Keywords

1 Introduction

Vigilance has become increasingly salient as a critical human factors issue because of the increasing automation of technology (Warm et al. 2008). Vigilance tasks require the observer to keep their focused attention on information displays for prolonged periods of time and detect the appearance of predetermined critical signals, which often occur infrequently (Dillard et al. 2014; Funke et al. 2016; Helton and Russell 2011). The recent advancement of automation technology makes vigilance an even more salient issue. The role of workers in an automated system shifts from manual controllers to system supervisors, who need to monitor the system and act only in instances of problems or system failures (Sheridan 1992). Moreover, automation has been introduced into many safety-critical domains such as military surveillance (Dillard et al. 2014; Helton and Russell 2011), aircraft and air traffic control (Langan-Fox et al. 2009), vehicle operation (Young and Stanton 2007), nuclear power plant and process control (Baker et al. 1994), and health care (Drews 2013; Sheridan 2002). Therefore, suboptimal performance in vigilance tasks can lead to severe accidents and consequences, as reported by several studies (Hawley et al. 2006; Molloy and Parasuraman 1996; Warm et al. 2008).

Vigilance decrement is a major finding of prior vigilance research. It has been observed in a variety of industrial, transportation, security, and medical settings (Parasuraman 1986; Warm et al. 2008). Vigilance decrement refers to loss of detection of critical stimuli during a prolonged, continuous work shift (Kamzanova et al. 2014). In laboratory studies, vigilance decrement is commonly observed within the first 20 to 30 min of performance (Molloy and Parasuraman 1996) and on some tasks, over even shorter durations (Gillberg and Åkerstedt 1998; Temple et al. 2000). The resource model theory, proposed by Parasuraman and Davies (1977), posits that the need to make continuous signal-noise discriminations under great uncertainty without rest depletes information-processing resources that cannot be replenished during task performance, leading to a decline in performance efficiency. From the perspective of attention, mindlessness theory asserted that the repetitive and monotonous nature of vigilance tasks lead to a lack of attentional focus on the task (Manly et al. 1999; Robertson et al. 1997). During sustained attention, the attention often drifts to some thoughts unrelated to task over time (Smallwood et al. 2004).

Due to vigilance decrement and the importance of vigilance performance for system safety, many scholar efforts have been devoted to study factors that affect vigilance performance. Some factors are related to features of stimuli, such as frequency of signals, temporal and spatial consistency of signals, and signal salience. Due to the valuable implication for personnel selection and training, many researches focus on influence of individual difference on vigilance performance and subjective experience (Finomore et al. 2009; Matthews et al. 2014; Reinerman-Jones et al. 2011; Shaw et al. 2010). Dittmar et al. (1993) found perceptual sensitivity for critical signals favored males in the monitoring the repetitive presentation of spatial task, and females felt the spatial task significantly more frustrating, mentally demanding, and effortful than males. Considering the five factor model of personality proposed by Costa and McCrae (1992), introversion and consciousness have been found positively related to vigilance performance (Koelega 1992; Rose et al. 2002).

Despite the abundant research on impacts of individual difference on vigilance, no study has examined the impact of cultural backgrounds on vigilance. Culture is often defined by nationality, race and organization, and we focused on national culture in this study. Over 40-years cross-cultural psychology research has shown that a wide range of psychological processes are influenced by culture, including attention (Masuda et al. 2008), perception, memory (Rule et al. 2011), and decision making (Chu et al. 2005). Because sustained attention on the stimuli over prolonged periods of time is an essential component of vigilance task, it is suggested cultural difference in the attentional process may indicate differences in vigilance.

Prior cross-cultural studies have found difference in cognition and attention between East Asians and Westerners. First, the different thinking style would bring cultural variation in attention pattern between East Asians and Westerners. East Asians tend to develop a holistic strategy of thinking style that emphasizes the entire field and relationships between the focal object and the context, whereas Westerners tend to develop a focused strategy of thinking style that emphasizes focusing on focal objects and ignoring contextual information (Duffy and Kitayama 2007; Duffy et al. 2009). It was found that East Asians paid more attention to the context and that Westerners paid more attention to the focal object (Chua et al. 2005; Masuda and Nisbett 2006). Second, East Asians are more field-dependent and Westerners are more field-independence (Ji et al. 2000). Field dependence – independence refers to the degree to which perception of an object is influenced by the background or surroundings in which it appears. Ji et al. (2000) used rod and frame test to measure field independence between East Asians and Westerners. They found East Asians were more susceptible to the background, and Westerners were more confident about their performance.

The cross-culture difference in cognition and attention has been confirmed by a series of studies using different paradigms, including cognitive experimentation, eye movement patterns, and neuroscience evidence (Masuda and Nisbett 2001; Masuda et al. 2008; Masuda and Nisbett 2006). Goto et al. (2010) used event-related potential index N400 to test cultural difference in the visual processing of meaning, with a task of detecting incongruities between background and foreground objects. The result suggested that Asians were processing the relationship between foreground and background objects to a greater degree than Westerners, which was consistent with hypothesized greater holistic processing among East Asians. Researches on field dependence have found that field-dependent people had greater difficulty in maintaining attention on specific sectors of information (Avolio et al. 1981; Guisande et al. 2007; Jolly and Reardon 1985). Cahoon (1970) found that field dependence was negatively correlated with vigilance performance in terms of higher false alarm rate and lower perceptual sensitivity.

Therefore, cultural background influences the attention allocation. East Asians tend to allocate attention in a broad region (holistic attention strategy), whereas Westerners tend to focus attention on focal objects or events (focused attention strategy). It was also suggested that field-dependent people tend to have a broader, less efficient focused attention on the ongoing task, whereas field-independent people tend to have a narrow and efficient focused attention. Vigilance task is characterized by requiring observers to maintain focused attention on the focal targets or objects for a prolonged period of time. Considering cultural difference on attention pattern, we expect that culture has an influence on vigilance. Therefore, we propose the following hypothesis:

  • Hypothesis 1. Americans perform better than Chinese in single vigilance task.

  • Hypothesis 2. Chinese have higher levels of perceived workload than Americans in single vigilance task.

  • Hypothesis 3. Chinese have higher levels of perceived fatigue than Americans in single vigilance task.

The purpose of this research is to examine whether cultural differences (Chinese VS Americans) exist in performance and subjective experience associated with the vigilance task. The performance includes accuracy and response time. The subjective experience includes workload and fatigue. This study used Continuous Performance Test (CPT), a typical and simple vigilance task, in the experiment to test the cultural difference. It was shown to have high test-retest reliability (Borgaro et al. 2003).

2 Methodology

2.1 Participants

Twenty Chinese students (14 males, 6 females; age range = 20–28 years, mean age = 24 years) and twenty American students (14 males, 6 females, age range = 19–28 years, mean age = 24 years) at Tsinghua University participated in the experiment. The American participants had stayed in China for an average time of 5.25 months (SD = 3.52). Thirty percent (30%) of them were exchange students from the United States, and the rest were students who studied Chinese language at Tsinghua University. All participants had normal or corrected-to-normal vision.

2.2 Procedure

All observers monitored the 23-min repetitive presentation of symbols centered on the screen without interruption. The symbols were constructed in 100-point type on a white background. In order to assess culture differences on sustained attention, it is necessary to use culture-free stimuli in the experiment. We adapted the original CPT task by replacing the stimuli letters with a series of mathematical symbols (e.g. “⊥”, “⋀”, “×”, “±”, “∠”, “=”, “+”). A pilot test was conducted to confirm a suitable frequency of stimuli in order to avoid the ceiling effect of performance. As a result, each stimulus exposed for 550 ms and there was no interval between stimuli in the experiment.

For each observer, the order of presentation of stimuli was varied at random with the target signal (angle symbol, “∠”) occurred at a probability of p = 0.25. Observers signified their detection of target signals by pressing the spacebar key on a keyboard directly in front of them. Responses occurring within 550 ms during the presentation of a target signal were recorded as correct detections (hits). No response was required for other stimuli. All other responses were recorded as errors of commission (false alarms). Once the observer responded to a target signal or non-target signal, the stimulus disappeared immediately and the next stimulus appeared.

Perceived workload and fatigue were measured by a paper-pencil version of NASA-TLX and fatigue scale (Matthews and Desmond 1998). The scales were administered immediately following the CPT task. The NASA-TLX is a well-regarded instrument for the measurement of perceived mental workload (Hart and Staveland 1988). It provided a global measure of workload on a scale from 0 to 20 and also identified the relative contributions of six sources of workload: mental demand, temporal demand, physical demand, performance, effort, and frustration. Participants first provided ratings on the six subscales, and then conducted pair-wise comparisons to determine the relative importance of each subscale to the global workload score. The fatigue scale was an adaption of the scale developed by Matthews and Desmond (1998), which was designed to assess the fatigue induced by a vigilance task. After deleting some items unrelated to CPT task, fourteen items remained with three dimensions of fatigue, including perceptual fatigue, boredom and physical fatigue.

3 Results

3.1 Performance

Percentage of Correct Detections.

Mean percentage of correct detections for Chinese and Americans are plotted as a function of periods of watch in Fig. 1. It is evident in the figure that the detection scores were generally quite high, exceeding 80% in all cases. The overall detection probability was greater for Americans (M = 89.8%) than Chinese (M = 82.4%), and moreover both groups demonstrated a vigilance decrement over time. The figure shows that there was a performance advantage favoring Americans and that detection probability in both groups declined over the duration of the vigilance task. These observations were supported by an analysis of variance (ANOVA), 2 (cultural background) × 7 (period) mixed ANOVA, conducted on an arcsine transformation of the detection scores. In this analysis, and all subsequent CPT task analyses, the data was analyzed as seven sequential periods of work. Box’s epsilon was used to correct the degrees of freedom if the Sphericity assumption was violated. As a result, it was found that the main effect for cultural background was significant, \( F_{1,38} = \, 4.91 \), p = .033, \( \upeta_{\text{p }}^{2} = \, .114 \), but the main effect of period and the interaction between these factors were not significant (p > .05).

Fig. 1.
figure 1

Percentage of correct detections for Americans and Chinese

Full size image

Percentage of False Alarm.

False alarms were rare in this study. The mean percentage of false alarm was 1.7% for Americans and 2.0% for Chinese, respectively. A 2 (cultural background) × 7 (period) mixed ANOVA analysis, based on the square root transformation of false alarm score, revealed a significant interaction effect between these factors, as shown in Fig. 2, \( F_{5.5.211} = \, 2.29 \), p = .041, \( \upeta_{\text{p }}^{2} = \, .057 \). There was no significant main effect for cultural background or period. It can be seen in Fig. 2 that percentage of false alarm for Chinese are greater Americans in the previous periods while it fluctuations in the later periods.

Fig. 2.
figure 2

Percentage of false alarm for Americans and Chinese

Full size image

Reaction Time.

Mean reaction time for Chinese and Americans are plotted as a function of periods of watch in Fig. 3. The overall reaction time was 400.36 ms for Americans and 414.72 ms for Chinese. The figure also shows that there was a performance advantage favoring Americans. These observations were supported by an analysis of variance (ANOVA), 2 (cultural background) × 7 (period) mixed ANOVA. Significant main effects were observed for culture, \( F_{1,38} = \, 7.21 \), p = .011, \( \upeta_{\text{p }}^{2} = \, .159 \), and period, \( F_{5.4,204} \) = 3.30, p = .006, \( \upeta_{\text{p }}^{2} = \, .080 \), via a 2 (cultural background) × 7 (period) mixed ANOVA analysis. However, the interaction effect between these factors was not significant, as shown in Fig. 3.

Fig. 3.
figure 3

Reaction time for Americans and Chinese

Full size image

3.2 Subjective Experience

Workload.

The mean (and standard errors) for global workload scores and six dimensions on the NASA-TLX for Chinese and Americans are displayed in Table 1. Cultural differences on these six dimensions were examined with t-test, or Mann-Whitney U test, if the normality assumption was violated. The global workload of Chinese was marginally higher than that of Americans, t (38) = −1.88, p = .068, d = −0.59. In terms of dimensions, Chinese rated significantly higher than Americans on physical demand (U = 79.0, p = .001, r = 0.53) and effort (U = 122, p = .035, r = 0.33). Chinese participants felt higher levels of physical demand and paid more efforts than American participants. For both Americans and Chinese, the overall workload on the two tasks was substantial.

Table 1. Ratings of NASA-TLX scales for Americans and Chinese
Full size table

Fatigue.

The mean (and standard errors) for three dimensions of fatigue scales for Chinese and Americans are displayed in Table 2. The reliability was analyzed by Cronbach’s alpha. For perceptual dimension, Cronbach’s alpha of Chinse was 0.912 and Cronbach’s alpha of Americans was 0.815. For physical dimension, Cronbach’s alpha of Chinse was 0.813 and Cronbach’s alpha of Americans was 0.618. For boredom dimension, Cronbach’s alpha of Chinse was 0.916 and Cronbach’s alpha of Americans was 0.819. Cultural differences on these three dimensions were examined with t-test, or Mann-Whitney U test, if the normality assumption was violated. Other than perceptual fatigue, Chinese rated significantly higher than Americans on physical demand (U = 122.5, p = .035) and boredom (U = 121.5, p = .033). Chinse participants felt higher levels of physical fatigue and boredom than American participants.

Table 2. Ratings of fatigue scales for Americans and Chinese
Full size table

4 Discussion

We developed a typical and basic continuous performance task (CPT) to examine the effect of cultural background (Chinese and Americans) on performance and subjective experience in vigilance task. Generally speaking, Americans performed better than Chinese in the experiment. The overall detection probability was greater for Americans (M = 89.8%) than Chinese (M = 82.4%). The percentage of correct detections generally showed a downward trend, which was consistent with vigilance decrement. However, only the main effect of cultural background was significant on percentage of correct detections (\( F_{1,38} = \, 4.91 \), p = .033, \( \upeta_{\text{p }}^{2} = \, .114 \)). The main effect of period was not significant. A pilot test was conducted to confirm a suitable frequency of stimuli. Each stimulus exposed for 550 ms and there was no interval between stimuli in the experiment. The rate of stimuli in previous studies of CPT vigilance task was 60/min or less (Rosenberg et al. 2013; Temple et al. 2000). However, the task in this study was very simple because there is no mask covered the entire visual field and it was easy to differentiate target signals from non-target signals. A rate of 60/min result in ceiling effect since all target signals could be detected correctly. Therefore, stimulus exposed for 550 ms as a result of a pilot test. Previous studies of longer-duration CPTs have found a variety of changes in performance over time: While some have observed decrements over 30 min (Grier et al. 2003; Helton et al. 2005). The CPT vigilance task lasted about 23 min in this study. It showed a downward trend of correct detections and vigilance decrement may be significant during a longer period.

Besides accuracy, Americans responded to target signals faster than Chinese. The overall reaction time was 400.36 ms for Americans and 414.72 ms for Chinese. The main effect of cultural background was significant on reaction time (\( F_{1,38} = \, 7.21 \), p = .011, \( \upeta_{\text{p }}^{2} = \, .159 \)). However, the main effect of period was not significant on reaction time. The reason may be the same as the percentage of correct detections. Because the task in this study was simple and duration of the task was relatively short, vigilance may not obviously decline. For percentage of false alarm, the main effects of cultural background and period were not significant. Percentage of false alarm was rare in this study because it was 1.7% for Americans and 2.0% for Chinese. It has little effect on performance evaluation in vigilance task. Therefore, the result of CPT in this study showed that Chinese participants missed more target signals and responded more slowly than American participants.

This result can be explained by difference in attention allocation between East Asians and Westerners. East Asians tend to develop a holistic strategy of attention and Westerners tend to develop a focused strategy of attention (Duffy and Kitayama 2007; Duffy et al. 2009). Moreover, East Asians are more field-dependent and Westerners are more field-independence (Ji et al. 2000). East Asians tend to allocate attention in a broad region, whereas Americans are inclined to focus attention on focal objects or events (Chua et al. 2005; Masuda and Nisbett 2006). The vigilance task is characterized by requiring observers to maintain focused attention on the focal targets or objects for a prolonged period of time. The divided attention strategy of East Asians may lead them to be more vulnerable to vigilance failures than Americans, who are dominated by focused attention strategy.

In terms of six dimensions of NASA-TLX scales, Chinese rated significantly higher than Americans on physical demand (U = 79.0, p = .001, r = 0.53) and effort (U = 122, p = .035, r = 0.33). Chinese participants felt higher levels of physical demand and paid more efforts than American participants. For three dimensions of fatigue scales, Chinese rated significantly higher than Americans on physical demand (U = 122.5, p = .035) and boredom (U = 121.5, p = .033). Chinese participants felt higher levels of physical fatigue and boredom than American participants. In CPT vigilance task, Chinese participants felt more physical demand and higher levels of physical fatigue. Moreover, Chinese participants paid more effort to complete the CPT vigilance task but found they felt higher levels of boredom, loss of motivation during the experiment.

The subjective experience in the task was consistent with the explanation of difference in attention allocation between East Asians and Westerners. East Asians tend to allocate attention in a broad region and Americans are inclined to focus attention on focal objects or events. During the vigilance task, observers are required to focus on the focal object for a long period. According to mindlessness theory proposed by Robertson and his colleagues (Manly et al. 1999; Robertson et al. 1997), the repetitive and tedious nature of vigilance tasks leads the observers to disengage the attention from the ongoing activity and approach it in a thoughtless manner. Using a modification of the standard vigilance paradigm, the authors demonstrated that the vigilance performance is primarily determined by the duration of time over which attention must be maintained on the tasks. Chinese tend to develop a holistic strategy of attention and they are used to switch attention within a board region. As a result, Chinese participants consumed more cognitive resources and physical demand to control their attention on the local object during the task but had higher levels of perceived fatigue at the end of the task.

5 Conclusions

5.1 Theoretical Implications

This study conducted a typical and basic Continuous Performance Test (CPT) to examine the effect of cultural background on vigilance task. Result of this study reveals the difference in patterns of attention allocation between Chinese and Americans. The vigilance task is an attention demanding assignment, which requires the observers to maintain their focus attention on targets for prolonged periods of time. American participants performed better than Chinese and they had lower levels of perceived workload and fatigue than Chinese. The difference in performance and subjective experience indicates that American participants are inclined to focus attention on focal objects and that Chinese participants tend to allocate attention in a broad region.

5.2 Practical Implications

Exploring the effects of culture on vigilance task is of great importance for practical applications of vigilance research. As noted by Chapanis (1974), “failure to take account of national and cultural variables may nullify the gains that one might reasonably expect to follow from the application of ergonomics in many areas of the world”. The personnel selection and assessment criteria for vigilant operators, which were primarily derived from Western cultures, may not be appropriate for Eastern cultures. Moreover, the automated systems which are designed according to Western minds may do not fit Eastern people. In the field of aviation, especially for cockpit operations, accumulating data indicated that there are substantial cultural differences in the way pilots conducts their work. Then, the researchers suggested that the training for crew resource management, which includes situation awareness and vigilance, should be adapted to national culture so as to make it more effective (Helmreich et al. 2001).

5.3 Limitations and Future Research

A major limitation of this study is controlling confounding variables that may influence performance and subjective experience in the task. Impact of individual differences on vigilance are not controlled in the experiment. In addition, the duration of the task is a little short since the main effect of period was not significant on correct detections and reaction time. Vigilance decrement is a major finding of prior vigilance research and longer duration may showed an obvious decrement of performance because the task in the experiment is very simple and basic.

The future research may be examine the effect of cultural background on performance and subjective experience in multiple vigilance tasks. Based on finding that East Asians tend to allocate attention in a broad region and Americans are inclined to focus attention on focal objects or events in this study, we expect that Chinese perform better than Americans in multiple vigilance task.