Human error studies tend to focus on identifying the relationship between human activity, errors ... more Human error studies tend to focus on identifying the relationship between human activity, errors and their consequences. Accidents and incidents report analysis has been the path followed by several authors in the human error studies field, as it will be discussed in this paper. However, error reports tend to detail technical aspects of the error occurrence but fail to explore the human-behaviour component that might have influenced it. In order to investigate the human behaviour and its relation with accidents and human errors the authors propose to observe individuals working during critical situations, while adopting a methodological approach supported by an experimental protocol, to ensure a rigorous systematization of the data gathering and analysis. This paper presents a cognitive model which accounts for the knowledge of the individual and situation characteristics when analysing the user behaviour during critical situations. It also presents the supporting experimental protocol, and discusses its application in the context of a decision making aid system, employed during maritime pollution crisis management.
We showed in a previous study an additive interaction between intrinsic and extraneous cognitive ... more We showed in a previous study an additive interaction between intrinsic and extraneous cognitive loads and of participants' alertness in an 1-back working memory task. The interaction between intrinsic and extraneous cognitive loads was only observed when participants' alertness was low (i.e. in the morning). As alertness is known to reflect an individual's general functional state, we suggested that the working memory capacity available for germane cognitive load depends on a participant's functional state, in addition to intrinsic and extraneous loads induced by the task and task conditions. The relationships between the different load types and their assessment by specific load measures gave rise to a modified cognitive load model. The aim of the present study was to complete the model by determining to what extent and at what processing level an individual's characteristics intervene in order to implement efficient strategies in a working memory task. Therefore, the study explored participants' cognitive appraisal of the situation in addition to the load factors considered previously -task difficulty, time pressure and alertness. Each participant performed a mental arithmetic task in four different cognitive load conditions (crossover of two task difficulty conditions and of two time pressure conditions), both while their alertness was low (9 a.m.) and high (4 p.m.). Results confirmed an additive effect of task difficulty and time pressure, previously reported in the 1-back memory task, thereby lending further support to the modified cognitive load model. Further, in the high intrinsic and extraneous load condition, performance was reduced on the morning session (i.e. when alertness was low) on one hand, and in those participants' having a threat appraisal of the situation on the other hand. When these factors were included into the analysis, a performance drop occurred in the morning irrespective of cognitive appraisal, and with threat appraisal in the afternoon (i.e. high alertness). Taken together, these findings indicate that mental overload can be the result of a combination of subject-related characteristics, including alertness and cognitive appraisal, in addition to welldocumented task-related components (intrinsic and extraneous load). As the factors investigated in the study are known to be critically involved in a number of real job-activities, the findings suggest that solutions designed to reduce incidents and accidents at work should consider the situation from a global perspective, including individual characteristics, task parameters, and work organization, rather than dealing with each factor separately.
The purpose of this study was to identify the influence of situation complexity and driving exper... more The purpose of this study was to identify the influence of situation complexity and driving experience on subjective workload and driving performance, and the less costly and the most effective strategies faced with a hazard pedestrian crossing. Four groups of young drivers (15 traditionally trained novices, 12 early-trained novices, 15 with three years of experience and 15 with a minimum of five years of experience) were randomly assigned to three situations (simple, moderately complex and very complex) including unexpected pedestrian crossings, in a driving simulator. The subjective workload was collected by the NASA-TLX questionnaire after each situation. The main results confirmed that the situation complexity and the lack of experience increased the subjective workload. Moreover, the subjective workload, the avoidance strategies and the reaction times influenced the number of collisions depending on situation complexity and driving experience. These results must be taken into account to target the prevention actions.
Proceedings of 11th international conference of naturalistic decision making, 2013
The aim of the present study is to identify which of subjective workload dimensions are influence... more The aim of the present study is to identify which of subjective workload dimensions are influenced by driving experience and situation complexity, and which ones influence driving performance. Method: Fifty-seven young drivers [15 traditionally trained novices, 12 early-trained novices, 15 with three years of experience and 15 with at least five years of experience] were randomly assigned to three situations [simple, moderately complex and very complex] in a driving simulator. Self-reported levels of workload during unexpected pedestrian crossings were collected by a questionnaire [NASA-TLX] between each situation. Results and discussion: Three workload subscales decreased with experience and increased with situation complexity, and one subscale was correlated to driving performance.
A study was conducted among 10 operators of 18 and 112 call center of Marseille city. These opera... more A study was conducted among 10 operators of 18 and 112 call center of Marseille city. These operators are fremen responsible for the management of calls. Management of calls includes taking calls, flling about a call via an interface in parallel with information acquisition on the situation, the diagnosis of the situa- tion, the mobilization of appropriated resources (tanker, emergency vehicle, etc) or referral to another ser vice. This call centre operates continuously, and operators must be constantly on site, being confronted thus shiftwork schedules. The aim of this study was to determine the infuence of organisational factors and operators’ perception of work situation on alertness and tension levels. Indeed, factors such as time-of-day, accumulated fatigue, or work activity have an impor- tant effect on these variables. Thus, alertness level is determined by circadian and homeostatic components. Circadian component corresponds to biological rhythm characterised by an alertness increase throughout the day to reach a maximum at 18-19 hr, and an alertness decrease during the night with a minimum at 3-4 hr. Homeostatic component corresponds to accumulated fatigue during wakefulness. According to Thayer’s multi-dimensional theory, tension increase could compen- sate alertness decrease to maintain acceptable performance. Questionnaires were flled by operators every 4 hours during 24 hours. These questionnaires allowed estimating self-reported alertness and tension levels (Thayer’s adjective check- list), and perception of work situation (physic demand, psychological demand, work life/private life interferences, decisional latitude, professional social support, and resources available). In parallel, continuous obser vations were made when operators were on shift in order to obtain an estimation of work amount. Results reveal a modulation of work amount by time-of-day on physical and psycho- logical demands (signifcant effect only at the morning) and show a time-of-day effect on alertness and tension with lower levels at night that during the day. Analyses suggest that this is not the task that is problematic (alertness decrease or tension increase) but the context in which it is performed. Task can be stimulating. Otherwise, different patterns appear depending on work activity characterising the period of work considered and suggesting that operators implement control strategies when work amount is high that would explain low effect of perception of work situation in these periods.
The present study tested the hypothesis of an additive interaction between intrinsic, extraneous ... more The present study tested the hypothesis of an additive interaction between intrinsic, extraneous and germane cognitive load, by manipulating factors of mental workload assumed to have a specific effect on either type of cognitive load. The study of cognitive load factors and their interaction is essential if we are to improve workers' wellbeing and safety at work. High cognitive load requires the individual to allocate extra resources to entering information. It is thought that this demand for extra resources may reduce processing efficiency and performance. The present study tested the effects of three factors thought to act on either cognitive load type, i.e. task difficulty, time pressure and alertness in a working memory task. Results revealed additive effects of task difficulty and time pressure, and a modulation by alertness on behavioral, subjective and psychophysiological workload measures. Mental overload can be the result of a combination of task-related components, but its occurrence may also depend on subject-related characteristics, including alertness. Solutions designed to reduce incidents and accidents at work should consider work organization in addition to task constraints in so far that both these factors may interfere with mental workload.
International Journal of Aviation Psychology, 2007
In real traffic conditions, air traffic controllers rated alertness and tension according to time... more In real traffic conditions, air traffic controllers rated alertness and tension according to time of day (recordings every 6 hr) and time on duty (1st vs. 2nd half of afternoon shifts). The 2 measures were inversely correlated on the 2nd shift half, suggesting a compensatory effect of increased tension on decreased alertness. Task performance was decreased at 7:00, overall in a free recall task, and according to list length (9- vs. 6-item) and presentation modality (visual vs. auditory) in a recognition task. The data suggest that changes in controllers' memory processes with time of day depend on the processing strength induced by task characteristics.
International Journal of Aviation Psychology, 2010
This study investigated the response strategies displayed by air traffic controllers (ATCs) recal... more This study investigated the response strategies displayed by air traffic controllers (ATCs) recalling visually and auditory presented verbal material in a free recall task. The end-of-list advantage for auditory compared to visually presented item lists (modality effect) was explored by using the procedure described by . Results revealed that ATCs' response sequences frequently included ordered end subsequences of 2 to 6 items. These end subsequences were recalled as an initial run at a similar rate in both modalities, although they occurred more frequently in positions other than initial recall positions in the auditory modality. These results replicate and extend those reported in controlled laboratory studies, and this is despite ATCs' extended experience with processing visual information. The results are discussed in terms of the functional limitations of visual and auditory item processing and their relevance to air traffic control activities, training, and selection.
The effects of shift and time-on-shift on alertness and perceived tension, and related physiologi... more The effects of shift and time-on-shift on alertness and perceived tension, and related physiological variables were investigated in satellite controllers working a rapid forward rotating 3-shift system. In controlled laboratory conditions subjective tension and heart rate have been reported to display circadian variations and a marked sensitivity to external factors. We examined whether in real-job conditions circadian variations were masked for these particular variables, unlike for alertness and body temperature which have been repeatedly shown to display circadian variations in these conditions. This hypothesis was tested in a repeated measures design by collecting alertness and tension self-reports, and recording operators' sublingual temperature on 3 occasions on each shift, and heart rate continuously throughout shifts. Alertness and body temperature varied according to a typical diurnal trend, subjective tension was only enhanced on the initial recording of each shift (compared to the remaining), while heart rate displayed an intermediary trend. Intra-subjects correlations revealed a positive relationship between alertness, oral temperature and heart rate, while no such relationship was found for subjective tension.
The study investigated time-of-day effects on task performance in shift workers in different task... more The study investigated time-of-day effects on task performance in shift workers in different tasks (reaction time, discrimination, probe recognition, free recall), by varying task-specific features. On each of six recordings, each programmed on a different day and in a randomised order, operators rated alertness and performed different tasks. Selfrated alertness varied according to a typical diurnal trend. Time of day also affected reaction time (slower responses at 03:00 hours), discrimination performance (lower accuracy at 03:00 hours in the most difficult condition) and recall (superior recall at 07:00 and 11:00 hours following deeper processing at encoding). The data demonstrated time-of-day effects on cognitive processes also involved in many real-job activities, despite the lack of control for a number of exogenous factors known to interfere with performance in work settings. Since in the cognitively more loaded tasks, time-of-day effects depended on task conditions, the findings are of operational concern in shiftwork situations involving differential task requirements. In a real-job setting, performance variations were observed according to time of day and task requirements in a set of cognitively more or less demanding tasks. Task-specific research across the 24-h day enables a better understanding of operators' tasks and the development of supporting technology.
Human error studies tend to focus on identifying the relationship between human activity, errors ... more Human error studies tend to focus on identifying the relationship between human activity, errors and their consequences. Accidents and incidents report analysis has been the path followed by several authors in the human error studies field, as it will be discussed in this paper. However, error reports tend to detail technical aspects of the error occurrence but fail to explore the human-behaviour component that might have influenced it. In order to investigate the human behaviour and its relation with accidents and human errors the authors propose to observe individuals working during critical situations, while adopting a methodological approach supported by an experimental protocol, to ensure a rigorous systematization of the data gathering and analysis. This paper presents a cognitive model which accounts for the knowledge of the individual and situation characteristics when analysing the user behaviour during critical situations. It also presents the supporting experimental protocol, and discusses its application in the context of a decision making aid system, employed during maritime pollution crisis management.
We showed in a previous study an additive interaction between intrinsic and extraneous cognitive ... more We showed in a previous study an additive interaction between intrinsic and extraneous cognitive loads and of participants' alertness in an 1-back working memory task. The interaction between intrinsic and extraneous cognitive loads was only observed when participants' alertness was low (i.e. in the morning). As alertness is known to reflect an individual's general functional state, we suggested that the working memory capacity available for germane cognitive load depends on a participant's functional state, in addition to intrinsic and extraneous loads induced by the task and task conditions. The relationships between the different load types and their assessment by specific load measures gave rise to a modified cognitive load model. The aim of the present study was to complete the model by determining to what extent and at what processing level an individual's characteristics intervene in order to implement efficient strategies in a working memory task. Therefore, the study explored participants' cognitive appraisal of the situation in addition to the load factors considered previously -task difficulty, time pressure and alertness. Each participant performed a mental arithmetic task in four different cognitive load conditions (crossover of two task difficulty conditions and of two time pressure conditions), both while their alertness was low (9 a.m.) and high (4 p.m.). Results confirmed an additive effect of task difficulty and time pressure, previously reported in the 1-back memory task, thereby lending further support to the modified cognitive load model. Further, in the high intrinsic and extraneous load condition, performance was reduced on the morning session (i.e. when alertness was low) on one hand, and in those participants' having a threat appraisal of the situation on the other hand. When these factors were included into the analysis, a performance drop occurred in the morning irrespective of cognitive appraisal, and with threat appraisal in the afternoon (i.e. high alertness). Taken together, these findings indicate that mental overload can be the result of a combination of subject-related characteristics, including alertness and cognitive appraisal, in addition to welldocumented task-related components (intrinsic and extraneous load). As the factors investigated in the study are known to be critically involved in a number of real job-activities, the findings suggest that solutions designed to reduce incidents and accidents at work should consider the situation from a global perspective, including individual characteristics, task parameters, and work organization, rather than dealing with each factor separately.
The purpose of this study was to identify the influence of situation complexity and driving exper... more The purpose of this study was to identify the influence of situation complexity and driving experience on subjective workload and driving performance, and the less costly and the most effective strategies faced with a hazard pedestrian crossing. Four groups of young drivers (15 traditionally trained novices, 12 early-trained novices, 15 with three years of experience and 15 with a minimum of five years of experience) were randomly assigned to three situations (simple, moderately complex and very complex) including unexpected pedestrian crossings, in a driving simulator. The subjective workload was collected by the NASA-TLX questionnaire after each situation. The main results confirmed that the situation complexity and the lack of experience increased the subjective workload. Moreover, the subjective workload, the avoidance strategies and the reaction times influenced the number of collisions depending on situation complexity and driving experience. These results must be taken into account to target the prevention actions.
Proceedings of 11th international conference of naturalistic decision making, 2013
The aim of the present study is to identify which of subjective workload dimensions are influence... more The aim of the present study is to identify which of subjective workload dimensions are influenced by driving experience and situation complexity, and which ones influence driving performance. Method: Fifty-seven young drivers [15 traditionally trained novices, 12 early-trained novices, 15 with three years of experience and 15 with at least five years of experience] were randomly assigned to three situations [simple, moderately complex and very complex] in a driving simulator. Self-reported levels of workload during unexpected pedestrian crossings were collected by a questionnaire [NASA-TLX] between each situation. Results and discussion: Three workload subscales decreased with experience and increased with situation complexity, and one subscale was correlated to driving performance.
A study was conducted among 10 operators of 18 and 112 call center of Marseille city. These opera... more A study was conducted among 10 operators of 18 and 112 call center of Marseille city. These operators are fremen responsible for the management of calls. Management of calls includes taking calls, flling about a call via an interface in parallel with information acquisition on the situation, the diagnosis of the situa- tion, the mobilization of appropriated resources (tanker, emergency vehicle, etc) or referral to another ser vice. This call centre operates continuously, and operators must be constantly on site, being confronted thus shiftwork schedules. The aim of this study was to determine the infuence of organisational factors and operators’ perception of work situation on alertness and tension levels. Indeed, factors such as time-of-day, accumulated fatigue, or work activity have an impor- tant effect on these variables. Thus, alertness level is determined by circadian and homeostatic components. Circadian component corresponds to biological rhythm characterised by an alertness increase throughout the day to reach a maximum at 18-19 hr, and an alertness decrease during the night with a minimum at 3-4 hr. Homeostatic component corresponds to accumulated fatigue during wakefulness. According to Thayer’s multi-dimensional theory, tension increase could compen- sate alertness decrease to maintain acceptable performance. Questionnaires were flled by operators every 4 hours during 24 hours. These questionnaires allowed estimating self-reported alertness and tension levels (Thayer’s adjective check- list), and perception of work situation (physic demand, psychological demand, work life/private life interferences, decisional latitude, professional social support, and resources available). In parallel, continuous obser vations were made when operators were on shift in order to obtain an estimation of work amount. Results reveal a modulation of work amount by time-of-day on physical and psycho- logical demands (signifcant effect only at the morning) and show a time-of-day effect on alertness and tension with lower levels at night that during the day. Analyses suggest that this is not the task that is problematic (alertness decrease or tension increase) but the context in which it is performed. Task can be stimulating. Otherwise, different patterns appear depending on work activity characterising the period of work considered and suggesting that operators implement control strategies when work amount is high that would explain low effect of perception of work situation in these periods.
The present study tested the hypothesis of an additive interaction between intrinsic, extraneous ... more The present study tested the hypothesis of an additive interaction between intrinsic, extraneous and germane cognitive load, by manipulating factors of mental workload assumed to have a specific effect on either type of cognitive load. The study of cognitive load factors and their interaction is essential if we are to improve workers' wellbeing and safety at work. High cognitive load requires the individual to allocate extra resources to entering information. It is thought that this demand for extra resources may reduce processing efficiency and performance. The present study tested the effects of three factors thought to act on either cognitive load type, i.e. task difficulty, time pressure and alertness in a working memory task. Results revealed additive effects of task difficulty and time pressure, and a modulation by alertness on behavioral, subjective and psychophysiological workload measures. Mental overload can be the result of a combination of task-related components, but its occurrence may also depend on subject-related characteristics, including alertness. Solutions designed to reduce incidents and accidents at work should consider work organization in addition to task constraints in so far that both these factors may interfere with mental workload.
International Journal of Aviation Psychology, 2007
In real traffic conditions, air traffic controllers rated alertness and tension according to time... more In real traffic conditions, air traffic controllers rated alertness and tension according to time of day (recordings every 6 hr) and time on duty (1st vs. 2nd half of afternoon shifts). The 2 measures were inversely correlated on the 2nd shift half, suggesting a compensatory effect of increased tension on decreased alertness. Task performance was decreased at 7:00, overall in a free recall task, and according to list length (9- vs. 6-item) and presentation modality (visual vs. auditory) in a recognition task. The data suggest that changes in controllers' memory processes with time of day depend on the processing strength induced by task characteristics.
International Journal of Aviation Psychology, 2010
This study investigated the response strategies displayed by air traffic controllers (ATCs) recal... more This study investigated the response strategies displayed by air traffic controllers (ATCs) recalling visually and auditory presented verbal material in a free recall task. The end-of-list advantage for auditory compared to visually presented item lists (modality effect) was explored by using the procedure described by . Results revealed that ATCs' response sequences frequently included ordered end subsequences of 2 to 6 items. These end subsequences were recalled as an initial run at a similar rate in both modalities, although they occurred more frequently in positions other than initial recall positions in the auditory modality. These results replicate and extend those reported in controlled laboratory studies, and this is despite ATCs' extended experience with processing visual information. The results are discussed in terms of the functional limitations of visual and auditory item processing and their relevance to air traffic control activities, training, and selection.
The effects of shift and time-on-shift on alertness and perceived tension, and related physiologi... more The effects of shift and time-on-shift on alertness and perceived tension, and related physiological variables were investigated in satellite controllers working a rapid forward rotating 3-shift system. In controlled laboratory conditions subjective tension and heart rate have been reported to display circadian variations and a marked sensitivity to external factors. We examined whether in real-job conditions circadian variations were masked for these particular variables, unlike for alertness and body temperature which have been repeatedly shown to display circadian variations in these conditions. This hypothesis was tested in a repeated measures design by collecting alertness and tension self-reports, and recording operators' sublingual temperature on 3 occasions on each shift, and heart rate continuously throughout shifts. Alertness and body temperature varied according to a typical diurnal trend, subjective tension was only enhanced on the initial recording of each shift (compared to the remaining), while heart rate displayed an intermediary trend. Intra-subjects correlations revealed a positive relationship between alertness, oral temperature and heart rate, while no such relationship was found for subjective tension.
The study investigated time-of-day effects on task performance in shift workers in different task... more The study investigated time-of-day effects on task performance in shift workers in different tasks (reaction time, discrimination, probe recognition, free recall), by varying task-specific features. On each of six recordings, each programmed on a different day and in a randomised order, operators rated alertness and performed different tasks. Selfrated alertness varied according to a typical diurnal trend. Time of day also affected reaction time (slower responses at 03:00 hours), discrimination performance (lower accuracy at 03:00 hours in the most difficult condition) and recall (superior recall at 07:00 and 11:00 hours following deeper processing at encoding). The data demonstrated time-of-day effects on cognitive processes also involved in many real-job activities, despite the lack of control for a number of exogenous factors known to interfere with performance in work settings. Since in the cognitively more loaded tasks, time-of-day effects depended on task conditions, the findings are of operational concern in shiftwork situations involving differential task requirements. In a real-job setting, performance variations were observed according to time of day and task requirements in a set of cognitively more or less demanding tasks. Task-specific research across the 24-h day enables a better understanding of operators' tasks and the development of supporting technology.
Uploads
Papers by Edith Galy