Measuring emotional variables in occupational performance: A scoping review

Abstract

BACKGROUND:

As interest in job-related psychology increased, the need to focus on understanding workplace stress was emphasized. Negative emotional states such as anxiety and stress permeate the organization and, if uncontrolled, can negatively impact the health and work performance of workers. Therefore, attempts to analyze various signals to understand human emotional states or attitudes may be important for future technological development.

OBJECTIVE:

The purpose of this study was to identify what biological variables can discriminate emotions that can significantly affect work results.

METHODS:

Databases (Embase, PsychINFO, PubMed, and CINAHL) were searched for all relevant literature published as of December 31, 2019.

RESULTS:

Brain activity (BA) and heart rate (HR) or heart rate variability (HRV) are adequate for assessing negative emotions, while BA, galvanic skin response (GSR), and salivary samples (SS) can confirm positive and negative emotions.

CONCLUSION:

In the future, researchers should study measurement tools and bio-related variables while workers perform tasks and develop intervention strategies to address emotions associated with work. This may enable workers to perform tasks more efficiently, prevent accidents, and satisfy clients.

Keywords

Bio-signal emotion technology work review

1 Introduction

Emotion is defined as “the mind or mood that occurs when a person faces a phenomenon or event.” The human state of mind can affect behavior; thus, human beings are greatly affected by emotional changes [1]. In recent years, human–machine interaction research has begun to focus on emotion-recognition techniques to make natural interactions a reality [2]. Emotion-awareness technology is a human-centered interaction in which a machine recognizes a person’s emotions and reacts accordingly. Therefore, emotions have become an important aspect of human–machine interaction [3].

In adapting to a fast-changing modern society, people tend to experience frequent emotional changes when they face insecurities in their lives. Therefore, physiological reactions and behavioral patterns as to how humans accept and respond to these modern societies have gained significance. Humans express emotions in various ways, which can be received through hearing (language), vision (facial expression), and biological signals (heart rate), and emotions can play an important role in predicting the state of the human brain. Therefore, attempts to analyze various signals to understand human emotional states or attitudes may be important for future technological development [4, 5].

However, there have been limitations in the research on emotion recognition centered on facial expression and voice. In the case of facial expressions, it may be difficult to extract the facial region in an environment where substantial lighting changes occur or when the user moves. When using voice, ambient noise may cause problems in extracting actual data. In contrast, biological signals (include brain activity, heart rate, eye movement, muscle activity, skin response, and respiration) are less affected by the environment and have the advantage of indicating emotional states physiologically. Additionally, they can be extracted using electrodes or infrared sensors and are less sensitive to social and cultural differences among users [6, 7]. Given the advantages, there has been active research in this area. Valenza and colleague suggests that human signals can be extracted through biological signals such as the heart’s electrical activity and heart rate variability and can help diagnose mental problems such as mood swings [8].

In recent years, much research has been focused on understanding the employee’s emotions and worker performance. Additionally, as interest in job-related psychology increased, the need to focus on understanding workplace stress was emphasized [9, 10]. It has been reported that negative emotional states such as anxiety and stress permeate the organization and, if uncontrolled, can negatively impact the health and work performance of workers. In this way, negative emotional states cause memory biased toward negative events, and negative emotional experiences can rob workers of cognitive resources for job performance [11]. Additionally, these emotional endeavors can negatively affect the maintenance of self-control in subsequent work. Consequently, negative emotions are associated with excessive cognitive evaluation and reduced self-esteem, leading to irrational choices and low work effort [12, 13]. Regehar and LeBlanc’s study [14] reported the results of a study on the negative emotional experience (posttraumatic stress disorder; PTSD) of experts in emergency services and pointed out the problematic high PTSD level of paramedics and police-related workers. In particular, this study demonstrated the possibility of cognitive change and performance decline due to acute stress. In other words, it is explained that the acute stress that occurs depending on the situation increases the possibility of misinterpreting the external situation even if it is mild, distracting more easily, and causing a bias in response.

Jiang and colleagues supported the evidence for biological correlations and pointed out that when human stress is uncontrollable or prolonged, cortisol production increases and acute cortisol elevation affects verbal and social memory impairment and selective attention [15]. Subsequently, emotions that appear as primary responses to stressful situations motivate and promote behavioral responses and increase the likelihood of negative consequences in terms of health, safety, and performance [16]. In another study, emotions influenced learning, and positive emotions helped in motivation, coping strategies, and efficiency [17 –19]. Therefore, recognizing and controlling the emotional state will be an important factor for humans to perform their work efficiently.

In this study, we examined different types of studies to gain a comprehensive understanding of the available information and where it is being researched and developed. This study aimed to identify what biological variables can discriminate emotions that can significantly affect work results.

2 Methods

Scoping reviews provide answers to specific questions rather than assessing an entire body of literature. They survey the literature, synthesize quantitative data about what has been achieved in previous studies, and then summarize and interpret the literature of a particular research field. To date, there have been no systematic reviews of this topic because the understanding of how biological variables can discriminate emotions that can significantly affect work results is a relatively new concept. In particular, it is important to understand how these variables are affected by jobs, the accidents that can be prevented when a subject’s emotional state is properly understood, and the variables that can affect the positive and negative aspects of job performance.

We searched databases for studies published until December 31, 2019. We used the framework proposed by Arksey and O’Malley, which lays down the five stages of a review process [20]:

Stage 1: Identifying the research question

Stage 2: Identifying relevant studies

Stage 3: Study selection

Stage 4: Charting the data

Stage 5: Collating, summarizing, and reporting the results

To select the research questions, a review of studies involving bio-signals helped define the study’s scope. Emotion is an important factor that can affect the efficiency and instability of work, and it is important to pay attention to finding measurable variables that can detect emotions in advance through recent research and a technical analysis of biometric systems using biometric signals [21]. Therefore, this study was a scoping review based on the following exploratory research questions.

What biometric variables can identify emotional states when performing work?

How do you measure emotional variables related to work, and have they been studied?

In the second step, we established the criteria to collect data for inclusion in the study and search for relevant papers. In this study, four search engines (Embase, PsychINFO, PubMed, CINAHL) were used to find related papers. The criteria of the research data included in this study were as follows: (1) journal article type, (2) published from 2008 to the present, (3) written in English, and (4) aimed at confirming bio-related variables that can predict human emotion.

The keywords used in this study were performance, occupation/human emotion, motivation/sensor, and bio-signal. The search was then repeated to select key papers that matched each topic.

In the third step, we identified studies to be included in the scoping review. We integrated the search results from different search engines and removed duplicate papers appearing in more than one engine. We reviewed the title and abstract to exclude articles that do not fit this study. Figure 1 illustrates the process of selecting 14 papers by searching and screening.

Fig. 1

Study selection process.

In the fourth step, from the 14 final articles, we extracted six categories used to analyze the full-text review. Information on task performance, bio-signal variations, attachment site, emotional variations, and highlights were tabulated. If necessary, the categories were modified, and the data extraction scheme was changed accordingly. The questions and disagreements that arose when changing the framework were discussed and resolved by the team. Discrepancies in the extracted data were discussed between the two reviewers until agreement was reached or, if necessary, were resolved through arbitration of a third-party reviewer. In the fifth step, the data were systematically categorized and organized using a data charting form developed in Microsoft Excel [Table 1].

Table 1

Summary of study characteristics and key results from 14 studies

Article citation	Task performance	Bio-signal variations (Questionnaires)	Attachment site	Emotional variations	Highlights
Alazrai et al. [22]	Watching videos	BA	Head	Sadness, Unhappiness, Happiness, Delight	–Using the EEG channels of C₄(all brain region) provides the best emotion classification performance. –Different time-frequency features, which was detected from different EEG channels, could be discriminate various emotion-related information.
Chen et al. [23]	Watching documents	MR	Eyebrow	Anger, Fear, Sadness, Surprise, Disgust	–Showed that different sEMG amplitudes when participants express anger, fear, sadness, surprise and disgust. –Disgust was recognized best. –sEMG has not provide information about specific emotions such as sadness or fear.
Choi et al. [24]	Watching horror movies	BR, FT, BA (Subjective evaluation)	Head	Fear	–FT and subjective evaluation are more reliable than EEG and BR.
Dissanayake et al. [25]	Watching video	HR	Fingers	Joy, Sadness, Pleasure, Anger, Fear, Neutral	–TFB method is an optimal method for emotion recognition using ECG signals. –Ensemble learning method-based model was more effectively.
van Fenema et al. [31]	Musical performance on the stage	HRV, ST, SC, RR, (Psychological questionnaires)	Left pectoral muscle	Stress	–Wearable biosensor patch is appropriate tool for assessing stress. –No correlated between bio-signals and questionnaires.
Kassam et al. [33]	Math tasks	HR, CO, TPR, (Behavior self-reports)	HR (Right arm + Left leg) CO (Neck and torso)	Shame, Anger	–HR, CO and TPR were the main measurable variables of emotional change. –Self-reports of emotions can have a significant effect on physiological responses.
Masood et al. [26]	Writing down and recalling the frightening event	BA	Head	Fear	–BA has been identified as a bio-signal capable of detecting human fear emotions. –Using together two tasks that trigger emotions is effective in analyzing brain signals for human fear feelings.
Matthews et al. [27]	UAS simulation	BA, HR, HC, CBFV (AnTI, HS, SGS, DSSQ, NASA-TLX)	Head	Stress, Anxiety, Worry	–The BA and HR bio-signals were excellent in discriminating emotions such as anxiety and anxiety.
Mozos et al. [28]	Trier Social Stress Test (TSST)	SC, BVP, HRV (Sociometric badge: voice + degree and direction of movement)	Wrist-finger	Stress, Anxiety	–Capable to distinguish between a stressful situation and a neutral situation. –More effective to use two sensors together than to use each one.
Phitayakorn et al. [29]	Operating room simulation	GSR (STAI)	Wrist	Anxiety	–No correlation between self-report and GSR. –Self-report anxiety score is likely to mislead feelings.
Singhal et al. [34]	Emotional oddball paradigm	BA	Head	Fear, Sadness, Neutral	–ERP could be detect emotions well. –Fearful interfered with concentration and had more influence on performance than sad and neutral.
Valenzano et al. [30]	Controlling the RPA	SS, GSR (STAI)	Upper arm	Stress, Anxiety	–SS and GSR was proper bio-signals for measuring stress and anxiety.
Villanueva et al. [32]	Activities in classroom	SS, GSR (Self-report)	Wrist	Active, Afraid, Alert, Ashamed, Attentive, Determine, Distressed, Enthusiastic, Excited, Guilty, Hostile, Inspired, Interested, Irritable, Jittery, Nervous, Proud, Scared, Strong, Upset	–SS did not show significant changes in conditions. –GSR change was greater during the interview, compared to the surveys.
Wells et al. [35]	Playing musical instrument on the stage	HRV	Chest, Wrist	Anxiety	–High state anxiety linked an increase in HRV.

AnTI: Anxious thoughts inventory, BA: Brain activity, BVP: Blood volume pulse, BR: Eye blinking rate, CBFV: Cerebral blood flow velocity, CO: Cardiac output, DSSQ: Short Dundee stress state questionnaire, ERP: Event-related brain potentials, FT: Facial temperature, GSR: Galvanic skin response, HC: Hemodynamic changes, HR: Heart rate, HS: Hardiness scale, HRV: Heart rate variability, MR: Muscle response, NASA-TLX: NASA task load index, RPA: Remotely piloted aircraft, RR: Respiratory rate, SC: Skin conductance, sEMG: Surface electromyography, SGS: Short grit scale, SS: Salivary samples, ST: Skin temperature, STAI: State-trait anxiety inventory, TFB: Ten frequency band, TPR: Total peripheral resistance, UAS: Unmanned aerial systems.

3 Results

3.1 Descriptive study characteristics

A total of 1,331 papers, published from 2008 to 2019, were extracted from the databases using the selected keywords. Of those, a total of 66 papers were removed that were either duplicates or did not provide full text. After screening the titles and abstracts of the remaining articles, 1,184 articles were removed that did not have human subjects, were not related to emotions, or had no reference to bio-signals. Next, the remaining fifty-four papers were reviewed; forty papers containing no experiments, or papers that did not study the work environment, were removed.

3.2 Characteristics of the study settings

The common purpose of all studies extracted was to measure the emotional and physiological variables detected by sensors as humans participate in work. The participants were physically and mentally healthy people, such as medical workers, remote pilots, orchestra performers, and students. There were also studies of adolescent groups experiencing emotional control disorders such as attention deficit hyperactivity disorder (ADHD). Experiments were carried out in the working environment (n = 2) or artificial experimental environment (n = 12). Recently published studies were conducted in a relatively near-real work environment rather than an artificial experimental environment.

3.3 Bio-signals associated with emotion

The analysis demonstrated that the following bio-signals were used in the studies: brain activity (BA), muscle response (MR), eye blinking rate (BR), facial temperature (FT), heart rate (HR), skin temperature (ST), skin conductance (SC), galvanic skin response (GSR), respiratory rate (RR), cardiac output (CO), total peripheral resistance (TPR), hemodynamic changes (HC), cerebral blood flow velocity (CBFV), blood volume pulse (BVP), and salivary samples (SS). Eye blinking rate was used for facial recognition, and emotion could be measured through subjective evaluation methods such as self-reports. Most research papers involved BA (n = 5), and HR (or HRV) (n = 6) and relatively few studies involved GSR (n = 3), ST (or FT) (n = 2), and SC (n = 2). Based on the fourteen studies, BA confirmed sadness, unhappiness, fear, happiness, and delight, and MR identified emotions such as anger, fear, sadness, surprise, and disgust. HR identified sadness, anger, fear, neutrality, joy, and pleasure. Some studies have also used two or more bio-signals together to extract various emotions such as fear, stress, anxiety, shame, anxiety, pride, inspiration, and alertness.

Negative emotional states such as stress can be identified using HR, SC, GSR, ST, FT, SS, MR, and BR [25 , 31]. In particular, non-invasive wearable systems measurements can be used to detect a person’s stress in real-time and provide accurate information. Negative emotions with low arousal levels were extracted using BA, HR, HRV, MR, SS, CO, TPR, and BVP [22 , 32–34]. Kassam and colleague’s study used their own reports and HR, CO, TPR, and BVP, and there was a significant correlation between bio-signals and self-report evaluation [33]. Positive emotions with low arousal levels were extracted using SS and GSR [32]. Conversely, positive emotions with high arousal levels were extracted from BA, GSR, SS, and MR [22 , 32]. Villanueva says that bio-signals, such as SS, can be used with self-report assessments to more accurately identify human emotions [32].

Fig. 2

Emotional valence and arousal in literature.

To date, the research has largely focused on higher arousal states than lower arousal states and negative emotions rather than positive ones. The emotions derived are tabulated based on the emotional dimensions of Mehrabian and Russel [36]. Thus, it was confirmed that the papers mainly focused on negative emotions such as sadness and anger rather than positive emotions such as happiness.

3.4 Suitability of choosing bio-signals

The papers used in this study demonstrate that most bio-signals and other tools, such as BA, HR, HRV, and MR are adequate for assessing negative emotions. Among them, BA, GSR, MR, and SS could confirm positive and negative emotions. The following is a summary of the results of the selected papers in Table 1.

Using HRV, a study evaluating the impact of slow breathing protocols on music players was able to identify changes in anxiety and stress. Valenzano and colleagues conducted research using SS and STAI, including GSR, to determine the anxiety and stress levels of pilots operating remotely controlled aircraft. The result was a significant increase in GSR and SS values, especially at takeoff and landing [30]. The results of the five emotional expressions (happiness, sadness, fear, disgust, and anger) were analyzed through electrical signals using surface EMG (sEMG: Surface electromyography). Each emotional expression had a different amplitude, and aversion was one of the most detected emotions. In a study that suggested the validity of EEG (Electroencephalogram)-based emotion recognition, the qualitative time-frequency distribution approach was used [22 , 34]. Subsequently, EEG captured varied emotion-related information through various time-frequency features extracted from different EEG channels and measured emotions according to four emotion classification methods. In particular, Alazrai and colleagues [22] explained that it is effective to analyze BA in a wide range of brain regions to discern the difference between emotions. In a study of adolescents with impaired emotional control and attention, the ERP markers identified fearful emotions among three emotions: sad, neutral, and fearful. There was also a greater difference between the three emotions in the symptom group than in the non-symptom group [34].

4 Discussion

The 14 selected papers used BA, HRV, GSR, and SS to measure stress and anxiety. MR was used to identify emotions through facial expressions. BA was used to identify emotions according to four emotion classification methods. This study’s main findings are based on the variables, and the results of the selected studies are as follows.

First, among the studies on occupation-based emotional variable measurements, few used bio-signals. Among papers related to the subject from 2008 to 2019, there were 0 to 3 studies each year since 2012. In the Kassam and Mendes paper, emotion was examined using self-report evaluations and questionnaires; this may have led to the measurement error (misclassification) of physiological outcomes, affecting the study validity [33].

Second, there is a lack of validity for using self-report assessments when using bio-signals to recognize emotions. In a paper experimenting with GSR and STAI, there was no correlation between the two tools [29]. A paper using BA claimed that using self-report assessments as well as bio-signals could be more efficient in recognizing emotions. Valenzano’s study, using GSR, SS, and STAI together to measure pilot anxiety and stress, demonstrated significant correlations across all three instruments, positively assessing self-report assessments [30].

Third, emotions identified through biometric tools were concentrated on the negative side. Except for Alazrai’s mention of happiness and delight, all other studies dealt only with negative emotions [22]. The researchers attempted to balance positive emotions and negative emotions, but the research focused on negative emotions; therefore, understanding human emotions in the occupational context is limited. No answer was found on how positive emotions can affect occupation. Positive sentiment expands the range of individual thinking and action alternatives, increasing the capacity to make comprehensive judgments and increase openness to new experiences, helping to explore creative alternatives [37]. Positive emotion also regulates response and recovery from stress, as well as resilience from frustration [38, 39]. Thus, if humans work with positive emotions, they will be able to cope with problems effectively in the face of a crisis and perform their work more efficiently. Therefore, the study of variables measuring positive emotions is important.

Fourth, bio-signal measurements may be more effective in emotion recognition when used in combination with other tools. Choi’s study conducted an experiment to measure horror while watching horror movies [24]. BA did not indicate meaningful emotional identification. It was suggested that using BA with other bio-signals such as HR or ST may yield more meaningful results than using only BA. Mozos’ work also argued that extracting emotions using two bio-signals together can recognize emotions more effectively [28].

Fifth, there were very few studies performed in real situations. When conducting experiments by attaching tools to measure bio-signals to extract emotions, it is impossible to create natural situations because the size and type of equipment limit the experiment’s time and space. There were many relatively simple and abstract situations in which the experimental situation was exposed to video watching or picture stimulus; however, little research has been conducted to measure human emotion in the work environment [22 , 34].

Humans are easily stimulated by individual feelings and their surroundings, and their emotions change rapidly. It is important to provide appropriate interventions to control this, especially because emotionally vulnerable humans are more susceptible to stimulation than healthy humans. For example, one study demonstrated that slow breathing improves work quality by stabilizing HRV, better regulating physiological arousal in high-risk musicians, and reducing anxiety [35]. Additionally, Horner and colleague’s study explained the relationship between the language memory and work ability of paramedics or police recruits who are highly likely to be exposed to PTSD. In particular, they explained that a greater loss of work ability could be experienced in higher situations than in less stressful situations [39 –42]. These emotional responses can also affect cognitive responses, and it will be important to adjust emotional aspects to avoid these negative effects. Therefore, it is necessary to develop intervention methods for controlling human emotions, such as coping with stress, and such interventions are expected to have a positive effect on occupational performance [43 –45].

5 Conclusion

In the future, we should actively study measurement tools and bio-related variables in the performance of emotional tasks and develop mediation strategies that control emotions to enable individuals to perform tasks more efficiently, prevent accidents, or satisfy clients. Additionally, as both positive and negative opinions were indicated in the use of self-report evaluation when recognizing emotions, it should be necessary to study the correlations between bio-signal and self-report evaluation in the future. It is also necessary to research a wide range of emotion recognition rather than focusing on negative emotions. Although some studies have suggested that the combined use of several tools would be more efficient, only one experimental study found this approach effective. Therefore, there is a lack of consensus on the effect of combining two or more tools, and further research is needed to identify emotions that can be detected more effectively recognizing when combining tools. Finally, we suggest that studies should be carried out by presenting more specific situations so that results are generalizable to work scenarios.

This scoping review must be considered within some limitations. First, the review was based on an extensive search using key keywords for peer reviews and gray literature sources, but it is possible that other related publications were not included. We did not evaluate the existing literature’s quality because it is not expected in the evaluation scope. However, as the research increases, the quality evaluation of research on evaluation tools and intervention methods will become an important research method.

Conflict of interest

None to report.

References

Chan

, Singhal

, The emotional side of cognitive distraction: implications for road safety, Accident Analysis & Prevention. 2013;50:147–54.

Shafer

, Matveychuk

, Penney

, O’Hare

, Stokes

, Dolcos

, Processing of emotional distraction is both automatic and modulated by attention: evidence from an event-related fMRI investigation, Journal of Cognitive Neuroscience. 2012;24(5):1233–52.

Lee

, Park

, Emotion recognition for intelligent man-machine interface, Telecommunications Review. . 2012;22(1):446–58.

, Kang

, Lim

, Lee

, Design and implementation of an emotion recognition system using physiological signal, The Journal of the Institute of Internet Broadcasting and Communication. 2010;10(1):57–62.

Lee

, Yoo

, The design of feature selection classifier based on physiological signal for emotion detection, Journal of the Institute of Electronics and Information Engineers. 2013;50(11):2910–20.

Meng

, Romera-Paredes

, Bianchi-Berthouze

, Emotion recognition by two view svm 2k classifier on dynamic facial expression features. Face and Gesture. 2011: IEEE; 2011.

Reisenzein

, Studtmann

, Horstmann

, Coherence between emotion and facial expression: evidence from laboratory experiments, Emotion Review. 2013;5(1):16–23.

Valenza

, Nardelli

, Lanata

, Gentili

, Bertschy

, Paradiso

, et al., Wearable monitoring for mood recognition in bipolar disorder based on history-dependent long-term heart rate variability analysis, IEEE Journal of Biomedical and Health Informatics. 2013;18(5):1625–35.

Tang

, A reciprocal interplay between psychosocial job stressors and worker well-being? A systematic review of the “reversed” effect, Scandinavian Journal of Work, Environment & Health. 2014;40(5):441–56.

10.

Ford

, Cerasoli

, Higgins

, Deccesare

, Relationships between psychological, physical, and behavioural health and work performance: A review and meta-analysis, Work & Stress. 2011;25(3):185–204.

11.

Smith

, Rasmussen

, Mills

, Wefald

, Downey

. Stress and performance: Do service orientation and emotional energy moderate the relationship? Journal of Occupational Health Psychology. 2012;17(1):116–28.

12.

Green

, Draper

, Helton

, The impact of fear words in a secondary task on complex motor performance: A dual-task climbing study, Psychological Research.. 2014;78:557–65.

13.

Rabenu

, Yaniv

, Elizur

, The relationship between psychological capital, coping with stress, well-being, and performance, Current Psychology. 2017;36:875–87.

14.

Regehr

, LeBlanc

, PTSD, acute stress, performance and decision-making in emergency service workers, The Journal of the American Academy of Psychiatry and the Law. 2017;45(2):184–92.

15.

Jiang

, Tran

, Madison

, Bakker

, Acute stress-induced cortisol elevation during memory consolidation enhances pattern separation, Learning and Memory. 2019;26(4):121–7.

16.

Kemp

, Quintana

, Felmingham

, Matthews

, Jelinek

, Depression, comorbid anxiety disorders, and heart rate variability in physically healthy, unmedicated patients: Implications for cardiovascular risk, PLoS One. 2012;7(2):e30777.

17.

Sippel

, Marshall

, Posttraumatic stress disorder and fear of emotions: The role of attentional control, Journal of Trauma Stress. 2013;26(3):397–400.

18.

Montoro

, Useche

, Alonso

, Cendales

, Work environment, stress, and driving anger: a structural equation model for predicting traffic sanctions of public transport drivers, International Journal of Environmental Research and Public Health. 2018;15(3):497.

19.

Jennings

, Children’s understanding of their emotions and the influence of their emotions on their learning. University of California, Berkeley; 1995.

20.

Arksey

, O’Malley

, Scoping studies: Towards a methodological framework, International Journal of Social Research Methodology. 2005;8(1):19–32.

21.

Noh

, Kim

, Smart emotion management system based on multi-biosignal analysis using artificial intelligence, Journal of Institute of Electrical and Electronics Engineers. 2017;21(4):397–403.

22.

Alazrai

, Homoud

, Alwanni

, Daoud

, EEG-based emotion recognition using quadratic time-frequency distribution, Sensors. 2018;18(8):2739.

23.

Chen

, Yang

, Wang

, Eyebrow emotional expression recognition using surface EMG signals, Neurocomputing. 2015;168:871–9.

24.

Choi

, Bang

, Heo

, Park

, Evaluation of fear using nonintrusive measurement of multimodal sensors, Sensors. 2015;15(7):17507–33.

25.

Dissanayake

, Roshan Ragel

, Nawnne

, An ensemble learning approach for electrocardiogram sensor based human emotion recognition, Sensor. 2019;19(20):4495.

26.

Masood

, Farooq

, Investigating EEG patterns for dual-stimuli induced human fear emotional state, Sensor. 2019;19(3):522.

27.

Matthews

, Panganiban

, Wells

, Wohleber

, Reinerman-Jones

, Metacognition, hardiness, and grit as resilience factors in unmanned aerial systems(UAS) operations: A simulation study, Frontiers in Psychology. 2019;26, 640.

28.

Mozos

, Sandulescu

, Andrews

, Ellis

, Bellotto

, Dobrescu

, et al., Stress detection using wearable physiological and sociometric sensors, International Journal of Neural Systems. 2017;27(2):1650041.

29.

Phitayakorn

, Minehart

, Hemingway

, Pian-Smith

, Petrusa

, Relationship between physiologic and psychological measures of autonomic activation in operating room teams during a simulated airway emergency, The American Journal of Surgery. 2015;209(1):86–92.

30.

Valenzano

, Moscatelli

, Messina

, Monda

, Orsitto

, Zezza

, et al., Stress profile in remotely iloted aircraft crewmembers during 2h operating mission, Frontiers in Physiology. 2018;9, 461.

31.

van Fenema

, Gal

, van de Griend , Maxime

, Jacobs

, Cohen

, A pilot study evaluating the physiological parameters of performance-induced stress in undergraduate music students, Digital Biomarkers. 2017;1(2):118–25.

32.

Villanueva

, Valladares

, Goodridge

, Use of galvanic skin responses, salivary biomarkers, and self-reports to assess undergraduate student performance during a laboratory exam activity, Journal of Visualized Experiments. 2016;(108):e53255.

33.

Kassam

, Mendes

, The effects of measuring emotion: Physiological reactions to emotional situations depend on whether someone is asking, PloS One. 2013;8(6):e64959.

34.

Singhal

, Shafer

, Russell

, Gibson

, Wang

, Vohra

, et al., Electrophysiological correlates of fearful and sad distraction on target processing in adolescents with attention deficit-hyperactivity symptoms and affective disorders, Frontiers in Integrative Neuroscience. 2012;6, 119.

35.

Wells

, Outhred

, Heathers

, Quintana

, Kemp

, Matter over mind: a randomised-controlled trial of single-session biofeedback training on performance anxiety and heart rate variability in musicians, PloS One. 2012;7(10):e46597.

36.

Mehrabian

, Russell

, (1974). An approach to environmental psychology. USA: The Massachusetts Institute of Technology. 1974. pp. 216-7.

37.

Fredrickson

, The role of positive emotions in positive psychology: The broaden-and-build theory of positive emotions, American Psychologist. 2001;56(3):218.

38.

Fredrickson

, Tugade

, Waugh

, Larkin

, What good are positive emotions in crisis? A prospective study of resilience and emotions following the terrorist attacks on the United States on September 11th, Journal of Personality and Social Psychology2003;84(2):365.

39.

Horner

, Mintzer

, Turner

, Edmiston

, Brawman-Mintzer

, Attentional functioning in patients with posttraumatic stress disorder: A preliminary study, CNS Spectrums. 2013;18(2):90–4.

40.

Ong

, Bergeman

, Bisconti

, Wallace

, Psychological resilience, positive emotions, and successful adaptation to stress in later life, Journal of Personality and Social Psychology. 2006;91(4):730.

41.

Kim

, A Systematic Review on the Effects of Intervention for Caregivers of People with Dementia to Reduce Their Burden, Therapeutic Science for Neurorehabilitation. 2014;3(1):19–29.

42.

Ahn

, Joung

, Hyun

, Seo

, Socio-psychological Process of Families with Institutionalized Dementia Elderly, Journal of Korean Academy of Community Health Nursing. 2004;15(1):122–31.

43.

Shrivastava

, Alduraywish

, Srivastava

, Alsharari

, Al-Johani

, Sghaireen

, et al., Assessment of knowledge and attitude of allied healthcare professionals about COVID-19 across Saudi Arabia, Work. 2021;68(2):305–15.

44.

Chomem

, Motter

, Analysis of physical and physiological workloads of nursing in the surgical center, Work. 2021;68(2):425–35.

45.

Kinoshita

, Tanigawa

, Maruyama

, Morimoto

, The effects of bright light treatment via ear canals on quality of sleep and depressive mood among overworked employees: A randomized-controlled clinical trial, Work. 2020;67(2):323–29.