Not All Spacings are Created Equal: The Effect of Text Spacings in On-the-go Reading Using Optical See-Through Head-Mounted Displays

2.2.1 Text Style. Typeface, size, and format significantly affect perceptions of legibility, sharpness, and ease of reading in a computer-displayed text [15]. In the mobile OHMD reading context, previous studies [40, 56, 68] have shown that simple green text (#00FF00 in hex) against a black transparent background [54, 87]) is most visible and sans serif typefaces such as Arial are most readable [31, 40, 51]. Furthermore, a minimum size of 30pt at a distance of 2m was recommended for the Microsoft Hololens OHMD [8]. We incorporated these recommended text styles in our studies as control variables.

2.2.2 Text Layout. Various approaches have also been proposed to optimise text layouts for OHMD mobile reading. Orlosky et al. [68] suggested an intelligent system for automated positioning, such that text is placed in an on-screen position that complements the user's background environment. Similarly, Tanaka et al. [95] determined the optimal screen area for displaying information by using a mounted camera to evaluate the sight image behind the OHMD. While these methods aim to display information in ideal positions, they are responsive to the dynamically changing environment in which the user is present. This may lead to continuous changes in virtual object positions, which adversely affects the user's ability to multitask, especially in mobile situations. More closely linked to mobile OHMD reading scenarios are dual-task applications. Chua et al. [23] investigated the effect of notification positioning on monocular OHMDs and found that middle-right and top-right positions were most suitable for long-duration usage, while central positions were optimal for less urgent scenarios. Rzayev et al.’s [87] study extended these findings by demonstrating that the centre or bottom-centre positions decreased subjective workload and improved text understanding compared to the top-right position. Since mobile OHMD reading is a dual-task application that may be used for prolonged duration, we consider bottom-centre and middle-right positions to be comparable state-of-the-art strategies.

2.2.3 Text Quantity. The amount of text displayed on a single OHMD screen could affect user perception. According to Chen et al. [22], low-text quantities might include labels with a few words, while high-text quantities might consist of detailed description of objects. Previous studies on OHMD mobile reading have primarily focused on short texts (i.e., low-text quantities), such as presenting only several words [54, 87] or one sentence at a time [43]. While these strategies can help reduce the cognitive load associated with mobile reading, they limit the amount of content that can be shown.

Studies have shown that displaying small chunks of words on the screen can negatively impact users’ reading comprehension. Dillon et al. found that splitting sentences between pages often results in a frequent return to the previous page to reread the text [32]. This splitting will likely disrupt the comprehension process by placing an extra burden on the limited capacity of working memory. Additionally, 10-20% of the eye movements made when reading in this condition are regressions to earlier fixated words. Previous research [62, 85] also indicated that larger text sizes are more readable than smaller ones.

In the mobile learning context, Ram et al. [77] suggested that users can comprehend 6-8 chunks of the information displayed on-screen for controlling the information density. They further recommended that the information be persisted on the same screen to ease the temporal load on the working memory. Previous research has also highlighted the importance of data persistence for enhancing the understanding of information presented on OHMDs [66, 76], which suggests the potential value of displaying longer texts on these devices. However, Fukushima et al. [38] investigated presenting 10-line text using default spacing settings on an OHMD while walking on a treadmill and found that the text blocks displayed were challenging to read while walking and felt “overwhelming”. Therefore, we investigate how to adjust text spacing so that users can access longer text (e.g., emails or articles) intuitively and efficiently for OHMD mobile reading.

2.3.1 Inter-word spacing. Inter-word spacing refers to the space between words. It is an important factor in the way words are identified, as it helps the reader delineate between the beginning and end of words [75]. Inter-word spacings also guide eye movements and direct the eye towards target reading positions [81]. Previous studies affirm that inter-word spacings are essential, demonstrating a slower reading speed when removed [37, 74, 80, 81, 83].

2.3.2 Inter-line spacing. Similarly, inter-line spacing, which is defined as the space between vertically adjacent lines, assists in guiding eye movement [75, 81]. Previous studies found a slight increase in reading speed in peripheral vision for normally sighted individuals, but no improvement was found for reading using central vision in desktop reading scenarios [18, 21, 24].

Overall, the default inter-word and inter-line spacing are more optimal for conventional reading scenarios. Yet mobile reading on OHMD is a unique use case with vastly different characteristics. The default spacing arrangement may not be optimal for this new context. It remains inconclusive whether the same insights translate to real-world OHMD reading and physical navigation scenarios. This motivated us to delineate factors influencing mobile OHMD reading performance in a series of full lab studies.

4.4.1 Primary reading task performance. This was measured using the reading goodput and comprehension. Similar to Ku et al. [54], we calculated reading goodput by dividing the total number of words correctly read out by the total time spent in each text spacing condition. Its units are words-per-minute (WPM). For comprehension, we measured the percentage of questions answered correctly [72, 86].

4.4.2 Secondary navigation tasks. This was measured using the Percentage of Preferred Walking Speed (PPWS) and reading task accuracy [43, 77]. For PPWS, we measured the walking speed of each text spacing condition and calculated the percentage based on the normal walking speed (measured at the beginning of the study procedure). For accuracy, we measured the percentage of locations correctly read out.

4.4.3 The task-switching performance. This was evaluated via:

1. Switch-back Duration (SBD). This measures the average switching time between primary (OHMD reading) and secondary (navigation) tasks [70]. Specifically, we started measuring the duration at the completion of a secondary task and ended at the beginning of a primary task.
   
2. Switch-back Error Rate (SBER). We measured the percentage of words repeated, skipped, or incorrectly read aloud while the participant resumed the primary task from the secondary task [13, 91].
   

4.4.4 Subjective measures.

1. NASA-TLX. This procedure computes participants’ perceived task load [44].
   
2. Overall preference ranking across all conditions.
   

5.3.1 Primary Task Performance. This was evaluated by Reading Goodput (number of correct words read per minute) and Reading Comprehension Scores. In summary, uniformly increasing inter-word and inter-line spacing improves the mobile OHMD reading performance, but its benefits diminish as the spacing increases to 150 pixels, suggesting that while increasing text space can potentially be beneficial, there is an optimal range. Text spacings that go below or above this range may adversely affect performance. A detailed analysis is provided below.

Comparing the Reading Goodput (wpm), we found a main effect of Text Spacing (F_{3, 33} = 6.86, p < .05, $\eta ^{2}_{G}=0.06$) on Reading Goodput. Pairwise comparisons revealed that 50 pixels (M = 119.12 wpm) had a higher Reading Goodput than both 0 pixels (M = 103.46 wpm, p < .05) and 150 pixels (M = 104.26.26 wpm, p < .05) (see Fig. 4). 100 pixels (M = 116.56 wpm) also had a higher Reading Goodput than 0 pixels (M = 103.46 wpm, p < .05). However, no significant difference between 50 pixels (M = 119.12 wpm) and 100 pixels (M = 116.56 wpm) was found. This shows that text spacings are more suitable for mobile OHMD reading than the default spacing (0 pixels). However, text spacings should not exceed 100 pixels because an increase from 50 to 100 pixels resulted in only a 2.1% decline in reading performance, but increasing it from 50 to 150 pixels led to a 14.3% performance drop. In addition, there was no significant effect of Text Spacing on Reading Comprehension Scores, suggesting that although text spacings can affect Reading Goodput, it does not significantly affect comprehension.

5.3.2 Secondary Task Performance. This was measured using reading Task Accuracy and Percentage of Preferred Walking Speed (PPWS). In summary, we found a positive correlation between the amount of text spacing and navigation performance. In other words, the larger the text spacings, the faster and easier the navigation. Details are provided below.

There was no main effect of Text Spacing (p > .05) on Task Accuracy, indicating that participants were similarly accurate with spaced texts as well as without. Also, we found a main effect of Text Spacing (F_{3, 33} = 29.59, p < .001, $\eta ^{2}_{G}=0.12$) on PPWS (see Fig. 4). Pairwise comparisons revealed that text spaced with 150 pixels ($M=48.5\%$), 100 pixels ($M=46.9\%$), and 50 pixels ($M=45\%$) all resulted in a higher PPWS than 0 pixels ($M=40.7\%$) (all p < .001). In addition, 150 pixels ($M=48.5\%$) also incurred a higher PPWS than 50 pixels ($M=45\%$) (p < .05). The relationship between text spacing and walking performance is positively correlated (i.e. improvements in walking performance taper as text spacings increase), though not linearly proportional (i.e. increasing from 0 to 50 pixels led to a 10.6% improvement, while increasing from 100 to 150 pixels only yielded a 3.4% improvement). These findings are in line with the Resource competition framework (RCF) [70]; greater spaces between words provide visual gaps for readers to look at their physical environment. Thus participants were better able to navigate the environment, which explains the faster walking pace.

5.3.3 Overall Task-Switching Performance. This was evaluated using Switch-Back Duration (SBD) and Switch-Back Error Rate (SBER: Percentage of words repeated, skipped, or incorrectly read). In summary, we found that increasing text space improves task switching, but its improvement rate roughly followed a power curve (see Fig. 4) instead of being linear. In other words, the improvement is much greater with initial space increments but slowly approaches maximum benefit as the increment increases. Note that with a limited range, it is difficult to judge whether further increments of text spacing will approach a theoretical limit or reverse the trend.

The results of SBD and SBER exhibit similar trends. We found a significant main effect of Text Spacing on both SBD (F_{3, 33} = 14.43, p < .001, $\eta ^{2}_{G}=0.09$) and SBER (F_{3, 33} = 10.34, p < .05, $\eta ^{2}_{G}=0.30$). Pairwise comparisons revealed that text spaced with 50 pixels (M_SBD = 1.25s, $M_{SBER}=4.3\%$), 100 pixels (M_SBD = 1.03s, $M_{SBER}=2\%$) and 150 pixels (M_SBD = 0.97s, $M_{SBER}=2.6\%$) incurred less SBD and SBER than 0 pixels (M_SBD = 1.57s, $M_{SBER}=9.3\%$) (all p < .05). These results suggest that adding text spacings improves OHMD multitasking. Text with 100 or 150 pixels could save participants around 38.2% of task switch-back time and improve task switch-back accuracy by 76.3-78.5% as compared with default spacing (0 pixels).

5.3.4 Subjective measures: The NASA TLX results formed a pattern similar to task-switching performance. Its improvement rate roughly followed a power curve (Fig. 4) with much greater gains during initial space increment and slowly approaches maximum benefit as the increment further increases. A detailed analysis is provided below.

We observed a main effect of Text Spacing (F_{3, 33} = 9.99, p < .001, $\eta ^{2}_{G}=0.07$) on perceived Task Workload. Pairwise comparisons revealed that 50 pixels (M = 43.45), 100 pixels (M = 40.92) and 150 pixels (M = 39.28) had significantly lower task workloads than 0 pixels (M = 53.14) (all p < .05). This suggests that greater text spacings can reduce task workload (i.e. 50, 100, and 150 pixels outperformed 0 pixels). These findings also show similar trends to PPWS, SBD and SBER results.

In addition, participants were asked to provide feedback on two aspects, i.e. the most preferred and the maximum acceptable condition (upper bound of text spacings). All participants preferred having text spacings over default spacing. Of all spacing configurations, seven participants preferred 100 pixels as their first choice, while the remaining five preferred 50 pixels. Eight participants felt 150 pixels were their upper bound, while four chose 100 pixels.

5.4.1 Increased text spacings outperform the default spacing arrangement, and improve OHMD multitasking performance. We found that greater text spacings (100 and 150 pixels) considerably outperformed default spacing in terms of reducing participants’ task workload, resulting in a significantly higher PPWS, and lower SBD and SBER measures of multitasking. This suggests that text spacing could positively influence the way in which participants shifted their attention between the digital display and the physical environment. Participants’ feedback showed that with increased text spacings, they could more easily view digital information while simultaneously locating their physical environment through the gaps between words. Besides, these spaces seemed to help participants detect task-switching boundaries, which caused less disruption and reduced the overall multitasking workload.

As a result, all participants preferred having some text spacing over default spacing. Participants found that default spacing made it “difficult to see the surrounding environment through the text” (P8, P9). Also, spacings between words “facilitate[d] smooth reading and switching between the primary and secondary tasks" (P1, P12).

5.4.2 There is an acceptable range for text spacings. Although different measures provided different conclusions, considered together, we conclude that the benefit of spacing increment lies within a certain range. Beyond its range, the overall benefits either plateaued or diminished. For example, as the spacing increased beyond 100 pixels, the reading performance decreased. While deciding whether to pick 100 or 150 as the suitable range, we looked at the data as well as consulted our participants. Based on the subjective feedback, eight participants found 150 pixels to be “too spacious” (P2, P9), and as it supported much fewer words per screen, content had to be spread “across multiple pages/screens” (P8). The downside of this is that participants had to click and scroll more frequently, which “[added to] more mental and physical demand” (P1). Taking into account objective results and subjective feedback, we chose 100 pixels as the upper limit for our follow-up studies.

Three types of text display format were evaluated in the Pilot study: (i) W0; (ii) W100 Unaligned; (iii) W100 Aligned with 5 columns per screen.

6.3.1 Primary reading performance: For inter-word spacing, we only found a main effect on Reading Goodput (F_{2, 34} = 17.27, p < .001, $\eta ^{2}_{G}=0.06$) (see Fig 7). Pairwise comparisons revealed that W0 (M = 113.08 wpm) resulted in the highest Goodput (all p < .001), followed by W50 (M = 106.9 wpm) and finally W100 (M = 99.26 wpm). Also, W50 was significantly higher than W100 (p < .05). This suggests that inter-word spacing reduces reading speed, which is in line with previous studies findings. For inter-line spacing, there was a main effect on Reading Goodput (F_{2, 34} = 49.97, p < .001, $\eta ^{2}_{G}=0.15$). Pairwise comparisons revealed that L0 (M = 93.09 wpm) resulted in the lowest Reading Goodput (all p < .001), but L50 (M = 111.1 wpm) and L100 (M = 115.05 wpm) were comparable (p > .05). These results suggest that while additional inter-word spacing reduces reading goodput, additional inter-line spacing increases it, but only for the first 50 pixels.

6.3.2 Secondary Task Performance: There was no main effect of inter-word spacing on Task Accuracy and PPWS, while we found a significant main effect of inter-line spacing on PPWS (F_{2, 34} = 71.16, p < .001, $\eta ^{2}_{G}=0.05$) but no effect on Task Accuracy (see Fig 7). Pairwise comparisons show that L100 ($M=43.33\%$) and L50 ($M=42.43\%$) had higher PPWS than L0 ($M=37.87\%$) (all p < .001, see Fig 7). This suggests that greater inter-line spacing could generate higher walking speeds.

6.3.3 Overall Task Switching Performance: There was no main effect of inter-word spacing on SBD and SBER. We found significant main effects of inter-line spacing on SBD (F_{2, 34} = 60.94, p < .001, $\eta ^{2}_{G}=0.27$) and SBER (F_{2, 34} = 8.54, p < .001, $\eta ^{2}_{G}=0.12$). Pairwise comparisons showed similar trends between both measures, which revealed that L100 (M_SBD = 0.81s, $M_{SBER}=1\%$) and L50 (M_SBD = 1s, $M_{SBER}=2.5\%$) had lower SBD and SBER measures of multitasking than L0 (M_SBD = 1.55s, $M_{SBER}=5.13\%$) (all p < .001) (see Fig 7). In addition, L100 (M_SBD = 0.81s, $M_{SBER}=1\%$) had a lower SBER than L50 (M_SBD = 1s, $M_{SBER}=2.5\%$) (p < .001), which suggests that greater inter-line spacing can generate task switching speeds and accuracy in multitasking scenarios.

6.3.4 Subjective measures: The NASA TLX assessment revealed one main effect of inter-line spacing (F_{2, 34} = 26.39, p < .001, $\eta ^{2}_{G}=0.10$) (see Fig 7). Pairwise comparisons revealed that L100 (M = 34.83) and L50 (M = 35.75) both had a lower task workload than L0 (M = 45.9) (p < .001, each). This suggests that increasing inter-line spacing may help participants lower task workloads in multitasking scenarios. Participants’ preferences also showed a similar trend. Eleven participants preferred having only inter-line spacing (without inter-word spacing) between words: nine preferred W0_L100 while two preferred W0_L50. Yet, only six participants preferred combining both spacings: Five preferred W50_L100 (i.e. inter-line spacing larger than inter-word spacing) and one preferred W100_L100. Only one participant preferred W100_L0 (inter-word spacing without inter-line spacing) because excessive inter-line spacing made it difficult for her to move from one line to the next as she read.

6.4.1 Inter-line spacings increase OHMD reading and walking speed, as well as overall multitasking performance. Our results show that text with inter-line spacings (of 50 and 100 pixels) considerably outperformed those without additional inter-line spacings (L0) in terms of task workload, reading speed, walking speed, as well as multitasking speed and accuracy. These benefits were also evident from participant comments. With increased inter-line spacing, they could “easily track where [they] stopped when switching back to the primary reading task” (P9 and P4). With small or no inter-line spacing, they “felt confused about where to locate the next line that should read” (P11 and P10). Overall, we find support for hypotheses H1.1 (inter-line spacing improves reading performance) and H1.2 (inter-line spacing increases walking speed).

6.4.2 Inter-word spacings reduce reading speed while walking. While almost all objective measures show no effect for inter-word spacing, participants produced the highest goodput with W0 conditions. In fact, adding 50 pixels and 100 pixels of inter-word space reduces goodput by 5.5% and 12.22% respectively. This undesirable effect is consistent with our qualitative results, where almost all participants (17/18) disliked W100_L0. As hypothesised, participants felt that the additional 50 pixels result in the eyes having “to move rapidly and [is] thus very distracting” [P9]. Distracted participants could, as a result, accidentally read from the wrong line. Therefore, we find no support for H2.1, which hypothesises that there is an optimal inter-word spacing. We consider the possibility that spacings between W0 and W50 (e.g. W10, W20, W30, W40) may influence outcomes and suggest this as a direction for future exploration. In addition, our results support H2.2, which hypothesises that inter-word spacing does not affect walking speed.

6.4.3 Inter-line spacing is more beneficial than inter-word spacing in OHMD multitasking scenarios. Inter-word spacing is more important than inter-line spacing in stationary reading contexts [18, 42], and our previous studies also hint towards the fact that this translates to the mobile OHMD reading situation. Despite this, our results do not support H3.1, revealing no interaction effect between inter-word and inter-line spacings on all measurements. In fact, our results reveal that inter-line spacings have significant positive effects on reading goodput, PPWS, task switching performance and task workload. On the other hand, inter-word spacings negatively affect reading goodput and have no effect on all other measurements. Hence, we find support for H3.2, that inter-line spacings are more advantageous than inter-word spacings.

From our studies so far, we have gained two main insights. Firstly, the mechanism by which inter-word spacings affect reading speed differs from that of inter-line spacings; while inter-word spacing influences the reading of consecutive words, inter-line spacing affects the reading of successive lines. Since the frequency of word-to-word reading is higher than that of line-to-line reading, “undesirable inter-word spacing could [more] easily have a negative effect on reading speed” (P11). Participants also found it challenging to locate words when switching between secondary and primary tasks (P7 and P15). Despite the negative effects of inter-word spacing on reading goodput, Study 1 showed that equal inter-word and inter-line spacings improved reading performance. This suggests that inter-line spacings can function to compensate for the negative impacts of increased inter-word spacings. Secondly, results demonstrated that inter-line spacings positively affect task-switching performance (between primary to secondary tasks). Taken together, we recommend W0_L100 as the optimal spacing, as it facilitates reading and walking performance. We formally evaluate W0_L100 in the following study.

7.3.1 Primary reading performance: There was a significant effect of the Display Method (F_{2, 68} = 16.76, p < .01, $\eta ^{2}_{G}=0.04$) on Reading Goodput (see Fig. 9). Pairwise comparisons revealed that W0_L100 (M = 105.4 wpm) had a significantly higher Reading Goodput than both the bottom centre (M = 100.32 wpm, p < .05) and middle right (M = 94.19 wpm, p < .001). Also, there was a significant difference between the bottom centre and the middle right (p < .05). This suggests that participants read faster and more accurately with our text spacing interface than with positioning strategies. There was no significant effect of the Display Method (p > .05) on Reading Comprehension. This suggests that participants could comprehend the passages well with text spacing as well as the positioning strategies.

7.3.2 Secondary Task performance: There was no significant effect of the Display Method (p > 0.05) on Task Accuracy (see Fig. 9), suggesting that participants were similarly accurate when completing the secondary tasks with text spacing and positioning strategies. There was a significant effect of Display Method (F_{2, 68} = 3.97, p < .05, $\eta ^{2}_{G}=0.003$) on PPWS. Pairwise comparisons revealed that W0_L100 ($M=39.1\%$) had a significantly higher PPWS than the middle right ($M=37.69\%$, p < .001). But no significant difference was found between W0_L100 and bottom centre ($M=38.67\%$). This suggests that participants walked faster with the text spacing interface than the middle right position.

7.3.3 Overall Task switching performance: There was a significant effect of Display Method on both Switch-Back Duration (F_{2, 68} = 12.97, p < .001, $\eta ^{2}_{G}=0.06$) and Switch-Back Error Rate (F_{2, 68} = 3.87, p < .05, $\eta ^{2}_{G}=0.05$) (see Fig. 9). Pairwise comparisons revealed that W0_L100 (M_SBD = 1.15s, $M_{SBER}=3.71\%$) had a significantly lower switch-back duration and reading error rate than the middle right (M_SBD = 1.53s, $M_{SBER}=7.52\%$) (all p < .001) and bottom centre (M_SBD = 1.52s, $M_{SBER}=7.66\%$) (all p = < .001) positions. However, no significant difference between the bottom centre and the middle right was found. This suggests that participants could switch back quickly with the text spacing interface to their primary task, and they could read more accurately with our text spacing interface upon resuming their primary task.

7.3.4 Subjective measures: There was a significant effect of Display Method (F_{2, 68} = 6.90, p < .05, $\eta ^{2}_{G}=0.05$) on perceived task workload (see Fig. 9). Pairwise comparisons revealed that W0_L100 (M = 50.8) had a significantly lower subject workload than the middle right (M = 57.13, p < .05). But no significant difference was found between W0_L100 and the bottom centre (M = 53.43), as well as the bottom centre and middle right. While Rzayev et al.’s study suggested that text read from the OHMD's bottom centre position can impose a lower subjective workload compared with other positions (i.e., middle right), this may be due to differences in task complexity; while Rzayev et al. utilised a simple walking scenario, our navigation task simulated real-world complex situations that naturally resulted in a higher workload.

For the simple walking path, eight participants preferred our text spacing interface over the two text positioning methods. Six participants preferred the bottom centre, while the remaining four preferred the middle right position. For the complex walking path, twelve participants preferred our text spacing interface over the two text positioning methods. Three participants preferred the bottom centre, while the remaining three preferred the middle right position. Between the bottom centre and middle right, eight participants preferred the former. This suggests that our text spacing strategy becomes more advantageous as the difficulty of the navigation path increases.

7.4.1 Is the optimal text spacing interface identified from Study 2 better than the middle right and bottom centre positioning strategies for mobile OHMD reading?. In this study, a mixed design method was used, in which the same experiment was conducted on two groups with different walking path difficulties. The results obtained from both groups demonstrated a consistent pattern, providing strong evidence for the generalizability of the findings to a broader population and to varying levels of walking path difficulty. In previous studies [23, 87], positional approaches were not tested in scenarios displaying large amounts of text. While we do not claim to surpass these approaches, we were inspired by them in developing our own methods for displaying large chunks of text. We have identified three strategies for displaying such text, each of which maintains a similar total number of words to leave the same space for users to maintain their awareness of their surroundings. Although we kept the same remaining spaces on OHMD across the three Display Methods, the results revealed that our text spacing interface yields notably higher reading and walking performance over the two recommended position methods in both the simpler and the more complex walking paths. Participants’ preferences supported these because increasing inter-line space (W0_L100) “allow[ed] greater readability” (P5, P21). The spacing between lines also facilitated greater temporal awareness (P10, P25). P6 and P27 felt it was “easy [to continue] after switching between tasks”. Compared to the two positioning methods, participants felt that the text spacing strategy made it easy for them to locate where they had last stopped, requiring them to only “read a couple of words instead of glancing through sentences" (P7, P23). Additionally, participants’ preferences also suggested that our text spacing strategy becomes more beneficial as the navigation path's difficulty increases. Overall, results demonstrated that the spacing strategy identified in Study 2 could bring more benefits to the activity of mobile OHMD reading, perhaps more so than existing position-based layout strategies.

The reader's eye traces a Zig-Zag route when they read. As they scan from the top left to right, they form an imaginary horizontal line. As they continue from the top right to the next line, they create an imaginary diagonal line. This repeats until they arrive at the last word on the screen. (a) A text message with 38 words arranged in 3 lines within 1024x768 screen size. We assume the text width is 885 pixels and the default spacing is 4 pixels between lines with the font height to be 30 pixels. Increasing its (b) inter-word spacing and (c) inter-line spacing by 100 pixels increases the eye-movement cost by 150% and 0.4% respectively.

The inter-line spacing between the labels needs to be adjusted for a better viewing experience on OHMDs.