier to read than grotesque because the serifs aids the eye in staying on the line.3 Even modern studies seem to support this view.4 So serif fonts, or Antiqua fonts are to be preferred. Font Size Given the right font type a font size of 10 to 12 pt. is to be preferred. In this study 12 pt. is preferred to minimize any problems in regards to light or visibility. Text Alignment A left alignment with a ragged margin is usually the easiest to read. A forced block is not helpful as the distance between the words and sometimes the letters are forced and uneven. The best would be a block that allows some minor edge but this can only be done with professional software. As this is not being used by teachers it can not be part of the setup. Hyphenation – Syllabification In a reader-friendly design hyphenations should be avoided. Words should not be separated. 2 Willberg, 2001 3 Tschichold, 2001; Kunz, 1950 4 van Rossum, 1998 Herbert Schmidt | mail@berufsschule.education 5/16 Clear Headlines Headlines give structure to a text and makes it easy for students to find relevant passages. Headlines in the text should be roughly 30 % larger than the text itself. This study uses headlines of 16 pt. which are set in bold. After the headline a space line is inserted. The different designs used in the study The different designs are specified as follows: Reader-friendly Design (RFD) Margins: top, left and right: 3 cm, bottom: 4 cm Font Type: Garamond Font Size: 12 pt. Line Spacing: 1.2 Line Length: 130 mm Text Alignment: left with ragged margin Hyphenation: none Headlines: 16 pt., bold with space line Non reader-friendly Design 1 (NRFD1) Margins: 2 cm Font Type: Arial Font Size: 12 pt. Line Spacing: 1 Line Length: full length Text Alignment: left with ragged margin Hyphenation: automatic Headlines: 12 pt., bold Non reader-friendly Design 2 (NRFD2) Margins: 2 cm Font Type: Comic Sans Font Size: 12 pt. Line Spacing: 1 Line Length: full length Text Alignment: forced block Hyphenation: automatic Headlines: 12 pt., bold Herbert Schmidt | mail@berufsschule.education 6/16 Teacher generated Design (TGD) The fourth design is a design that was done by a teacher without any previous instructions. The teacher was being asked to design the text in the same way that is used in everyday teaching. This design is therefore not foreseeable but nevertheless interesting because it constitutes an example from the working environment. Similar designs are being used by a number of teachers at the chosen school. In total there are four designs, one reader-friendly design (RFD) and two different non-readerfriendly designs (NRFD1 + NRFD2) as well as one teacher generated design (TGD). The text The main problem of the study was to eliminate all possible influences. To achieve this, the texts had to be suitable for the age group of the participants. The content had to be new for all students otherwise the result would have been biased because of the previous knowledge of some students. Any increase of knowledge could then not be attributed clearly to working with this text. To avoid this bias, the text had to be as neutral as possible and the topic unknown to the students. In each run, preliminary questions assured that the topic as well as the content was unknown for all participants and that there was minimal previous knowledge. If there would have been previous knowledge or any strong emotional responses, either positive or negative, the text as well as the topic was discarded. It proved to be very difficult to find texts which offered the least number of possible biases. The difficulty of the text. To ensure that the level of the text itself wasn’t a problem, the text had to be suitable for the age group. Best of all, it would be suitable for a slightly younger age group because even students with reading difficulties could participate. The text was tested to be of the appropriate level for the students. The text difficulties were established using the following three methods: • Flesch-Kincaid • LIX • Wiener Sachtextformel The tools used to establish these level of the texts were a program (LIX-Rechner) and an online tool for writers. 5 5 Lix-Rechner and http://www.schreiblabor.com/textanalyse/ Herbert Schmidt | mail@berufsschule.education 7/16 Flesch-Kincaid: The Flesch-Kincaid formula was developed by Rudolf Flesch and J. Peter Kincaid. Their formula is one of the most widely used methods to establish the reading difficulty of a text.The Flesch-Kincaid formula used was adapted for the german language to accommodate the fact that there are longer words and a different syntax.6 The levels are given as follows (translation by the author): over 80: extremely easy (appropriate for 11 year old pupils) 71 – 80: very easy 61 – 70: easy 46 – 70: average (appropriate for 13 – 15 year old pupils) 36 – 45: difficult below 35: very difficult (academic) LIX: The LIX uses the method of Björnsson7 to establish the reading difficulty of a text. The LIX scale is set as follows: below 40: childrens literature 40 – 50: fiction 50 – 60: non-fiction over 60: academic Wiener Sachtextformel (WSTF): The Wiener Sachtextformel was established by Richard Bamberger and Erich Vanecek. This formula lists for which grade a text is suitable. Thus the WSTF lists grades instead of difficulties. Above the grade 12 it would be better to see it as grades of difficulties rather than actual school grades. Thus a level of WSTF 4 is a very easy text, whereas a level of WSTF 15 would be a quite difficult one. The grades of the vocational school would be 10 to 13. In reality the grades need to be set lower because a lot of the students are on a lower reading level. Thus taking the numerical correct grade would be too difficult and thus distort the result. A grade of WSTF 9 would be more than appropriate. 6 The exact formula can be found on several homepages as well as in different publications. 7 Björnnsen 1968, more on: https://www.psychometrica.de/lix.html Herbert Schmidt | mail@berufsschule.education 8/16 The Study The chosen text was designed in the four designs RFD, NRFD1, NRFD2 and TGD (definitions see above). The TGD was done by a teacher who was not involved in the study. She was simply being asked to design the text the way she usually does. This serves as a sort of control group (or control design) because the design is not done with a certain goal as the teacher has no design or typographic background. Additionally, as the TGD design reflects usual practice it serves also the double purpose to compare the RFD with a normal standard design. The goal of the study as well as the procedure were explained to the students. The texts were distributed in class at random together with an answer sheet with 4 questions and 2 QRCodes. The QR-Codes were labelled 1 and 2. The four questions on the questionnaire were: • Are you male or female. • How old are you. • How long did you take to read the text. • Did you find the text… easy difficult hard The first two questions allow a grouping of results. The age also serves as a benchmark for the correct level of readability of the text. The third question is directly linked to the readability. The easier to read, the shorter the reading time should be. The last one is a purely subjective and can’t be the basis of a judgment for the text. It is however a perceived level of difficulty which could hint at how the reading difficulty was experienced. This question also engages the student and makes clear that it is a judgement of their own abilities. The students were familiarized with the procedure to ensure that any time differences were not due to technical problems or because of an unclear understanding. 1. The students read the text in one go. 2. They used a stop-watch app on their mobile phones to mark the time. 3. After that they scanned QR-Code 1 and answered the questions anonymously. 4. After having answered QR-Code 1, they scanned the second code. This led them to the same questions again but this time they used the text to find the answers. This setup allows for the extraction of the following data. • How long did it take to read the text with the given design. –> reading speed • How long did it take to answer the first quiz. –> time of testing Herbert Schmidt | mail@berufsschule.education 9/16 • How well were the questions answered. –> knowledge intake, how well was the content remembered • How long did it take to answer the second quiz. This shows how fast and easy the relevant information was found in the text. This is direct evidence for the ease of navigating the text. –> information retrieval • Are there any differences in the number of correctly answered questions when using the text. –> quality of information retrieval Herbert Schmidt | mail@berufsschule.education 10/ 16 Results There have been 134 participants in total. The age of the subjects participating in the study ranged from 15 to 25 years with half of the participants being 17 and 18 years old. overview age of participants number age 12 15 19 16 35 17 32 18 9 19 8 20 6 21 4 22 6 23 1 24 1 25 0 26 1 27 The results in seconds and percentage: measured NRFD 1 NRFD 2 TGD RFD reading time in seconds 428 404 413 360 time spent on test without consulting the text 220 226 200 171 time spent on test with consulting the text 260 275 356 277 % of correct answers from memory – without consulting the text 64.8 62.1 58.5 72 % of correct answers with consulting the text 73.6 76.8 77.7 75.35 Herbert Schmidt | mail@berufsschule.education 11/ 16 Interpretation of results Reading speed When working in a classroom environment time is a scarce resource. Wether students read faster does make a difference, especially with long texts. The biggest difference was 18.8 % faster compared with the NRFD 1 and 14.7 % in regards to the TGD. Thus with a reader friendly design texts would be read almost 15 % faster than with the design currently used and almost 19 % faster than with a badly designed text. reading time in seconds 430 428 404 410 413 390 370 350 360 NRFD 1 NRFD 2 TGD RFD % of correct answers from memory – without consulting the text Information remembered Reading fast is on only desirable if the 80 information is remembered well. In the first test 72 72,5 students answered purely from memory. As we 64,8 excluded previous knowledge the results are a 65 62,1 58,5 good indicator wether they remembered more 57,5 from a reader friendly designed text. 50 The results are clear: 72 % of correct answers NRFD 1 NRFD 2 TGD RFD after having read a reader friendly designed text compared to 58.5 % after having read the teacher designed text. This hints at a better information retention purely through better text design. As the text itself was the same and only the design differed this difference of 13.5 % is quite significant. Time of testing Not only did the students who worked with the reader friendly design read faster (by 18.8 %) and had better results (by 13.5 %) but they also finished the test faster. This hints at the availability of the remembered information. time spent on test without consulting the text 250 226 212,5 200 171 175 137,5 100 Herbert Schmidt | mail@berufsschule.education 220 NRFD 1 NRFD 2 TGD RFD 12/ 16 The biggest difference was 32.1 % in comparison with the NRFD 2. Compared with the TGD there was still a difference of 16.9 %. This could be due to the students actively trying to remember the contents. time spent on test with consulting the text Time of information retrieval When students were allowed to consult their text for information retrieval the students working with NRFD 1 were the fasted followed closely by NRFD 2 and RFD. The students working with TGD where 36.9 % slower than the other designs. 356 370 327,5 285 260 277 275 242,5 200 NRFD 1 NRFD 2 TGD RFD Result of information retrieval Interestingly almost all students received similar results once they were allowed to consult the test. The biggest difference amounts to no more than 5.5 % (TGD). The percentage of correct answers ranged from 73.6 % to 77.7 %. Thus it can be safely said that the design is not of relevance in respect of the quality of information retrieval when the text can be used to answer questions or in other words: when information has to be retrieved actively. Herbert Schmidt | mail@berufsschule.education % of correct answers with consulting the text 80 73,6 76,8 77,7 75,35 72,5 65 57,5 50 NRFD 1 NRFD 2 TGD RFD 13/ 16 Conclusion Reader friendly designed texts proved to be faster to read (18.8 %), better remembered (13.5 %) and they tested faster (32.1 %). Information retrieval from the text was up to 36.9 % faster from a reader-friendly designed text than from a teacher designed text. In regards of the quality of information retrieved from the text there was no statistically significant difference. Apparently the different designs were no hindrance to retrieving specific information although with the teacher designed text it took almost 37 % longer. The results of the quality of information retrieved though was similar. This clearly shows that a reader friendly design not only helps students in reading but also in remembering the information. They also test faster. This could be a bonus especially with longer tests. As the documentation provided by the teachers is the basis to learn, this could make a noticeable difference. Ease of learning directly encourages learning and should be considered when designing texts for classroom use. When working with the text, the reader friendly designed texts proved to be easier to read, faster to use and helped in remembering the information. It is interesting that the texts designed in a standard teacher design came in last in almost all respects. This clearly shows that there is not enough emphasis on this point in the education and training of teachers. I would strongly encourage teachers, educators and school administrators to take a look into reader friendly design. I am quite sure that additional effects can be achieved by reader friendly phrasing. Herbert Schmidt | mail@berufsschule.education 14/ 16 Bibliography Bailin Alan and Grafstein Ann: Readability: Text and Content, 2016, Palgrave Mc Millan Burkard, Andreas: Lesefreundliche Textgestaltung, Verband der Schweizer Druckindustrie Dillon, A.: Reading from paper versus screens: a critical review of the empirical literature. Ergonomics, 35(10), 1297-1326., 1992 Forssman Friedrich, de Jong Ralf: Detailtypografie, 2002, Verlag Hermann Schmidt, Mainz Hiebert, Elfrida H.: The case for reader friendly articles, Educational Leadership, Vol 7, 2013 Kunz Hannes: Die künstlerische Schrift, 1950, Verlag Oskar Sachsperger Lukes Dominik: Dyslexia Friendly Reader: Prototype, Designs and Exploratory Study, 2015 van Rossum, M. C. W.: Measuring Font Legibility, Departement of Neuroscience, 1998, University of Pennsylvania Rottensteiner Sylvia: Structure, function and readability of new textbooks in relation to comprehension, 2010, Free University of Bolzano, Italy Swanson Gunnar: Reader Centered Design, 2001 Singer Trakhman Lauren, Alexander Patricia: Reading on Paper and Digitally: What the Past Decades of Empirical Research Reveal, Review of Educational Research Month 201X, Vol. XX, No. X, pp. 1–35 DOI: 10.3102/0034654317722961 © 2017 AERA. http://rer.aera.net Tarasov Dmitry A., Sergeev Alexander P.: The leading factor of readability: development of the methodology for educational use, 2013, 4th International Conference on New Horizons in Education Tschichold Jan: Erfreuliche Drucksachen durch gute Typographie, 2001, MaroVerlag, Augsburg (reprint of 1988) Umit Izgi, Burcu Sezginsoy Seker: Comparing different readability formulas on the examples of science-technology and social science textbooks, 2012 Herbert Schmidt | mail@berufsschule.education 15/ 16 Willberg Hans Peter: Wegweiser Schrift, 2001, Verlag Hermann Schmidt, Mainz Willberg Hans Peter, Forssman Friedrich: Lesetypo, 2005, Verlag Hermann Schmidt, Mainz Williams Sandra; Reiter Ehud and Osman Liesl: Experiments with discourse-level choices and readability. 2003, In: Proceedings of the 9th European Workshop on Natural Language Generation (ENLG-2003), 13-14 April 2003, Budapest, Hungary, pp. 127–134. Herbert Schmidt | mail@berufsschule.education 16/ 16