Research on Universal Design fonts for the visually impaired and the elderly

Research on Universal Design fonts for the visually impaired and the elderly

In order to pursue further possibilities for fonts and to seek the optimal fonts for the visually impaired and the elderly, Fontworks Inc. "Universal Design ( Universal Design) Fonts" was researched. Report its contents.

From the viewpoint of universal design Universal Design we conduct research on fonts and "Universal Design fonts that are effective designs for everyone" We have been developing with the aim of. Through this process, we came to the idea that rather than developing only one design that is effective for everyone, wouldn't it be better to change the character design to make it the optimal UD font for each specific group so that we can deliver the optimal font to the people in the whole group?

In the first study, we conducted an experiment on an elderly group, concluded that "there is not much difference between the elderly and the young," and released Universal Design font of our company. Following the idea I mentioned earlier, instead of saying that there is not much difference, we need to deal with small differences, so we decided to work on it from here.
In addition, in the discussion, there was an opinion that "it is better to incorporate not only the elderly but also the visually impaired", so we are researching and developing Universal Design fonts for the visually impaired and the elderly.

Research overview

1. Font comparison experiment

We conducted two experiments.

・ Readability evaluation experiment
Evaluate fonts in terms of "readability".

・Discrimination evaluation experiment
Fonts are evaluated from the viewpoint of "ease of distinguishing" characters that are easy to misread.

This research For the visually impaired and the elderly, "seeking the best font from among various fonts" for the purpose of "Evaluate mainly readability while considering discriminability" It was made.
We asked Mirairo Co., Ltd. and Do House Co., Ltd. to select the subjects.

We have a large selection of fonts with rich designs. In other words, the number of experimental trials to select just one font from all fonts is very large.
On the other hand, there was also the problem of conducting long-term experiments on visually impaired people. Therefore, in order to reduce the number of trials, design typefaces were removed from all fonts, and general Mincho, Gothic, and Maru Gothic fonts were used as test fonts.

It actually covers the following fonts:

Due to the pandemic, instead of face-to-face experiments, we created a web application for the experiment and conducted an online experiment by running it on the PC or tablet owned by the subject. The character display of the experimental application uses web fonts.

In addition, the following constraints were set for the subjects in the experiment.

・It is possible to answer from a tablet or PC (smartphones are not possible)
・ Participate in a bright room (no brightness setting)
・I don't normally browse in black and white reversed state.
・It is possible to view the text information displayed in the designated application without enlarging or reducing it.
・The distance from the screen should be 30 to 40 cm.

Font comparison experiment summary

(1) Differences from young people
In the previous research (joint research report vol.1 with Kyushu University ~Evaluation of universal design (Universal Design) fonts~), we focused on the difference in typeface, so we compared the same weight for both young and old people. I was. At that time, the Text uses medium weight fonts such as R and M.
"Those with visual impairments and the elderly are thought to prefer heavier weights." was found in this experiment.

(2) Recommended fonts
It assumes the environment set in the experiment.
・ Text
Suppose UD Marugo_Small Marugo_Small BEB or B is good. Accordingly, UD Kakugo_Small is also good.
・ Heading
UD Kakugo_Small BEB or B is good. Accordingly, UD Marugo_Small is also good.

(3) Discriminability
As expected, the results showed that it was difficult to distinguish characters with small differences, such as dakuten and handakuten. I think that it is necessary to try to improve the discriminability in the future.

Experiment details

■ Dividing the experiment

We conducted three experiments, including a preliminary experiment. The number of fonts used in one experiment was set to 6, and we decided to search for the appropriate font by repeating the experiment.

■ Readability experiment

■ Discrimination test

Experimental result

■ Survey ①

□ Experiment

In the preliminary questionnaire, there was an opinion that a thicker weight is better, so we decided to use "UD Kakugo_Small DBB", which is "DB for kanji and B for non-kanji".
First, we prepared "UD Kakugo_Small", which is "B for kanji weight and EB for non-kanji characters".
There was an option to use "UD Kakugo_Small E", but considering the possibility that the kanji counter would be destroyed if visually impaired people saw it, I prepared a special font.

In line with this, we conducted a survey (1) with the aim of finding appropriate weights and typefaces by mixing fonts with thinner weights for each typeface.

*Typeface: typeface
*Counter: Refers to the completely closed or partially closed part of the character margin space.

□ Readability result

There was no difference in evaluation results between the visually impaired and the elderly.

・ Text
Overall, I was able to split my rating into three.

・ Heading
Overall, I was able to split my rating into three.

□ Readability evaluation

R11) Both the Text and the Heading were highly evaluated for "UD UD Kakugo_Small BEB" and "UD Kakugo_Small DBB", and "UD Kakugo_Small BEB" was highly evaluated, so survey ② After that, I decided to verify mainly the weight of B.

R12) In the 2nd group, "UD Kakugo_Small" is equal to or better than "UD Kakugo_Large", so we decided that the size of the font area should be small.

□ Discriminability result

It was evaluated that the shorter the time it took to answer, the easier it was to discriminate.

・The visually impaired
UD Kakugo_Small BEB
UD Kakugo_Large DB
UD Kakugo_Small DB
UD Kakugo_Small DBB
UD Marugo_Small DB
Tsukushi Gothic D

·senior citizen
UD Marugo_Small DB
UD Kakugo_Large DB
Tsukushi Gothic D
UD Kakugo_Small DB
UD Kakugo_Small BEB
UD Kakugo_Small DBB

□ Discriminability evaluation

L11) Visually impaired people and elderly people were evaluated separately.
However, since this time we mainly evaluate readability, we decided to focus on the weight of B according to readability from Survey 2 onwards. For the discriminant investigations that follow, we aimed to understand the difficulty of discrimination for each pair.

■ Survey ②

□ Experiment

After receiving R11 and L11, I chose "UD UD Kakugo_Small DBB" and "UD Kakugo_Small BEB". I chose the weight B Maru Gothic, and the corresponding weight Tsukushi Gothic, Tsukushi Mincho, and Kaku Gothic Condensed.
Based on these, we conducted a survey (2) with the aim of finding an appropriate typeface.

□ Visually impaired readability result

There was a slight difference in evaluation results between the visually impaired and the elderly. Although there was a difference in ranking between amblyopia and color blindness, there was no difference in the evaluation groups, so all were combined.

・ Text

・ Heading

Figure 3: Visually Impaired Readability Results

□ Elderly readability results

・ Text

・ Heading

Figure 4: Elderly Readability Results

□ Readability evaluation

R21) Tsukushi Mincho and Condensed typefaces were rated poorly by both the visually impaired and the elderly. I believe that this is because the Mincho typeface has thin horizontal strokes, and the condensed typeface has relatively thin vertical Weight compared to other typefaces.
In addition, since Tsukushi Gothic was evaluated as standard, we decided to explore the weight mainly for Kaku Gothic and Maru Gothic from Survey 3 onwards.

■ Survey ③

□ Experiment

"Universal Design Kaku (Round) Go_Small DBB", "Universal Design Kaku (Round) Go_Small BEB", and "Universal Design Kaku (Round) Go_Small B" are fonts with weights as follows.

After receiving R21, I decided to investigate the differences between these fonts, round and square.

(1) Whether changing the weight of kanji and non-kanji is effective
(2) Which is superior, Round Gothic or Kaku Gothic?

The survey was conducted with the above (1) and (2) as the major objectives.

□ Readability result

Unlike Survey 2, the standard deviation was small because the typefaces were similar overall, making it difficult to see a large difference.
They were then divided into the following evaluation groups. "UD Maru UD Marugo_Small B", "UD Kakugo_Small DBB", and "UD Kakugo_Small BEB" were also used in Survey 2, so the results of Survey 2 (evaluation group - ranking) are shown in parentheses. describe.

□ Readability evaluation

R31) Overall, the difference was not large, but after that, we conducted a “comparison between round gothic and square gothic” and a “comparison between round gothic and square gothic” and evaluated them as follows.

Based on this result, Maru Gothic is superior to the Text for both the visually impaired (color blindness) and the elderly. In addition, Kaku Gothic has a slight advantage in Heading.

R32) Overall, DBB is in the 5th or 6th place, except for visually impaired people (color blindness) in 3rd place and elderly Text in 4th place. However, there does not seem to be a clear advantage between BEB and B.
In other words, it is considered that there is no effect for the purpose (1), "What kind of effect does changing the weight of kanji and non-kanji have?"

R33) Regarding "Heading", a similar tendency can be seen in the rankings for both the visually impaired and the elderly.