Keywords

1 Introduction

The effect of character display on visual search performance has been widely concerned by many researchers at home and abroad. The study found that the display ergonomics such as character font, size, character density, line spacing and alignment affect visual search performance. Ling (2006) compares the visual search performance of the Times New Roman font and the Arial font design in a web page. For Chinese fonts, Gongkun (2009) used visual search procedure and experimental introspection methods to examine the influence of Chinese character fonts on the visual search response. The experimental results show that when the font size is the same, the visual search of Song-style is very significantly less than that of the Regular Script. Shen (1990) also conducted systematic research on the factors that affect the ergonomics of Chinese characters display early. Text information density is one of the important factors affecting the visual search performance of the visual display interface. For example, Ling et al. (Ling and Schaik 2007) studied the effects of different line spacings and text alignment methods on visual search performance for web-based web interfaces. The results show that wider line spacing and left-aligned text can significantly improve search performance. In terms of the density of Chinese textual information, Jing et al. (2010) studied experimentally the relationship between web page character information density and recognition response time. Their experimental results show that the information display density has a significant effect on search efficiency, and there is an approximate inverted U-shaped curve relationship between the two. Environment may also affect visual search. Yuan et al. (2011) and Underwood et al. (2002) studied the impact of vehicle speed and text height on visual search patterns in a simulated traffic environment.

With the wide application of touch screen display, more and more physical operation interfaces are replaced by touch screen interfaces, which also makes the working environment of the touch screen interface more complicated. For example, the touch screen operation in a vibrating environment has a high requirement for the operator’s visual interpretation performance to avoid operational errors. Wu et al. (2015) analyzed the vibration environment of armored vehicle cabins, explained the impact of vibration on the ergonomics of occupant operations, and gave the effect of vibration on visual recognition.

However, so far, no study has examined the effect of text size on visual search performance in a vibration environment. Therefore, this study will mainly examine the visual interpretation performance of different character sizes in a vibration environment to analyze the effect of character size on interpretation performance.

2 Method

2.1 Participants

A total of 20 college students participated in the experiment, and all of them are males with age between 22 to 26 years old with normal visual ability and all with right-handedness.

2.2 Equipment

Simulated Vibration.

The factors that cause vehicle vibration during vehicle driving include road surface excitation, traction load, vehicle motion device, power device, etc. The road surface irregularity is the main cause of vehicle vibration. The vibration of the vehicle is mainly three-axis vibration. The positive X1 axis is defined from the left to the right, the Y1 axis is positive from the back to the chest, and the Z1 axis is directed from the foot to the head. Vehicle vibration is mostly low-frequency vibration, and the intensity of vibration in the Z1 direction is generally larger than that of the X1 and Y1 directions, which is the main factor affecting the ergonomics of the crew. The research needs to establish an environment simulation model to simulate different vibration environments, mainly including the vibration amplitude and vibration frequency, and the vibration amplitude and frequency are required to be adjustable. A self-oscillating platform was developed, as shown in Fig. 1. An eccentric was driven by a motor to make the platform vibrate. The initial state of the table and chairs is that the upper plate is substantially parallel to the ground, and the last state is that the upper plate vibrates to the maximum amplitude position. Taking into account the common installation position and ease of use, the desktop on the shake table is adjusted to a tilt angle of about 70°.

Fig. 1.
figure 1

Simulated vibration

Full size image

Because the vibration simulation model drives the vibration of the upper plate through the eccentric wheel, the amplitude of the vibration needs to be changed by changing the eccentric diameter and eccentricity. Two types of eccentric wheels with different diameters are designed for this project, there are two amplitudes. The vibration frequency was 2.5 Hz and 5 Hz.

Vibration Measuring Equipment.

According to the characteristics of the measuring equipment, vibration acceleration was used to characterize the vibration intensity. The vibration acceleration was measured using the MEMS triaxial accelerometer MPU6050 in the Mini IMU AHRS.

Character Search Device.

Experiments are performed on Lenovo Yoga3 computer with 14 inch touch screen, 1920 × 1080 screen resolution, 16:9 aspect ratio in simulated vibration environment and running real vehicle environment.

Visual Search Test Software.

The main purpose of the visual search test software development is to obtain the impact data of font size and spacing on visual search task performance in simulated vibration environment and running real vehicle environment. The visual search test software support text display to visual interpretation performance research.

2.3 The Experimental Design

Simulated Vibration.

This experiment is a 2*2 *4 within-subjects design. The independent variables include frequency, amplitude, and text size. The amplitude includes two levels of 2.5 Hz and 5 Hz. The amplitude includes two sizes. The text size includes four levels of 20 pt, 30 pt, 40 pt, and 50 pt. The dependent variable includes performance indicators and subjective evaluation indicator, and the performance indicators include the accuracy and the response time of search. A total of 20 college students have participated in the experiments.

Running Real Vehicle Vibration.

Beyond the simulation vibration station, experiments have also been carried out on one certain type of real vehicle running on the training field, and the vibration intensity of the real vehicle is weaker than the simulated vibration intensity. This experiment is a within-subjects design. The independent variable of the experiments is character size, including four levels: 20 pt, 30 pt, 40 pt and 50 pt. The real vehicle running at normal speed. A total of 8 college students have participated in the experiments.

2.4 Task

The task of the experiment is to ask the subject to visually search the screen for the number of target words in a line of text under the vibration state, and record the result by touch click.

2.5 Procedure

The subject sat on the table on simulated vibration or running real vehicle vibration. After the main experimenter introduces the instructions to the participants, the practice procedure is first started. When the participants confirm to understand the experiment tasks completely, the experiment will begin. In the formal experiment, the distance between the eyes of the participants and the screen is 40 cm. Firstly, the central fixation point “ + ” will be presented in the central of the screen. After 2 s, the target word which will be researched by the participants will appear, with a size of 20 pt and the font of Song Typeface, and it will disappear after 2 s. At this point, one line will scroll up from the bottom of the screen. The text has a total words of 19 characters, including the number of the target words from 3 characters to 5 characters randomly, and the positions of the target words in the text are also random. Participants need to find out the number of target words contained in the line as soon as possible in the case to ensure the correct situation, then enter the number of the target words by touching and clicking in the keystrokes area. The experimental program is shown in Fig. 1. Each participant needs to complete a total of 16 tests under the combination of three character sizes, two frequencies, and two amplitudes. Each group includes 12 judgment tasks. In order to balance the influence of the sequence of each group, the Latin square balance was used in the experimental sequence. At the same time, in order to control the effect of the text length of the line under different character sizes on the visual search difficulty, the program automatically adjusts the character spacing to keep the total length of the line text unchanged.

After the completion of each test, the participants need to make 5 score on the difficulty of the search of the size of this group characters under vibration.

The experimental procedure is complied with the C # language, and the program will automatically record the right and wrong of the participants and their reaction (Fig. 2).

Fig. 2.
figure 2

Schematic diagram for the task of the experiment

Full size image

3 Results

The experimental results are processed by SPSS19.0 soft package. Firstly, the data of extreme value outside the three standard deviations in all conditions are removed, which account for about 1.1% of the total data.

3.1 Visual Search Performance in Simulated Vibration Environment

Visual Search Accuracy under Different Frequency, Amplitude and Character Size Combinations.

Figure 3 is the search accuracy under different character sizes, vibration frequencies and amplitudes. The repeated measure with analysis of variance show that the main effect of characters is not significant (p > 0.05); The main effect of the vibration frequency is significant (p < 0.01); The main effect of the vibration amplitude is significant (p < 0.01); And the interaction effect of vibration frequency and amplitude is significant (p < 0.05). Afterwards, the tests show that the difference between 30 pt and 50 pt is significant (p < 0.05), the difference between 40 pt and 50 pt is significant (p < 0.05), and the difference between 30 pt and 40 pt is not significant (p > 0.05). This shows that character size has influence on visual search accuracy under vibration state, But the main effect is not significant more relative to vibration.

Fig. 3.
figure 3

Visual search accuracy of different character sizes under simulated vibration

Full size image

Visual Search Reaction Time for Different Frequency, Amplitude and Character size Combinations.

Figure 4 shows the results of search reaction times under different vibration and character sizes. The unit of the time is millisecond. The repeated measure with analysis of variance show that the main effect of the size of the characters is not significant (p > 0.05). The main effect of the vibration frequency is significant (p < 0.01). The main effect of the vibration amplitude is significant (p < 0.01). The interaction effect of vibration frequency and amplitude is significant (p < 0.05). Afterwards, the tests show that the difference between 20 pt and 50 pt has reached the significant level (p < 0.05), and the differences between other sizes are not significant (p > 0.05). From the Fig. 4, we can see that with the character size becoming larger, the reaction time is gradually decreasing, and the effect can be more obviously under higher frequency and larger amplitude conditions.

Fig. 4.
figure 4

Visual search reaction time of different character sizes under simulated vibration

Full size image

The Subjective Ratings of the Visual Search for Different Frequency, Amplitude and Character Size Combinations.

Figure 5 shows the results of ratings under different vibration and character sizes. The repeated measure with analysis of variance results of subjective ratings of the participants under different character sizes and vibrations show that the main effect of vibration frequency is not significant (p > 0.05). The main effect of vibration amplitude is not significant (p > 0.05), But the main effect of character size is significant (p < 0.01). There is no interaction between vibration frequency, amplitude and character size. The further analysis shows that the character size of 20 pt and the differences between 30 pt, 40 pt and 50 pt have reached the significant level (p < 0.01). It means that the smaller size of (20 pt) the characters is inappropriate.

Fig. 5.
figure 5

Subject ratings of different character sizes under simulated vibration

Full size image

3.2 Visual Search Performance in Running Real Vehicle Vibration Environment

Visual Search Accuracy under Different Frequency, Amplitude and Character Size Combinations.

Figure 6 is the search accuracy under different character sizes under real vehicle running vibration. The analysis of variance show that the main effect of character size is significant (p < 0.05). Afterwards, the tests show that the character size of 40pt and the differences between 20 pt and 30 pt have reached the significant level (p < 0.05).

Fig. 6.
figure 6

Visual search accuracy of different character sizes under vehicle vibration

Full size image

Visual Search Reaction Time for Different Frequency, Amplitude and Character Size Combinations.

Figure 7 is the search reaction time under different character sizes under real vehicle running vibration. The analysis of variance show that the main effect of character size is not significant (p > 0.05). Afterwards, the tests show that there is no significant difference between various sizes.

Fig. 7.
figure 7

Reaction time of visual search of different character sizes under vehicle vibration

Full size image

The Subjective Ratings of the Visual Search for Different Frequency, Amplitude and Character Size Combinations.

Figure 8 shows the results of ratings under different character sizes. The analysis of variance results of subjective ratings of the participants under different character sizes show that the main effect of character size is significant (p < 0.01). The further analysis shows that the character size of 20 pt and the differences between 30 pt, 40 pt and 50 pt have reached the significant level (p < 0.01).

Fig. 8.
figure 8

Subject ratings of different character sizes under vehicle vibration

Full size image

Under the vibration of running vehicle, the smaller size of (20 pt) the characters is inappropriate and even will improve subjective difficulties. But the measured performance show that the performance of character size 40 pt and 50 pt is worse than the performance of size 20 pt and 30 pt. This may because that the smaller size of character will inspire participants’ motivation to become more concentrated, which has promote the performance.

4 Discussion

In this study, the influence of different character sizes on visual search under different frequency and amplitude vibration environments was investigated through the accuracy, the reaction time and subjective scores in visual search. The results show that under the vibrating environment, as the size of characters increases, the reaction time will decrease. In the case of 2.5 Hz, the decreasing trend of the reaction time at both small amplitude and large amplitude is relatively gentle; at 5 Hz, the decreasing trend of the reaction time is significant, especially between 20 pt and 50 pt. In the case of a large amplitude of 5 Hz, as the font size becomes larger, the increase in the accuracy of interpretation becomes more pronounced. The subjective evaluation of subject’s difficulty in visual search also decreased significantly with the increase of font size, which was consistent with the performance results. It shows that in the vibration environment, choosing a larger font is more conducive to ensuring job performance. The stronger the vibration, the more obvious the effect of font size on performance.

The amplitude and frequency both have a significant effect on the accuracy and reaction time of Chinese characters’ visual search. And the interaction between amplitude and frequency is significant.

Due to the limited experimental conditions, the range of variables selected in this study is still relatively small. This may be the reason that the difference in the character size performance between reaction time and accuracy is not obvious. In addition, previous studies have shown that different font types, arrangements, and other factors will affect the performance of character interpretation search (Ling 2006; Gongkun 2009). In this study, the font type was selected as the Song Dynasty. In the follow-up study, we can examine the influence of different fonts and other factors on the character display performance judgment under the vibration environment.

5 Conclusion

This study shows that with the increase of vibration frequency and amplitude, larger fonts can improve visual interpretation performance, and subjects are subjectively perceived to be easier to operate. In the case of a large amplitude of 5 Hz, the effect of character size on performance is even more pronounced. No matter which kind of vibration state, when the character size is 20 pt, the interpretation performance and subjective evaluation will be significantly reduced. The results of this study for character display design under vibration conditions are that the character size should be greater than 20 pt.