Archives of Design Research

The present study examined the government mobile applications available on Android for public services in healthcare, taxes, and utilities in Thailand. A carefully selection process was employed to choose a range of widespread applications from the offices of the Thai government that are easily accessible to the public and offer valuable resources to Thai citizens. The five applications we chose were ทางรัฐ (Thang Rath), สมุดสุขภาพ (Smud Sukhaphap), RD Smart Tax, PEA Smart Plus, and PWA Plus Life (Table 1). As of 9 April 2024, the downloads for each app on Google Play were as follows: 500K+ for Thang Rath, 5K+ for Smud Sukhaphap, 500K+ for RD Smart Tax, 5M+ for PEA Smart Plus, and 1M+ for PWA Plus Life. The study analyzed 78 words and texts, each containing specific details, as outlined in Appendices 1–5.

An analysis was conducted to acquire comprehensive insights into the typography employed in these mobile applications. This analysis led to the identification of two primary categories of Thai typefaces: Thai conventional text fonts and Roman-like Thai fonts. The results of this analysis are visually depicted in Figure 1, and a comprehensive overview of the typefaces utilized in each of the selected Thai government mobile applications can be found in Table 1. According to Table 1 and Figure 1, Droid Sans Thai (Regular) was the conventional text typeface used in both the ทางรัฐ (Thang Rath) and PWA Plus Life mobile applications for titles, subtitles, headings, and body text. However, the PWA Plus Life also utilized Mitr (Regular), which is a Roman-like Thai typeface for the Words/Texts (W/T) 01, 02, 03, 08, 14, 15, and 16 (Appendix 5). In contrast, other selected mobile applications such as สมุดสุขภาพ (Smud Sukhaphap) and PEA Smart Plus used only the Roman-like Thai typeface Prompt (Regular), while RD Smart Tax used Athiti (Regular).

Figure 1 Alphabet set of the typefaces and examples of partial typefaces used in selected Thai government mobile applications

The typeface used in the ทางรัฐ (Thang Rath) application employs a conventional Thai typeface with type sizes ranging from 1.286 to 1.983 mm and contrast ratios between 2.2:1 and 15.1:1. The study found that Words/Texts 01 (W/T 01) had a higher misreading rate (16.70%) compared to other Words/Texts at a viewing distance of 40 cm when blur simulation goggles were worn (see Figure 2). Furthermore, the result of W/T 04 under similar visual conditions showed a slight “Read-correctly-but-difficult-to-read” (10%). Although the W/T 01 and W/T 04 had type sizes of 1.926 and 1.477 mm, respectively, the contrast ratios of 4.4:1 and 2.2:1 (Appendix 1) indicated low contrast conditions (not meeting WCAG 2.1 standards), which hindered the legibility of the text. However, there were no errors in those short words that had smaller type sizes and low contrast ratios, such as W/T 07 and W/T 08. Additionally, the Chi-Square test results in Figure 2 indicate a significant difference at 0.05 only for the viewing distance of 40 cm when the goggles were worn, with no significant differences found in the other results.

Figure 2 Comparative reading performance rates of ทางรัฐ (Thang Rath) application varied by viewing distances, with and without blur simulation goggles

*W/T = Words/Texts

The typeface displayed in this application is a Roman-like Thai typeface. The range of type sizes used in the study was 1.025–1.991 mm, with contrast ratios ranging from 2.6:1 to 21:1. The results showed that when wearing the goggles, participants exhibited a significantly high rate of “Read-correctly-but-difficult-to-read” and “Misread.” This was evident in the performance of W/T 1 (type size = 1.862, contrast ratio = 11:1), W/T 22 (type size = 1.861, contrast ratio = 14.9:1), W/T 03 (type size = 1.990, contrast ratio = 20.1:1), W/T 04 (type size = 1.860, contrast ratio = 14.5:1), W/T 05 (type size = 1.859, contrast ratio = 3.7:1), and W/T 18 (type size = 1.414, contrast ratio = 21:1) (see Figure 3 and Appendix 2). Further analysis using the Chi-Square test revealed significant differences at the 0.05 level between the viewing distances of 40 cm and 35 cm, both with and without wearing the goggles (see Figure 3). However, significant differences were observed at viewing distances of 30 cm and 25 cm only when participants wore the goggles, whereas no such differences were observed when the goggles were not worn.

Figure 3 Comparative reading performance rates of สมุดสุขภาพ (Smud Sukhaphap) application varied by viewing distances, with and without blur simulation goggles

*W/T = Words/Texts

The results indicated that participants experienced a significantly high rate of reading errors and difficulties when using goggles, particularly in the case of W/T 01, where 20% of responses were “Read-correctly-but-difficult-to-read” and 23.30% were “Misread” at a viewing distance of 40 cm. Similarly, W/T 22 demonstrated a 20% “Read-correctly-but-difficult-to-read” and a 30% “Misread” at the same viewing distance. At a viewing distance of 40 cm, W/T 04, W/T 05, W/T 03, and W/T 18 exhibited a “Read-correctly-but-difficult-to-read” rate of 13.30% and a “Misread” rate between 10% and 16.70%. The readability errors and difficulties were primarily observed in W/T 01, W/T 03, and W/T 04 at different viewing distances. W/T 22 had a range of 13–16.70% for both the “Read-correctly-but-difficult-to-read” and “Misread” categories, while W/T 04 had a range of 6.70–16.70% for the same categories. The findings for W/T 01 showed that 23.30% of “Read-correctly-but-difficult-to-read” was at a viewing distance of 35 cm, 10% at a viewing distance of 30 cm, and 13.30% at a viewing distance of 25 cm. Moreover, W/T 03 exhibited 16.70% of “Read-correctly-but-difficult-to-read” and 10% of “Misread” at a viewing distance of 25 cm, while W/T 08 demonstrated a relatively high “Misread” rate of 20%, as shown in Figure 3.

The experiment findings conducted without the use of blur simulation goggles revealed that W/T 01 exhibited a 20% “Misread” rate, whereas W/T 04, W/T 05, W/T 18, and W/T 22 had “Misread” rates of 16.70%, 13.30%, 13.30%, and 13.30%, respectively (refer to Figure 3). For other viewing distances, the results indicated a relatively low percentage of “Read-correctly-but-difficult-to-read” and “Misread,” ranging from 3.30% to 10%.

The findings of the สมุดสุขภาพ (Smud Sukhaphap) application revealed that the utilization of Roman-like Thai typeface adversely impacts the accuracy and ease of reading, particularly in the case of longer texts, as observed in the results of W/T 01 and W/T 22. This is despite the display of larger type sizes (1.86 mm) and high contrast ratios (11:1, 14.9:1), as illustrated in Appendix 2. Moreover, the use of a Roman-like Thai typeface in shorter texts also affects reading proficiency, as indicated by the results of W/T 18. This is especially noticeable when the text is viewed from a distance of 40 cm and has a small type size (1.414 mm) and a high contrast ratio (21:1). Hence, it can be inferred that the implementation of Roman-like Thai typeface in small type sizes does not enhance effective reading.

The typeface displayed in this application is a Roman-like Thai typeface. The typeface had a range of type sizes from 1.348 to 1.798 mm, with contrast ratios ranging from 2.8:1 to 21:1. The findings revealed that the use of goggles resulted in a high rate of “Read-correctly-but-difficult-to-read” and “Misread,” specifically for type sizes of 1.605 mm with a contrast ratio of 4.7:1 (W/T 02), 1.348 mm with a contrast ratio of 5.7:1 (W/T 04), 1.413 mm with a contrast ratio of 2.8:1 (W/T 13), 1.604 mm with a contrast ratio of 10.7:1 (W/T 14), and 1.606 mm with a contrast ratio of 6.8:1 (W/T 11) (refer to Figure 4 and Appendix 3). Furthermore, the Chi-Square test results presented in Figure 4 indicate significant differences at the 0.05 level. All the results of the viewing distances show significant differences when wearing the goggles compared to when not wearing them.

Figure 4 Comparative reading performance rates of RD Smart Tax application varied by viewing distances, with and without blur simulation goggles

*W/T = Words/Texts

The findings revealed that at a viewing distance of 40 cm, W/T 02, which presented longer texts and had a lower contrast ratio, demonstrated significantly higher rates of difficulty and inaccuracy in reading. Specifically, the typeface showed 26.70% of “Read-correctly-but-difficult-to-read” and 23.30% of “Misread,” while at a viewing distance of 35 cm, the typeface showed 20% of “Read-correctly-but-difficult-to-read” and 13.30% of “Misread.” At a viewing distance of 30 cm, the typeface exhibited 16.70% of “Read-correctly-but-difficult-to-read” and 13.30% of “Misread.”

In contrast, the findings without goggles showed lower rates of difficulty and inaccuracy in reading for W/T 02, with “Read-correctly-but-difficult-to-read” and “Misread” rates ranging from 6.70% to 16.70% at viewing distances of 40, 35, and 30 cm. However, at a viewing distance of 25 cm, the “Misread” rate of W/T 02 increased to 23.30%. Additionally, W/T 04, which displayed the smallest type sizes in the application and had a lower contrast ratio, showed 16.70% of “Read-correctly-but-difficult-to-read” and 16.70% of “Misread” at a viewing distance of 40 cm and 20% of “Read-correctly-but-difficult-to-read” and 16.70% of “Misread” at a viewing distance of 35 cm.

Moreover, W/T 13, which presented text in a type size of 1.413 mm with a very low contrast ratio of 2.8:1, resulted in 10% of “Read-correctly-but-difficult-to-read” and 23.30% of “Misread” at a viewing distance of 40 cm. However, other typefaces with similar type size and contrast ratio properties, such as W/T 12, W/T 03, and W/T 09, had few reading problems. These typefaces presented shorter texts and were consistent with the results of previous studies. Using a Roman-like Thai typeface does not significantly affect the reading of single words or shorter texts, but it may hinder reading longer texts in small sizes.

This application employed a Roman-like Thai typeface, with type sizes ranging from 1.033 mm to 1.866 mm and contrast ratios ranging from 1.9:1 to 14.9:1. Upon wearing goggles, the results indicated a high rate of “Read-correctly-but-difficult-to-read” and “Misread,” particularly in texts W/T 02 (type size = 1.291, contrast ratio = 4.5:1), W/T 03 (type size = 1.225, contrast ratio = 4.5:1), W/T 09 (type size = 1.221, contrast ratio = 4.466:1), and W/T 14 (type size = 1.411, contrast ratio = 4.476:1) (refer to Figure 5 and Appendix 4). Further, the Chi-Square test results in Figure 5 indicate significant differences at a significant level of 0.05 in the results of the 40 cm viewing distance, both with and without wearing the goggles. Additionally, significant differences were observed in the results of the 35-cm viewing distance when the goggles were worn. However, the other findings showed no differences.

Figure 5 Comparative reading performance rates of the PEA Smart Plus application varied by viewing distances, with and without blur simulation goggles

*W/T = Words/Texts

The results of W/T 07, a well-known short word, indicated that using small font size (1.033 mm) and a very low contrast ratio (1.9:1) had a slightly negative impact on reading performance, particularly when viewed from a distance of 40 cm, with percentages of 6.70%, 3.30%, and 10% for “Read-correctly-but-difficult-to-read,” “Misread,” and “Cannot-read,” respectively. However, reading from other viewing distances did not result in any significant reading problems (see Figure 5).

Further, the study compared the results of W/T 02, W/T 03, and W/T 09, which had similar font sizes and contrast ratios. The percentage of “Misread” at a viewing distance of 40 cm was nearly identical for all three conditions. However, the percentage of “Read-correctly-but-difficult-to-read” varied across the conditions, with percentages of 13.30%, 10%, and 16.70% for W/T 02, W/T 03, and W/T 09, respectively. The reading problems significantly decreased when reading was done from different viewing distances (0–16.70%).

In contrast, the longer text W/T 14, with a font size of 1.411 and a contrast ratio of 4.476:1, showed a higher percentage of “Misread” (23.30%) when viewed at a distance of 40 cm. Additionally, it was found that 20% of participants reported that the text was “Read-correctly-but-difficult-to-read” when viewed from a distance of 35 cm. Even for W/T 07 with goggles, reading from different viewing distances still led to reading problems, such as 10% of “Read-correctly-but-difficult-to-read” and 10% of “Misread” at a viewing distance of 30 cm and 16.7% of “Read-correctly-but-difficult-to-read” and 10% of “Misread” at a viewing distance of 25 cm.

Finally, the results of the conditions without goggles showed a significant impact on reading performance at a viewing distance of 40 cm for W/T 03 (6.70% of “Read-correctly-but-difficult-to-read” and 6.70% of “Misread”), W/T 09 (10% of “Read-correctly-but-difficult-to-read” and 13.30% of “Misread”), and W/T 14 (13.30% of “Read-correctly-but-difficult-to-read”), as depicted in Figure 5. However, reading from other viewing distances did not result in any significant reading problems (refer to Figure 5).

This application employed both Roman-like Thai typefaces and a conventional Thai typeface. It featured a range of type sizes, spanning from 0.642 to 1.992 mm, and contrast ratios that varied from 2.6:1 to 21:1. The results obtained while wearing the goggles revealed a high incidence of “Read-correctly-but-difficult-to-read” and “Misread,” particularly in type sizes W/T 05 (type size = 0.642, contrast ratio = 21:1), W/T 04 (type size = 0.642, contrast ratio = 21:1), W/T 09 (type size = 1.670, contrast ratio = 14.4:1), W/T 16 (type size = 1.027, contrast ratio = 12.1:1), W/T 14 (type size = 0.897, contrast ratio = 4.7:1), and W/T 15 (type size = 1.030, contrast ratio = 5.1:1), as depicted in Figure 6 and Appendix 5. Moreover, the Chi-Square test results in Figure 6 indicate significant differences at a significance level of 0.05 in all the viewing distances, both with and without wearing the goggles.

Figure 6 Comparative reading performance rates of PWA Plus Life application varied by viewing distances, with and without blur simulation goggles

*W/T = Words/Texts

The results of the study revealed that despite being displayed in a conventional Thai typeface with a longer text and the highest contrast ratio of 21:1 (maximum), the text of W/T 05 failed to meet readers’ needs at all viewing distances, whether with or without goggles. This was primarily due to the diminutive size of the typeface used, which measured 0.642 mm. The participants experienced reading difficulties at a viewing distance of 40 cm with goggles. These difficulties included 20% of “Read-correctly-but-difficult-to-read,” 20% of “Misread,” and 53.30% of “Cannot-read.” Without goggles, the corresponding figures were 23.30% of “Read-correctly-but-difficult-to-read,” 23.30% of “Misread,” and 36.70% of “Cannot-read.” Similarly, at a viewing distance of 35 cm with goggles, participants experienced 23.30% of “Read-correctly-but-difficult-to-read,” 26.70% of “Misread,” and 46.70% of “Cannot-read.” Without goggles, the corresponding figures were 33.30% of “Read-correctly-but-difficult-to-read,” 26.70% of “Misread,” and 13.30% of “Cannot-read.” Likewise, at a viewing distance of 30 cm with goggles, participants experienced 36.70% of “Read-correctly-but-difficult-to-read,” 26.70% of “Misread,” and 20% of “Cannot-read.” Without goggles, the corresponding figures were 26.70% of “Read-correctly-but-difficult-to-read,” 36.70% of “Misread,” and 3.30% of “Cannot-read.” Finally, at a viewing distance of 25 cm with goggles, participants experienced 43.30% of “Read-correctly-but-difficult-to-read,” 33.30% of “Misread,” and 13.30% of “Cannot-read.” Without goggles, the corresponding figures were 33.30% of “Read-correctly-but-difficult-to-read,” 26.70% of “Misread,” and 3.30% of “Cannot-read.” Furthermore, the study found that the W/T 04 property, which was a short text presented in a conventional Thai typeface and had a contrast ratio similar to that of W/T 05, had a significant impact on reading difficulties, especially the problem of “Cannot-read,” at viewing distances of 40 cm and 35 cm with goggles. The corresponding figures were 26.70% and 30%, respectively (see Figure 6).

Notwithstanding the utilization of a conventional Thai typeface with a font size of 1.670 mm and a contrast ratio of 14.4:1 for the W/T 09 test (as described in Appendix 5), the outcomes of the said test with goggles revealed that 33.30% and 10% of the text were “Misread” at viewing distances of 40 cm and 35 cm, respectively (refer to Figure 6). Nevertheless, the remaining viewing distances did not exhibit any noteworthy reading difficulties.

The text in W/T 16 was presented in a Roman-like Thai typeface with longer text and a font size of 1.027 mm, as indicated in Appendix 5. This presentation resulted in significant reading errors, as revealed through experiments conducted with and without goggles, as illustrated in Figure 6. Specifically, with goggles, the reading difficulties amounted to 26.70% of “Read-correctly-but-difficult-to-read,” 16.70% of “Misread” at a viewing distance of 40 cm; 16.67% of “Read-correctly-but-difficult-to-read,” 10% of “Misread” at a viewing distance of 35 cm; 23.30% of “Read-correctly-but-difficult-to-read,” 6.70% of “Misread” at a viewing distance of 30 cm; and 6.70% of “Read-correctly-but-difficult-to-read,” 10% of “Misread” at a viewing distance of 25 cm. On the other hand, without goggles, the reading difficulties were 23.30% of “Read-correctly-but-difficult-to-read,” 13.30% of “Misread” at a viewing distance of 40 cm; 20% of “Read-correctly-but-difficult-to-read,” 10% of “Misread” at a viewing distance of 35 cm; 6.70% of “Read-correctly-but-difficult-to-read,” 13.30% of “Misread” at a viewing distance of 30 cm; and 13.30% of “Read-correctly-but-difficult-to-read,” 20% of “Misread” at a viewing distance of 25 cm.

The study results revealed that Roman-like Thai typefaces have an impact on reading accuracy and ease, particularly when it comes to longer texts. The สมุดสุขภาพ (Smud Sukhaphap) and RD Smart Tax applications employed Roman-like Thai typefaces, and both exhibited a high rate of reading errors and difficulties, specifically for longer texts. On the other hand, the ทางรัฐ (Thang Rath) application used a conventional Thai typeface, resulting in no errors for short words that had smaller font sizes and low contrast ratios.

The typeface’s impact on reading proficiency was also observed in shorter texts, such as the PEA Smart Plus and PWA Plus Life applications. In these cases, the use of a Roman-like Thai typeface led to a significantly higher rate of reading errors and difficulties. In contrast, the conventional Thai typeface used in the PWA Plus Life application resulted in significant reading difficulties, especially for longer texts with small type sizes and high contrast ratios.

According to Kamollimsakul et al. (2014a), a study on Thai fonts revealed that using a conventional Thai text font, known as a conservative font, results in faster reading and greater efficiency on web pages compared to a modern font, a Roman-like Thai font. It is noteworthy that both young and old adults showed a preference for conservative fonts over modern ones.

Mitchell (2014) observed that Thai typefaces that resemble Roman fonts have fewer features, which may improve visibility when used in small sizes. However, this may also affect legibility and precision because key features of Thai letterforms may be omitted. (Punsongserm et al., 2018a). The advantage of using Roman-like Thai fonts is their increased Bo Baimai height compared to conventional Thai fonts. This means that Roman-like Thai fonts can have more significant consonants than conventional Thai fonts at the same point size (Usakunwathana, 2015). However, the omission of loops and some features in the Roman-like Thai typeface leads to cramped letter spaces, which can be affected by the lack of loop representatives and jutting parts (Punsongserm & Suvakunta, 2022a).

A recent study by Punsongserm and Suvakunta (2022b) found that there may be better approaches than using Roman-like Thai typefaces on drug labels and documentation. The study revealed that Roman-like Thai typefaces, especially PSL Kittithada, which is highly used, were often misread over conventional text typefaces during a word accuracy identification task. The user manuals for Antigen Test Kits (ATK) that used Roman-like Thai typefaces also received low user satisfaction ratings in a user preference test. Additionally, a previous study by the same authors (Punsongserm & Suvakunta, 2022a) found that using Roman-like Thai typefaces negatively impacted participants’ reading time. However, the study also revealed that DB Ozone (a Roman-like Thai typeface) outperformed other Roman-like Thai typefaces in the word accuracy identification task and performed better than some in the reading time test. The previous study suggested that DB Ozone’s loop representatives (short horizontal lines) resulted in better letter spacing than loopless typefaces (Figure 7).

Figure 7 Comparing letterform characteristics and horizontal spacing of FT Manifest UD, PSL Kittithada, and DB Ozone in the same word (Source: Punsongserm & Suvakunta, 2022a)

Therefore, when choosing Thai typefaces for use in small sizes on mobile applications, it is crucial to consider distinct key letter features that enhance legibility. Most of these key letter features can be found in conventional Thai text typefaces. Additionally, providing examples of good typefaces can be helpful. In some cases, Roman-like Thai typefaces may be acceptable, but only if they include jutting parts and loop representatives and proper letter spacing when displayed in small type sizes. However, it is essential to note that these should only be used with large types or for headlines and subheads, as they may not be suitable for texts, particularly longer small texts.

The study identified that type size is a critical factor affecting the readability of text, particularly for longer texts. The results showed that even if the contrast ratio is high, the smaller type sizes resulted in a high rate of reading errors and difficulties, such as “Cannot-read,” “Misread,” and “Read-correctly-but-difficult-to-read.” Moreover, the results indicated that the reading difficulties associated with smaller type sizes were more pronounced when using blur simulation goggles.

The study also revealed that type size significantly affects the readability of Thai text. The findings indicated that longer texts in smaller font sizes resulted in higher reading difficulties. For instance, the PWA Plus Life application’s results demonstrated that the text of W/T 05 failed to facilitate readers’ needs at all viewing distances. This was due to the diminutive size of the typeface used, which measured 0.642 mm, even when displayed in a conventional Thai typeface with the highest contrast ratio (21:1). The study suggests that using larger font sizes can enhance the readability of Thai text in mobile applications.

According to a study by Kamollimsakul et al. (2014a), both younger and older adults tend to prefer the conservative font type (conventional Thai text font) over the modern font type (Roman-like Thai font) when browsing web pages. However, font size preferences vary depending on age groups. Younger adults tend to favor 14- and 16-point sizes over 12-point sizes, while older adults prefer 16- to 12- and 14-point sizes. The study recommends using the conservative font type for both age groups. However, it suggests choosing the appropriate font size based on font type. For the conservative font type, 12-point or larger font sizes are recommended for younger adults. In comparison, 14 points or larger are recommended for older adults. For the modern font type, 14 points or larger font sizes are suitable for younger adults. In comparison, 16 points or larger are acceptable for older adults. It is worth noting that the study involved a viewing distance of approximately 57 cm between the participants’ eyes (42 people) and the monitor. However, due to the fonts’ names not being provided in the study and their Bo Baimai height being unknown, it is impossible to determine the visual angles for each point size of the conservative and modern fonts tested in the experiment. However, assuming we have two well-known Thai fonts, Cordia New and Tahoma, we measured their Bo Baimai height at point sizes 12, 14, and 16. For Cordia New, the Bo Baimai height at these point sizes were 1.693 mm, 1.97 mm, and 2.252 mm, respectively. On the other hand, for Tahoma, the Bo Baimai heights at these point sizes were 2.311 mm, 2.696 mm, and 3.081 mm, respectively. If we view these fonts from 57 cm away, the physical sizes of Cordia New at each point size would convert to visual angles of 0.1702°, 0.1980°, and 0.2264°, respectively. In contrast, the physical sizes of Tahoma for each point size would convert to visual angles of 0.2323°, 0.2710°, and 0.3097°, respectively. Punsongserm and Suvakunta (2022b) have proposed that a minimum visual angle of 0.200° should be provided for optimal legibility. It is important to note that Tahoma’s point sizes 12, 14, and 16 exceed most readers’ minimum legibility threshold of 0.200°. However, point size 12 of Cordia New falls below the minimum legibility threshold. Nonetheless, point size 14 is almost at a visual angle of 0.200°, and point size 16 is larger than a visual angle of 0.200°.

Based on the results of studies conducted on various mobile applications, it is evident that the appropriate type size is crucial for ensuring easy readability and minimizing reading difficulties. Santayayon et al. (2011) recommended a minimum Thai-type size of 2 mm for a viewing distance of 50 cm, which corresponds to a visual angle of 0.2292°. However, as the viewing distance decreases, the recommended minimum type size should increase in order to maintain legibility. Punsongserm and Suvakunta (2022b) suggested that a range of type sizes between 1.3 and 2 mm in Bo Baimai height may be optimal for easy readability among readers from diverse backgrounds.

It is essential to consider the category of typefaces, thickness stroke, and letter spacing, among other factors, in addition to type size, when designing materials for mobile applications. Furthermore, the optimal font size may vary depending on the intended audience, the device being used, and the distance from which it is being viewed. For example, recent studies suggest that the mean viewing distance for smartphone usage is approximately 29.2 cm, resulting in a corresponding visual angle of 0.2551° (Long et al., 2017).

The present study found that contrast ratios play a significant role in determining the readability of text on mobile applications. The results revealed that low contrast ratios lead to inadequate legibility and hinder effective reading of Thai text. Even if the type size is large, low contrast ratios can lead to a high rate of reading errors and difficulties.

The WCAG 2.1 guidelines recommend a minimum contrast ratio of 4.5:1 to ensure that the text is legible for individuals with normal vision. The results of the ทางรัฐ (Thang Rath) application revealed that although the type sizes of W/T 01 and W/T 04 were relatively larger, the contrast ratios were indicative of low contrast conditions, which hindered the legibility of the text. The study suggests that high contrast ratios can significantly enhance the readability of Thai text in mobile applications. However, a study by Ojanpää and Näsänen (2003) investigated the impact of luminance and color contrast on searching for information on display devices. The research revealed that visual search times, the number of eye fixations, and mean fixation durations increased significantly when the luminance contrast decreased, even when color contrast was present. Therefore, achieving high color contrast does not necessarily ensure quick visual perception if the luminance contrast between foreground and background is insufficient. The study suggested that user interfaces require an apparent luminance (brightness) difference between foreground and background to ensure good visibility of alphanumeric information.

To improve mobile application standards or guidelines, further research is necessary to validate assumptions and provide more precise recommendations. However, based on the findings of the studies, it is recommended to use conventional Thai typefaces rather than Roman-like Thai typefaces. The results suggest that using Roman-like Thai typefaces adversely affects the accuracy and ease of reading, particularly in the case of longer texts. Even with larger font sizes and high contrast ratios, the readability and legibility of the text were hindered. Therefore, implementing conventional Thai typefaces is recommended to enhance effective reading. Also, it is recommended that a minimum type size of 1.3 mm in Bo Baimai height be used for reading body text. The font size should be increased for headlines, subheads, and text typed with Roman-like Thai typefaces. Additionally, it is important to maintain appropriate contrast ratios to ensure legibility. In addition, it is recommended that the contrast ratios for Thai mobile applications be higher than the recommendations of the WCAG guidelines. The study found that the contrast ratios of certain words/texts in the tested applications indicated low contrast conditions, which hindered the legibility of the text. Therefore, it is essential to ensure that the contrast ratios for Thai mobile applications are higher than the recommended guidelines in order to enhance reading effectiveness and accessibility for all users. Designers and developers should also consider the typeface, type size, and contrast ratio when designing applications. This is significant to ensure that users can easily read and comprehend the text accurately.