Sharing of research methods on plain text reading experience on mobile terminals

According to a survey, in 2014, the book reading rate of Chinese adults was 58.0%, and the digital reading exposure rate was 58.1%. The proportion of digital reading has surpassed traditional reading. The results also show that among emerging media, mobile reading leads the way, with an average daily mobile reading time of 33.82 minutes per person [1]. In order to provide a better article reading experience for QQ public account users, we conducted a study on the reading experience of pure text articles.
This article mainly shares the process of determining the research ideas and methods. The first part starts with the indicator system that affects mobile reading performance and introduces how to determine the evaluation indicators in the context of this study. The second part mainly introduces the research methods used in this study and future research prospects.

1. Determine the experience evaluation indicators
1.1 Mobile reading and its components

Mobile reading refers to all reading behaviors performed using mobile terminals such as mobile phones, tablets, e-readers, etc., including browsing web pages through browsers and reading through bookstore clients, news clients, information clients, magazine clients, Weibo, public account articles and other reading channels, and browsing novels, newspapers, books, magazines, animation, documents and other content.

In terms of the components of mobile reading, there are three main ones: mobile reading users (subjects), e-books (objects composed of content and mobile carriers), and behaviors (users’ attitudes and behavioral expressions), as shown in the figure below.

1.2 Indicators that affect mobile reading performance

There are many studies on the indicators that affect mobile reading performance. The relevant indicators can be divided into four categories: subject evaluation, digital content, hardware performance, and software functions (see the table below for details).

1.3 Evaluation indicators of plain text article reading experience

This study aims to provide a better article reading experience for public account users. Combined with the current status of the product, it has its own unique evaluation indicators. Among them, "digital content" is open and will be optimized through operational recommendations in the future, which is not within the scope of this study; "hardware performance" depends on the user's own mobile terminal and the impact of the mobile QQ version running, which is relatively complex. Therefore, the current research only focuses on display comfort and adapts to different screens through mainstream screen test results; "software functions" are not yet perfect, and will be differentiated in the future according to the value positioning of the article. Therefore, the current research mainly focuses on layout design, that is, the impact of factors such as fonts, font sizes, and line spacing on reading.

After determining that the purpose of this study is to optimize the reading experience of plain text articles and the corresponding secondary indicators ("*" in the table), we refined the indicators in combination with product characteristics. The display comfort of layout design can usually be evaluated from two perspectives: visual performance (i.e. usability) and visual subjective preference (i.e. aesthetics). According to previous studies, the two are significantly correlated, and those with higher subjective preference scores tend to have better recognition performance. This can also be verified from the correlation between elements that affect visual performance and design elements.

The impact of different layout designs on visual performance within a certain length of article is relatively limited (when users maintain a high level of concentration, the accuracy and time of task completion will not differ greatly), but it can be manifested through visual fatigue (which is also a manifestation of users' high concentration). In the relevant research on visual fatigue, the main conclusions are as follows:
A. Brightness contrast between text and background

When a white background meets black text, the reflectivity of the text is increased, making it easier to notice and understand, but the color difference is large, and long-term staring will cause fatigue, so the staring time is relatively short.

Users prefer positive polarity (dark text on light background), but negative polarity (light text on dark background) is actually less fatiguing.
B. Minimum acceptable viewing angle for text

It is usually determined by the font size and reading distance. The laboratory measurement method is shown in the figure below. In the e-book reading test, for Chinese characters that are close to square (i.e., the character height is equal to the character width), the generally acceptable minimum viewing distance is more than 30 cm, and the appropriate viewing distance is 50 cm. The "character height" of the youth group (20-35 years old) is at least 4.8 mm (24 pt), and the middle-aged group (36-50 years old) is at least 5.14 mm (25 pt). When the character height of the nearly square Chinese font is fixed at 4.85 mm and the line spacing is 3 mm, the character spacing of 0.61 mm or 1.21 mm will have better performance and will not increase visual fatigue [2].

In summary, when the font is determined, font color, background, font size, font spacing, and line spacing are the main factors affecting visual performance. In addition, screen brightness, screen size, screen resolution, ambient light, reading distance, user age, text length, reading purpose, etc. will affect the user's reading experience, which we will control in the study.

Visual subjective preference can be comprehensively evaluated through three indicators: clarity, aesthetics, and visual comfort.

2 Research Methods and Prospects
2.1 Research paradigm

This study is mainly divided into two parts. The first part is subjective debugging, as shown in the figure below. Users are asked to adjust the parameter values ​​related to the study on the computer. The visual effects corresponding to the parameters will be displayed on the mobile phone synchronously. Users need to adjust to their best visual experience.

(This tool was developed with the help of front-end teacher Zhang, and the parameter range is a combination of previous relevant research results and analysis results of reading-related competitive products)

The second part is a proofreading test, where users are asked to find typos in the proofreading area (about 500*2 words per article, a total of 15 typos), record the completion time and accuracy (performance measurement indicators), and conduct preference evaluation and fatigue level feedback (out of 10 points).

2.2 Variable Control

This study adopted a 2*2 within-subject design. The independent variables were reading viewing distance (Level 1: habitual viewing distance, Level 2: appropriate viewing distance - 50cm) and mobile phone models (Level 1: iPhone6, Level 2: iPhone6plus). Each test was balanced by the ABBA order effect. Subjective feedback was given after each test, and a 5-minute break was taken before the next test.

Using uniform ambient light and iPhone's adaptive screen brightness, the test subjects range in age from 14 to 35 years old. The length of each article and the distribution of typos are relatively balanced. Users are required to maintain a fixed viewing distance during the proofreading test and find as many typos as possible and write them down on paper.
2.3 Summary and Research Prospects

This study measured the visual performance of the user's subjective reading experience under several fonts selected by the designer to determine the parameter values ​​of design elements such as font color, background, font size, font spacing, and line spacing. Looking back at the results of this study, it can be found that the viewing distance of users when reading on mobile phones is closer than that of e-books. The average viewing distance is 30cm, and nearly half of the users have a viewing distance between 10-30cm. Screen size, viewing distance, and age all have different degrees of influence on parameters such as font size, line spacing, background, and font color. In addition, fonts play a very important role in this (for example, Lanting's main text is 36-38pt, and Hanyi's main text is 34-36pt). Different fonts will get different result parameters under the influence of factors such as different viewing distances. For details, please see the subsequent sharing articles by visual classmates.

Fonts serve text reading. Whether it is a designer who chooses fonts for reading design or a reader who repeatedly looks at fonts for text reading, they will gradually form their own evaluation criteria for fonts and text typesetting, consciously or unconsciously. Therefore, we usually think that there is a general visual cognitive psychology of the population and the comprehensive evaluation ability acquired by professional designers through long-term work experience. There may be differences between the two. For example, different individuals have different preferences for different fonts and different weights allocated between aesthetics and usability.

When using fonts, designers need to choose the right font according to the design requirements, set the corresponding font size, font spacing, line spacing, font color, etc., and meet the cost requirements to obtain the best visual effect. This process involves the balance between aesthetics and usability. This balance is particularly important when the font usage scenarios are complex and the users are very different.

Therefore, future research can be carried out in two directions:

1. Starting from font research, we can obtain users’ general visual cognitive psychology in different usage scenarios;

2. Starting from mobile reading performance, evaluate and optimize other aspects of the indicator system.

References

【1】Digital reading: Mobile reading leads the way in expanding content layout. Source: China News Publishing and Broadcasting Newspaper. December 28, 2015

[2] Wu Xinjie, Zheng Shunwen, Li Zhenglong, Zheng Yujing. Minimum acceptable viewing angle for different age groups when reading Chinese e-books. 13th Annual Conference and Seminar of the Human Factors and Engineering Society of the Republic of China. 2006