The Impact of Emotional States on Construction Workers’ Recognition Ability of Safety Hazards Based on Social Cognitive Neuroscience

Abstract

The construction industry is one of the most dangerous industries with grave situation owing to high accident rate and mortality rate, which accompanied with a series of security management issues that need to be tackled urgently. The unsafe behavior of construction workers is a critical reason for the high incidence of safety accidents. Affective Events Theory suggests that individual emotional states interfere with individual decisions and behaviors, which means the individual emotional states can significantly influence construction workers’ unsafe behaviors. As the complexity of the construction site environment and the lack of attention to construction workers’ emotions by managers, serious potential emotional problems were planted, resulting in the inability of construction workers to effectively recognize safety hazards, thus leading to safety accidents. Consequently, the study designs a behavioral experiment with E-prime software based on social cognitive neuroscience theories. Forty construction workers’ galvanic skin response signals were collected by a wearable device (HKR-11C+), and the galvanic skin response data were classified into different emotional states with support vector machine (SVM) algorithm. Variance analysis, correlation analysis and regression analysis were used to analyze the influence of emotional states on construction workers’ recognition ability of safety hazards. The research findings indicate that the SVM algorithm could effectively classify galvanic skin response data. The construct ion workers’ the reaction time to safety hazards and emotional valence were negatively correlated, while the accuracy of safety hazards recognition and the perception level of safety hazard separately had an inverted “U” type relationship with emotional valence. For construction workers with more than 20 years of working experience, work experience could effectively reduce the influence of emotional fluctuations on the accuracy of safety hazards identification. This study contributes to the application of physiological measurement techniques in construction safety management and shed a light on improving the theoretical system of safety management.

Full text

Frontiers in Psychology | www.frontiersin.org 1 June 2022 | Volume 13 | Article 895929
ORIGINAL RESEARCH
published: 16 June 2022
doi: 10.3389/fpsyg.2022.895929
Edited by:
Jiayu Chen,
City University of Hong Kong,
Hong Kong SAR, China
Reviewed by:
Chaojie Fan,
City University of Hong Kong,
Hong Kong SAR, China
Guangchong Chen,
City University of Hong Kong,
Hong Kong SAR, China
*Correspondence:
Anni Yu
15705732971@163.com
Specialty section:
This article was submitted to
Emotion Science,
a section of the journal
Frontiers in Psychology
Received: 15 March 2022
Accepted: 18 May 2022
Published: 16 June 2022
Citation:
Chong D, Yu A, Su H and
Zhou Y (2022) The Impact of
Emotional States on Construction
Workers’ Recognition Ability of Safety
Hazards Based on Social Cognitive
Neuroscience.
Front. Psychol. 13:895929.
doi: 10.3389/fpsyg.2022.895929
The Impact of Emotional States on
Construction Workers’ Recognition
Ability of Safety Hazards Based on
Social Cognitive Neuroscience
DanChong
1, AnniYu
1
*, HaoSu
1 and YueZhou
2
1 Department of Management Science and Engineering, Shanghai University, Shanghai, China, 2 Shanghai Urban
Construction Road Engineering Co., Ltd, Shanghai Road & Bridge (Group) Co., Ltd, Shanghai, China
The construction industry is one of the most dangerous industries with grave situation
owing to high accident rate and mortality rate, which accompanied with a series of security
management issues that need to betackled urgently. The unsafe behavior of construction
workers is a critical reason for the high incidence of safety accidents. Affective Events
Theory suggests that individual emotional states interfere with individual decisions and
behaviors, which means the individual emotional states can signicantly inuence
construction workers’ unsafe behaviors. As the complexity of the construction site
environment and the lack of attention to construction workers’ emotions by managers,
serious potential emotional problems were planted, resulting in the inability of construction
workers to effectively recognize safety hazards, thus leading to safety accidents.
Consequently, the study designs a behavioral experiment with E-prime software based
on social cognitive neuroscience theories. Forty construction workers’ galvanic skin
response signals were collected by a wearable device (HKR-11C+), and the galvanic skin
response data were classied into different emotional states with support vector machine
(SVM) algorithm. Variance analysis, correlation analysis and regression analysis were used
to analyze the inuence of emotional states on construction workers’ recognition ability
of safety hazards. The research ndings indicate that the SVM algorithm could effectively
classify galvanic skin response data. The construct ion workers’ the reaction time to safety
hazards and emotional valence were negatively correlated, while the accuracy of safety
hazards recognition and the perception level of safety hazard separately had an inverted
“U” type relationship with emotional valence. For construction workers with more than
20 years of working experience, work experience could effectively reduce the inuence
of emotional uctuations on the accuracy of safety hazards identication. This study
contributes to the application of physiological measurement techniques in construction
safety management and shed a light on improving the theoretical system of
safety management.
Keywords: safety management, construction workers, emotional states, safety hazards, galvanic skin response

Frontiers in Psychology | www.frontiersin.org 2 June 2022 | Volume 13 | Article 895929
Chong et al. Emotional States Impact Recognition Ability
INTRODUCTION
With a large number of employees, the construction industry
is a typical labor-intensive industry worldwide. Every year,
over 60,000 work-related fatalities are reported from construction
workplaces around the world (Lingard, 2013). e construction
industry is a pillar industry in China, from 2000 to 2020, the
number of people in the construction industry has increased
from 19.94 million to 53.67 million, with an annual benet
of 729.96 billion RMB in the construction industry in 2020,
an increase of 3.5% over the previous year (China National
Bureau of Statistics, 2020). However, the occupational safety
of construction workers is not guaranteed with frequent safety
accidents in the construction industry. e total number of
safety accidents in the construction industry has remained
high over the years (China Construction Industry Association,
2020). According to the Ministry of Housing and Urban–Rural
Development of China, 773 production safety accidents occurred
in 2019 in housing and municipal engineering in China, with
904 deaths, an increase of 39 safety accidents and 64 fatalities
over 2018, up5.31 and 7.62%, respectively (Ministry of Housing
and Urban–Rural Development of China, 2019). In other
countries, construction casualty rates are also disproportionately
high compared to the number of people employed (Al-Bayati
etal., 2019; Liu etal., 2020). Overall, the construction industry
continues to experience a disproportionate share of work-related
injuries and illnesses, which signicantly contributes to work-
related fatalities (Pandit etal., 2019). With the frequent occurrence
of safety accidents in the construction industry, the aim of
research on safety management has gradually changed from
the specic environment to human factors. Currently most of
the construction workers in China are migrant workers, who
have strong risk-taking and uke psychology. e coarse
management mode in the construction industry fails to take
care of the psychological needs of the employees and brings
serious hidden mental health problems.
Accident Causation eory suggests that human factors are
the main factor in the frequency of safety accidents (Salminen
and Tallberg, 1996). Accurate identication and assessment of
the potential consequences of construction site safety hazards
by construction workers is an important prerequisite for safety
management (Farmer and Chambers, 1929). e better the
construction workers’ ability to identify and assess the visible
or potential safety hazards in the construction sites, the smaller
are the chances of their unsafe behaviors occurring (Perlman
etal., 2014; Namian etal., 2016). rough a statistical analysis
of the causes of 75,000 injuries and fatalities occurred in
enterprises, American safety engineer Heinrich concluded that
more than 88% of safety accidents were caused by unsafe
human behavior (Heinrich, 1941). e analysis of safety accident
surveys in the construction industry also showed that unsafe
behavior of construction workers was a common cause of safety
accidents (Haslam et al., 2005). e main factors aecting
unsafe behavior of construction workers can be classied into
three aspects: individual factors, organizational factors and
environmental factors (Abdelhamid and Everett, 2000; Zhou
etal., 2008). Individual inuences include psychological factors,
physiological factors and the physical quality of the worker,
which lead to unsafe behavior of the worker under a single
or multiple factors (Alizadeh etal., 2015; Namian etal., 2016).
An individual’s safety behavior is aected by emotional state;
Aective Event eory (AET) suggests that employees’ behavior
and performance at work are largely determined by the changes
in their emotions at each moment rather than their attitudes
or personalities (Weiss and Cropanzano, 1996). e AET has
been demonstrated eectively in the areas of mine worker
safety behavior (Yang etal., 2020), driver driving safety (Muller
et al., 2014) and among other areas. Kajiwara veried that
emotions can inuence the productivity and accuracy of workers
in a logistics picking system (Kajiwara et al., 2019). Manzoor
developed an agent-based computational social agent model
to explore how decisions can be aected by regulating the
emotions involved, and how emotions are aected by emotion
regulation and contagion (Manzoor and Treur, 2015). However,
people do not always think rationally when they act, thus
they oen make irrational choices or decisions when they are
“emotionally driven.” erefore, employees’ emotional states
and the long-term emotions accumulated from emotional
fragments can interfere with individual decisions and behaviors
(Rachlin, 2003).
e psychological factors are considered as the most important
contributors of unsafe behavior for construction workers, and
the psychological activities are inuenced by individual emotions.
Ekman divided emotions into six basic emotions that are
sadness, happiness, anger, disgust, fear and surprise (Ekman,
1992b). Zelenski classied all emotional states into positive
and negative emotions (Zelenski et al., 2012). Emotions can
bemeasured in three dimensions including personal physiological
changes, subjective feelings and external expressions (Kim etal.,
2013). When an emotion occurs, it will increase the heart
rate, dopamine secretion or brain activity. ese changes can
be reected by physiological signals, such as galvanic skin
response, blood pressure, respiration amplitude and brain waves,
which can be collected by wearable devices in real time
(Dzedzickis et al., 2020). Galvanic skin response is a highly
relevant physiological signal for individual emotions. Four-
channel biosensors were used to measure electromyogram,
electrocardiogram, skin conductivity and respiration changes,
by using an extended linear discriminant analysis (pLDA), Kim
et al. developed a novel scheme of emotion-specic multilevel
dichotomous classication (EMDC) with an accuracy of 95%
(Kim and André, 2008). Zhao et al. (2018) used a sensor-
enriched wearable wristband to measure the three physiological
signals including blood volume pause, electrodermal activity
and skin temperature. ey classify the emotions into four
types in aspect of arousal and valence. Zhang choose four
physiological signals including photoplethysmography, galvanic
skin response, respiration amplitude and skin temperature.
Recursive Feature Elimination-Correlation Bias Reduction-
Support Vector Machine (SVM-RFE-CBR) algorithm was used
for the classication (Chen etal., 2020). Izard etal. determined
an individual’s emotions by questionnaire (Dougherty et al.,
1974). Watson proposed the positive and negative emotion
scale, ten adjectives were applied to express their individual

Chong et al. Emotional States Impact Recognition Ability
Frontiers in Psychology | www.frontiersin.org 3 June 2022 | Volume 13 | Article 895929
emotions at work, and the results reect the individual’s
accumulated emotions and emotional experience (Watson etal.,
1988). is scale can describe the emotions eectively for its
simplicity and interpretation. External expressions refer to the
external changes that can bevisually observed under a stimulus,
such as changes in facial expressions, tone of voice and behavior.
Nevertheless, this method lacks objectivity because the external
performance of individuals can be hidden or disguised
(Ekman, 1992a).
e aective generalization theory (Johnson and Tversky,
1983) suggests that emotions irrelevant to the decision-making
task will aect people’s judgments about the probability of events
with the same emotion valence. Specically, positive emotions
reduce the subjective estimate of risky events and people in
positive mood are prone to perform risky behaviors, while
people in negative emotions are prone to perform risky behaviors.
In contrast, Mood Maintenance Hypothesis (Isen and Patrick,
1983) is the other classic theory about emotional valence, and
it refers to people’s tendency to maintain positive mood states
and implies that positive mood is associated with less critical
thinking and reduced information processing, and is prone to
reduce their estimates of risk events and less likely to take
risky behavior. Positive emotions promote brain mental activity
and thus, can enable individuals maintain a higher level of
concentration (Pool et al., 2016). Fredrickson found the
relationship between positive emotions and unsafe behaviors
has a U-shaped eect (Fredrickson and Branigan, 2005), while
negative emotions can decrease individual’s attention,
responsiveness and reasoning abilities (MacLeod et al., 1986).
e more intense and emotional the workers are, the more
likely they are to commit intentional violations, leading to unsafe
accidents (Radenhausen and Anker, 1988; Golparvar, 2016).
Research on emotions suggests that there are diverse eects
of emotions on individuals’ behavior. eoretical controversies
over the emotion maintenance hypothesis and the aective
generalization theory remain. Given the specicities of
construction task and the construction worker population, this
paper examines the eects of emotions on construction workers’
recognition of safety hazards from a safety management
perspective. A wearable device (HKR-11C+) was used to collect
physiological signals for emotion classication, which is more
objective compared with subjective questionnaire traditionally
used in previous studies. In addition, this study achieves a
quantitative analysis between emotions and individual behavior
through the quantication of emotional valence. is research
is helpful for construction workers to regulate their self-safety
behaviors from an individual psychological perspective, as well
as provide theoretical safety management strategies that focus
on individual psychology for construction companies.
MATERIALS AND METHODS
e famous James Lange’s theory divides emotions into two
basic dimensions that are emotional arousal and emotional
valence (Lang, 1995). Emotional arousal refers to the level of
activation of an individual’s emotion in response to a stimulus
from passive to active, while emotional valence describes the
level of pleasant or unpleasant experience from negative to
positive. Emotions in this study were classied into positive,
neutral and negative emotions by emotional valence. e
construction workers’ recognition ability of safety hazards is
measured in three aspects including the reaction time to safety
hazards, identication accuracy of safety hazards, and the
perception level of safety hazards.
Participants
irty students from Shanghai University majoring in
construction engineering management were selected for the
pilot test to conrm the feasibility and validity of the experiment.
Forty construction workers from six Shanghai construction
engineering enterprises were recruited for the formal experiment.
Among all the subjects, 3 are construction workers and the
other 34 were workers in supervisory positions. ere are 3
subjects were below undergraduate level, 27 were undergraduates
and 10 were postgraduates. All subjects had received safety
management training of construction work. e studies involving
human participants were reviewed and approved by Ethics
Committee of Shanghai University. Participants selected for
the pilot and formal experiment were based on the following
criteria: (1) Familiar with the construction industry or have
long-term working experience on construction site, and familiar
with the operation codes on the construction site; (2) Physically
and mentally healthy without any psychological disorders; (3)
All are right-handed; (4) All provided written informed consent.
e basic information of the subjects are shown in Tabl e 1.
Procedures
e experiment was carried out in a closed construction site
conference room without interference. Forty rounds of
experiments were included in this study. Each round contains
three parts that are positive emotions, neutral emotions, and
negative emotions. ere are 120 samples in total. Aer informed
consent, all the participants were attached electrodes for the
physiological measurement. e experiment guidance is displayed
by computer, which explains the purpose and procedures of
the experiment. Following the instruction, a picture of the
targeted emotion will be shown on the screen to stimulate
the empathic eect of the participant, and the picture will
TABLE1 | Basic information of the subjects.
Category Number
Gender Male 37
Female 0
Years of working
experience
1–9 years 12
10–19 years 13
20 year above 12
Job type Operative workers 10
Supervisory workers 27
Education background Below undergraduate 6
Undergraduate 21
Postgraduate 10

Chong et al. Emotional States Impact Recognition Ability
Frontiers in Psychology | www.frontiersin.org 4 June 2022 | Volume 13 | Article 895929
TABLE2 | The emotional state after emotion arousal.
1 2 3 4 5 6 7 8 9
Very Negative Negative Neutral Positive Very Positive
last for 6 s, 15 pictures will be shown each time. Subsequently,
participants completed the Positive and Negative Aect Schedule
(PANAS) (Watson et al., 1988), followed by the identication
and assessments of safety hazards in the construction pictures.
e experimental procedure is shown in Figure 1.
Individual Emotional Arousal
Arousal of individual emotions using picture stimuli is one
common forms of emotional stimulation (Gerdes et al., 2014).
e International Aective Picture System (IAPS) database (Lang
et al., 1988) was used to evoke dierent emotional states of
the construction workers. irty images each of positive, neutral
and negative pictures were selected from the IAPS as emotional
arousal stimulus material. Researchers have found that incidental
emotions pervasively carry over from one situation to the next,
aecting decisions that unrelated to that emotion, known as
the carryover of incidental emotion (Loewenstein and Lerner,
2003; Lerner and Tiedens, 2006; Keltner and Lerner, 2010).
In the IAPS database, values of valence indicating the level of
enjoyment and the values of arousal indicating the level of
excitement. e positive pictures selected in this paper include
life scenes, animal activity pictures and baby pictures. Neutral
pictures include pictures of static objects, abstract artwork and
pictures of natural environment. Negative pictures include
catastrophic events, violent and brutal scenes and pictures of
disabled individuals. Images of the IAPS database cannot
be displayed as a result of a condentiality agreement. e
valence value of the negative neutral and positive mood pictures
were 1.78, 4.92, 7.83, and the arousal value was 6.36, 3.37,
5.14, respectively. e mean squared deviation of the pictures
was less than 2.4, ensuring the validity of the pictures (Zhang
et al., 2018). e subjects were randomly shown one set of
emotional pictures, and level of emotional arousal was evaluated
using the PANAS Emotional Self-Rating Scale; subjects whose
emotions were not aroused were excluded from the results of
the experiment, the emotional scale is shown in Tab le 2 .
FIGURE1 | Procedures of the experiment.

Chong et al. Emotional States Impact Recognition Ability
Frontiers in Psychology | www.frontiersin.org 5 June 2022 | Volume 13 | Article 895929
Aer the 15 emotion pictures were displayed, the participants
will give a self-evaluation of their emotions by ling an emotion
scale adapted from PANAS, with a scale ranging from 1 to
9. e higher scores indicate stronger positive emotions, while
the lower scores indicate stronger negative emotions. e
emotional scale is shown in Tab le 3 .
Galvanic Skin Response Measurement
Aer cleaning the skin surface, the participants were attached
the sensor to the construction worker’s nger. e electrodes
were stick in the sensor around the index and middle nger.
e galvanic skin response data were recorded and sent to a
computer. e picture of the GSR equipment and the electrode
attachment location is shown in Figure 2. e galvanic skin
response at 5–6 s aer the emotional arousal was used for
identifying and classifying of the emotions. e experiment
was carried out in a laboratory at a room temperature of
22°C, the subjects sitting still in front of a computer for the
galvanic skin response measurement, with the temperature and
humidity remaining constant throughout the experiment. e
signal processing procedures are elaborated in Section 3.
The Measurement of the Recognition
Ability of Safety Hazards
e recognition of safety hazards was measured by identifying
the safety hazards from construction site pictures, and images
of construction sites containing ve types of safety hazards
were collected for this study from 12 construction sites in
Shanghai, China. ese images were obtained from safety
incident reports, and the opinions of 10 experts were collected
to evaluate the two dimensions of the selected images, (1)
whether the images visually represented the safety hazards of
that type of construction and (2) whether the images were
prevalent in the construction site. e 120 images were retained
aer deleting the lower scoring images. e images were
displayed randomly according to category, and 16 images were
presented on screen in a set for the subjects to evaluate the
safety hazards. e distribution of the 16 images by category
was: 7 fall from height, 2 electric shock and re, 2 object
strikes, 1 collapse hazard, 1 mechanical injury and 3 no hazard
images. To avoid a learning eect in the subjects, each picture
was presented only once at random, with no repeated
presentations in all sets for one participants. Some examples
of the images chosen are shown in Figure 3.
e cognitive level of safety hazards was measured by
behavioral experiment. e participants were requested to view
the construction site images and determine if the pictures
contain safety hazards (Tixier et al., 2014). e participants
pressed “1” on the keyboard if they think there are safety
hazards, while “0” on the keyboard if there are not. e
computer automatically records the time taken to identify safety
hazards and the accuracy of the safety hazards assessment.
e perception level of safety hazards of construction workers
was measured by safety hazards perception assessment form,
which was completed simultaneously when workers believe
there is a safety hazards in the given picture. Hallowell pioneered
the use of this form by quantifying safety hazards perception
as the product of the expected frequency and severity of injury
(Hallowell, 2010). e corresponding scores of perception level
of safety hazards are shown in Tabl e 3.
Statistical Analysis
Forty subjects participated in this study, excluding three who
failed in emotional arousal with 37 remaining subjects.
TABLE3 | Expected frequency and severity of safety hazard.
Severity/Frequency Very
common Common Uncommon Very
uncommon
Negligible 0.19 0.04 0.00375 0.000375
Emergency aid 1.13 0.27 0.0226 0.00226
Seek medical advice 3.2 0.77 0.064 0.0064
Hospitalization 6.4 1.53 0.128 0.0128
Permanent
disablement or fatality
340.48 81.55 6.81 0.681
No risk 0 0 0 0
FIGURE2 | GSR equipment and attachment location.

Chong et al. Emotional States Impact Recognition Ability
Frontiers in Psychology | www.frontiersin.org 6 June 2022 | Volume 13 | Article 895929
FIGURE4 | Diagram of the ltering process results.
Aer 37 sets of experiments, each containing 3 categories of
emotional stimuli, 16 pictures of safety hazards for each category,
a total of 1776 data were collected. Finally, 1,650 valid data were
obtained with a 92.9% validity aer removing invalid questionnaires.
e collected data were smoothed and ltered with a median
lter and a third-order Butterworth low-pass lter. A cut-o
frequency of 0.3 HZ was used to lter out abnormal signals,
and the signals were eliminated from the baseline interference
to reduce the inuence caused by the measurement instrument
itself and the current and voltage. e ltered galvanic skin
response signals were extracted and normalized from both
time domain signal features and descriptive features, and then,
principal component analysis is applied to reduce the
dimensionality of the acquired features to obtain the signal
features for classication.
e reaction time to safety hazards, accuracy of safety
hazards identication and perception level of safety hazards
of construction workers in dierent emotional states were
analyzed. One-way ANOVA was used to investigate the dierences
in recognition ability of safety hazards among workers within
dierent age groups under dierent emotional states. Pearson
correlation and regression analyses were used to quantify the
eects of emotions on the reaction time to safety hazards, the
accuracy of safety hazards identication, and perception level
of safety hazards, respectively.
RESULTS AND DISCUSSION
Classication of Galvanic Skin Response
Signals
e collected data were smoothed and ltered with a median
lter and a third-order Butterworth low-pass lter, and the
cut-o frequency was set as 0.3 HZ to lter out the abnormal
signals and retain the valid signals. e original waveform,
the waveform aer median ltering and the waveform aer
low-pass ltering are shown in Figure 4; the obtained signals
were de-baselined to reduce the eects caused by the measurement
instrument itself and the current and voltage; the ltered skin
electrical signals were extracted from both time domain signal
features and descriptive features. e ltered electrical skin
signal was extracted from both time domain signal features
and descriptive features, and the extracted features are shown
in Tab le 4 . Principal component analysis was used to reduce
dimension, the results are shown in Ta bl e 5 , and the coecient
of the feature GSR_di_Std is low, so the feature is deleted
from the subsequent classication training. e nal 10 features
FIGURE3 | Examples of safety hazards pictures.

Chong et al. Emotional States Impact Recognition Ability
Frontiers in Psychology | www.frontiersin.org 7 June 2022 | Volume 13 | Article 895929
retained aer normalization were obtained for the subsequent
Support Vector Machine classication.
e 120 sample points were used as the training data set;
the Support Vector Machine (SVM) model was applied for
classication training. In a ratio of 2:1, these 120 samples are
divided into a training set (90 samples containing 30 positive,
30 neutral, and 30 negative samples) and a test set (30 samples).
Both classication accuracy and model validation were improved
by classied labelling of these 120 sample points and supervised
learning of the model, which was implemented by Matlab
R2016b. In the prediction experiment, 30 sample were validated
and the results are shown in Figure 5; Tabl e 6. e vertical
coordinates 1,2,3 correspond to negative, neutral and positive
samples, respectively. When calculating the sensitive, specicity
and precision of the examples, for each emotion, the emotion
itself is considered a positive example, while the remaining
two emotional states are considered negative examples. e
classication results indicate that the picture triggering method
in this study achieves eective emotional arousal.
Recognition Ability of Safety Hazards of
Construction Workers in Different
Emotional States
e statistical results of recognition ability of safety hazards of
construction workers in dierent emotional states are shown in
Tabl e 7 . Workers in the positive emotional state had the shortest
reaction time (5.61 s), and workers in the negative and neutral
emotional states required longer reaction time (8.08 s and 6.91 s,
respectively). Construction workers in the neutral emotional state
had the highest identication accuracy of safety hazards (92.25%)
and perception level of safety hazard (24.52), while in the negative
emotional state, the construction workers had the lowest
identication accuracy of safety hazards (75.41%) and perception
level of safety hazards (0.75) than other emotional states.
e study found that the reaction time for identifying safety
hazards was longer in negative emotions than in neutral and
positive emotions. When construction workers were in a positive
emotion, the feedback time for identifying safety hazards in
construction site pictures was 5.61 s, which was less than the
6.91 s in a neutral emotion and 8.08 s in a positive emotion.
is conclusion is consistent with Fredrickson’s ndings that
positive emotions serve to improve individual’s physical, intellectual,
and perceptions (Fredrickson, 1998). Although the reaction time
to safety hazards became shorter, the accuracy identication of
safety hazards decreased when the emotional valence increased
from negative to neutral. e identication accuracy of safety
hazards was 92.25% under neutral emotional state, which was
greater than the negative emotional state (75.41%) and the positive
emotional state (80.60%). e ndings are similar to the ndings
on the eect of emotion on driver performance, with drivers
performing better in a neutral state (Jeon et al., 2014) hazards.
The Effect of Emotions on the Recognition
Ability of Safety Hazards of Different
Working Age Groups
Given that the construction workers’ emotions may beinuenced
by the age (Kappes et al., 2017), age was chosen as the
independent variable and a one-way ANOVA was used to
explore the eect of emotions on construction workers’
recognition ability of safety hazards at dierent working ages.
Reaction Time
e results of the one-way ANOVA for the reaction time to
safety hazards at dierent working ages are shown in Table  8.
e signicant dierences between the dierent age groups indicate
that age has a signicant eect on the reaction time to safety
hazards for construction workers. Correlation-type eect size eta
square (Ferguson, 2009) and power reect that age has a small
estimates eect size on the reaction time of construction worker.
e reaction time to safety hazards under dierent working
ages is shown in Figure6. In all three emotional states, construction
workers with more than 20 years of experience had longest reaction
time. Workers with more than 19 years of working experience
had longer reaction time than the 10–19 years working experience
group by more than 0.5 s, and more than 0.8 s than those with
less than 10 years. is suggests that as construction workers
increase in years of experience, the reaction time to recognize
safety hazards increases, while inuence of emotions on the reaction
time to safety hazards decreases. is experiment required computer
operation and construction workers with more than 20 years of
experience, who were generally older and less skilled at operating
the devices may have contributed to longer reaction time.
TABLE4 | Selected variables.
No. Selected Variables Codename
1 Mean value of GSR GSR_Mean
2 Standard deviation of GSR GSR_Std
3 Minimum value of GSR GSR_Min
4 Maximum value of GSR GSR_Max
5 First difference mean value of GSR GSR_diff_Mean
6 First difference standard deviation of GSR GSR_diff_Std
7 First difference minimum value of GSR GSR_diff_Min
8 First difference maximum value of the of GSR GSR_diff_Max
9 Average magnitude of GSR GSR_AveMag
10 Average rise time of GSR GSR_AveRt
11 Average energy of GSR GSR_AveE
TABLE5 | Results of principal component analysis.
1 2 3 4 5
GSR_diff_Min 0.829 0.452 −0.254 0.196 0.021
GSR_diff_Mean 0.826 0.457 −0.251 0.200 0.014
GSR_diff_Max 0.769 0.526 −0.275 0.223 0.017
GSR_Min 0.767 −0.552 0.293 −0.124 0.069
GSR_Max 0.767 −0.551 0.293 −00.124 0.070
GSR_Mean 0.766 −0.552 0.293 −0.124 0.069
GSR_AveRt 0.084 0.711 0.569 −0.343 −0.074
GSR_AveE 0.060 0.701 0.675 −0.126 −0.162
GSR_AveMag −0.063 0.087 0.493 0.732 −0.209
GSR_diff_Std 0.152 0.522 −0.251 −0.547 0.090
GSR_Std −0.243 0.261 0.263 0.197 0.875

Chong et al. Emotional States Impact Recognition Ability
Frontiers in Psychology | www.frontiersin.org 8 June 2022 | Volume 13 | Article 895929
TABLE6 | Support vector machine classication simulation training results.
Category Sensitive Specicity Precision Accuracy f1
Positive 77.8% 90.5% 70% – 0.74
Neural 90% 85% 75% – 0.82
Negative 90.9% 84.2% 76.9% – 0.83
Integral – – – 86.7% 0.80
TABLE7 | Safety hazard cognition results in different emotional states.
Emotional State Reaction time
(s)
Accuracy
(%)
Safety hazard
perception
Negative 8.08 75.41 0.75
Neural 6.91 92.25 24.52
Positive 5.61 80.60 3.10
Identication Accuracy
The results of the one-way ANOVA for the identification
accuracy of safety hazards at different working ages are
shown in Tab le 9 . The significant differences between the
different age groups indicate that age has a significant effect
on the identification accuracy of safety hazards for
construction workers. Correlation-type effect size eta square
(Ferguson, 2009) and power reflect that age has a small
estimates effect size on the identification accuracy of
construction worker.
e identication accuracy of safety hazards under dierent
working ages is shown in Figure 7. e graph illustrates
that the construction workers with more than 20 years of
working experience have a higher accuracy in safety hazards
identication under positive and negative emotional states
than construction workers in the other two age groups. While
under neutral emotions, workers with more than 20 years of
working experience and the group with 10–19 years have
similar accuracy in evaluate safety hazards, at 93.73 and
93.58%, respectively, which are both at a high level. e
above results indicate that the accuracy of safety hazards
identication of experienced workers with more than 20 years
of experience is less aected by their emotional state, which
means that work experience can eectively reduce the impact
of emotional uctuations on the accuracy of safety
hazards identication.
FIGURE5 | SVM classication prediction result.

Chong et al. Emotional States Impact Recognition Ability
Frontiers in Psychology | www.frontiersin.org 9 June 2022 | Volume 13 | Article 895929
TABLE8 | One-way ANOVA results for reaction time at different working ages.
Emotional
states Working years Mean value Standard
deviation Minimum Maximum FSig η2Power Multiple
Comparisons
Negative 0–9 7.74 1.384 4.87 10.50 9.843 0.000 0.127 0.096 1 < 3
10–19 8.02 1.675 3.35 12.72
20 and above 8.50 1.564 5.06 15.50 2 < 3
Neural 0–9 6.35 1.779 1.43 9.75 21.203 0.000 0.136 0.103 1 < 2
10–19 6.98 1.523 2.85 10.12 1 < 3
20 and above 7.49 1.617 5.05 18.50 2 < 3
Positive 0–9 4.42 1.931 0.55 8.95 70.670 0.000 0.096 0.076 1 < 2
10–19 5.71 1.321 1.03 8.47 1 < 3
20 and above 6.44 1.052 3.60 9.33 2 < 3
1 represents the 0–9 working years group, 2 represents the 10–19 years working group and 3 represents the 20 and above years working group.
FIGURE6 | Reaction time to safety hazard under different working ages.
TABLE9 | One-way ANOVA results for identication accuracy at different working ages.
Emotional
states Working years Mean value Standard
deviation Minimum Maximum FSig, η2Power Multiple
Comparisons
Negative 0–9 0.74 0.459 0.510 0.902 4.663 0.010 0.096 0.076 1 < 3
10–19 0.81 0.446 0.603 0.895
20 and above 0.85 0.370 0.537 0.916
Neural 0–9 0.90 0.242 0.645 0.923 3.978 0.042 0.082 0.069 1 < 3
10–19 0.94 0.149 0.668 0.952
20 and above 0.94 0.236 0.636 0.914 2 < 3
Positive 0–9 0.71 0.439 0.554 0.886 6.166 0.002 0.063 1 < 3
10–19 0.73 0.395 0.527 0.893 0.061
20 and above 0.86 0.351 0.611 0.921
1 represents the 0–9 working years group, 2 represents the 10–19 years working group and 3 represents the 20 and above years working group.

Chong et al. Emotional States Impact Recognition Ability
Frontiers in Psychology | www.frontiersin.org 10 June 2022 | Volume 13 | Article 895929
FIGURE7 | Identication accuracy of safety hazard under different working ages.
Safety Hazards Perception
e results of the one-way ANOVA for the perception level
of safety hazards at dierent working ages are shown in Table10.
e signicant dierences between the dierent age groups
indicate that age has a signicant eect on the perception
level of safety hazards for construction workers. Correlation-
type eect size eta square (Ferguson, 2009) and power reect
that age has a small estimates eect size on the safety hazards
perception of construction worker.
e perception level of safety hazards under dierent working
ages is shown in Figure8. As illustrated in the gure, the highest
level of safety hazards perception was found in the group of
workers aged 10–19 years old in the neutral mood state, at 40.24,
and the group of workers aged 10–19 years old had a higher
level of safety hazards recognition in all emotional states. In all
age groups, the level of perceived safety hazards increased from
low to high then decreased as construction workers changed
from extreme negative to extreme positive emotions.
TABLE10 | One-way ANOVA results for perception level at different working ages.
Emotional
states
Working
years Mean value Standard
deviation Minimum Maximum FSig η2Power Multiple
Comparisons
Negative 0–9 2.7177 12.7158 0.0037 81.5500 3.004 0.041 0.028 0.052 1 < 2
10–19 7.1499 20.2831 0.0640 81.5500
20 and
above
4.1766 16.6419 0.0128 81.5500
Neural 0–9 13.6066 25.9091 0.1280 81.5500 10.945 0.000 0.023 0.051 1 < 2
10–19 40.2416 38.2904 0.2700 81.5500 1 < 3
20 and
above
20.3367 30.7954 0.2700 81.5500 2 > 3
Positive 0–9 0.57601 1.4384 0.0000 6.8100 33.901 0.000 0.027 0.052 1 < 2
10–19 6.7472 21.2234 0.0003 81.5500
20 and
above
1.1089 2.16974 0.0003 6.8100 2 > 3
1 represents the 0–9 working years group, 2 represents the 10–19 years working group and 3 represents the 20 and above years working group.

Chong et al. Emotional States Impact Recognition Ability
Frontiers in Psychology | www.frontiersin.org 11 June 2022 | Volume 13 | Article 895929
Quantitative Relationship Between
Emotional Valence and Recognition Ability
of Safety Hazards
According to the results of safety hazards reaction time,
identication accuracy of safety hazards and perception level
of safety hazards, correlation and regression analysis were used
to explore the quantitative relationship between emotional
valence and construction workers’ recognition ability of
safety hazards.
Emotional Valence and Reaction Time
e Pearson correlation coecient (r = −0.556, p = 0.000) indicates
a moderate negative correlation between emotional state and
reaction time to safety hazards. e regression model passed
the F-test (p = 0.000), and the emotional valence explains 30.9%
of the workers’ reaction time to safety hazards (R2 = 0.309,
SE = 0.832, F = 664.413).
e regression results are shown in Table 11, and the
quantitative relationship between safety hazards reaction time
and emotional valence of construction workers is shown in
Equation (1).
RT
EE
V
=− −− +
9 952 50556
..
ε
(1)
Where
RT is safety hazards reaction time,
EV is emotional valence,
ε is error term, which indicates unexplained variability of
the data.
e coecient of emotional state and reaction time to
safety hazards (β1 = −0.556) is less than zero, indicating that
the reaction time to safety hazards decreases as the emotional
state increases, as shown in Figure 9. Either in positive or
negative emotional state, the reaction time to safety hazards
FIGURE8 | Perception level of safety hazard under different working ages.
TABLE11 | Coefcients between reaction time and emotional valence.
Unstandardized coefcients Standardized coefcients
tSig.
BStandard error Beta
Constant(β0)−9.952E-5 0.022 −0.005 0.996
Emotional valence(β1)−0.556 0.022 −0.556 −25.776 0.000

Chong et al. Emotional States Impact Recognition Ability
Frontiers in Psychology | www.frontiersin.org 12 June 2022 | Volume 13 | Article 895929
FIGURE9 | The effect of emotional valence on the reaction time to safety
hazard.
increased by 0.556 units for each unit decrease in the
construction worker’s emotional valence. e reaction time
to safety hazards decreases continuously as the emotional state
of construction workers change from extreme negative to
extreme positive.
Emotional Valence and Identication Accuracy
Descriptive statistics found that the identication accuracy of
safety hazards was higher for construction workers in neutral
emotions than in both positive and negative emotions. e
Pearson correlation coecient shows a moderate negative
correlation between positive emotion and identication accuracy
of safety hazards (r = −0.526, p = 0.000), while negative emotion
is negatively correlated with identication accuracy of safety
hazards (r = −0.356, p = 0.000).
e correlation analysis shows that the relationship between
the emotional state and the identication accuracy of safety
hazards is an inverted U-shape, so curvilinear regression model
is established to show a quadratic expression. e regression
model passed the F-test (p = 0.000) and the emotional valence
explains 21.3% of the workers’ identication accuracy of safety
hazards (R2 = 0.213, SE = 0.834, F = 201.79). e regression results
are shown in Table 12, and quantitative relationship between
identication accuracy of safety hazards and emotional valence
of construction workers is shown in Equation (2).
IA EV EV=+
−+
0510021 0 443 2
.. .
ε
(2)
Where
IA is identication accuracy of safety hazards,
EV is construction worker’s emotional valence,
ε
is error term, which indicates unexplained variability of
the data.
e coecient of the quadratic term (
α
2
= − 0.443) b etween
the emotional state and the accuracy of safety hazards
identication is negative, indicating that the accuracy of safety
hazards identication increases and then decreases as the
emotional valence increases, reaching the highest point when
the emotional state of workers is in neutral emotion as shown
in Figure 10. Specically, when the construction workers are
in positive emotional state, the identication accuracy of safety
hazards decreases by 1.350 units for each unit increase in
emotional state. While the construction workers are in a negative
emotional state, the identication accuracy of safety hazards
decreases by 1.308 units for each unit decrease in their
emotional states.
Emotional Valence and Perception Level of
Safety Hazards
e correlations between the positive emotion, negative emotion
and the perception level of safety hazards were explored,
TABLE12 | Coefcients between identication accuracy and emotional valence.
Unstandardized coefcients Standardized coefcients
tSig.
BStandard error Beta
Constant(
0
α
) 0.51 0.031 16.380 0.000
Emotional state(
1
α
) 0.021 0.022 0.023 0.981 0.327
Emotional state(
2
α
)−0.443 0.022 −0.459 19.769 0.000
FIGURE10 | The effect of emotional valence on the identication accuracy.

Chong et al. Emotional States Impact Recognition Ability
Frontiers in Psychology | www.frontiersin.org 13 June 2022 | Volume 13 | Article 895929
respectively, by Pearson correlation coecient. e results
indicate a low negative correlation between positive emotion
and perception level of safety hazards (r = −0.256, p = 0.000),
while negative emotion is moderately negative correlated with
the perception level of safety hazards (r = −0.520, p = 0.000).
e correlation analysis shows that the relationship between
the emotional state and perception level of safety hazards is
an inverted U-shape, so curvilinear regression is used to establish
a quadratic expression for regression analysis. e regression
model passed the F-test (p = 0.000), and the emotional valence
explains 35.7% of the workers’ perception level of safety hazards
(R2 = 0.357, SE = 0.802, F = 365.999). e regression results are
shown in Table 13, and quantitative relationship between
perception level of safety hazards and emotional valence of
construction workers is shown in Equation (3).
2
PL 0.683 0.079EV 0.617EV
= − − +ε
(3)
Where
PL is perception level of safety hazards,
EV is construction worker’s emotional valence,
ε
is error term, which indicates unexplained variability of
the data.
e coecient of the quadratic term(
γ
2
= − 0.617) b etween
the emotional state and the perception level of safety hazards is
negative, indicating that the perception level of safety hazards
increases and then decreases as the emotional valence increases,
reaching the highest point when the emotional state of workers
is in neutral emotion as shown in Figure 11. Specically, when
construction workers are in positive emotion, the perception level
of safety hazards decreases by 1.93 units for every each increase
in their emotional state. When construction workers are in negative
emotion, the perception level of safety hazards decreases by
1.772 units for each unit decrease in their emotional state. erefore,
the perception level of safety hazards of construction workers
in negative and positive emotions is lower than that in neutral
emotions. As construction workers’ emotions change from extremely
negative to extremely positive, the perception level of safety
hazards changes from low to high and then lower.
The Safest Emotion
ere was a negative correlation between reaction time to safety
hazards and emotional valence, while the accuracy of safety
hazards identication and the perception level of safety hazards
had an inverted “U” shape relationship with emotional valence.
When workers are under positive emotional valence, the ndings
are consistent with the aective generalization theory (Johnson
and Tversky, 1983), where positive emotion drive construction
workers to make optimistic judgements about the construction
environment, and therefore lower levels of perception of safety
hazards. When under the negative emotional valence, the ndings
are more in line with the Mood Maintenance Hypothesis (Isen
and Patrick, 1983), where construction workers tend to take more
risky and aggressive decisions in order to escape from their current
negative emotional state, thus underestimating the risks of the
environment and lowering the level of safety hazards perceived
by workers. In addition, the ndings of this study are support
the Aective states as information hypothesis proposed by Schwarz
and Clore (1981). is theory suggests that emotions simplify
people’s risk decision-making process, which people judge things
based on their feelings rather than their features, and that emotions
can lead to overestimation of events of the same valence. is
nding is related to information acquisition and cognitive processes,
and subsequent research could be further explored from this
perspective. Workers should avoid overexcited emotional states,
for each unit increase in emotional valence, the reaction time
to safety hazards reduced by 0.556 units. Meanwhile the identication
accuracy of safety hazards reduced by 1.35 units, and the level
of safety hazards perception reduced by 1.93 units when the
TABLE13 | Coefcients between perception level of safety hazard and emotional valence.
Unstandardized coefcients Standardized coefcients
tSig.
BStandard error Beta
Constant(
0
γ
) 0.683 0.034 20.354 0.000
Emotional state(
1
γ
)−0.079 0.021 −0.084 −3.747 0.000
Emotional state(
2
γ
)−0.617 0.023 −0.604 −27.055 0.000
FIGURE11 | The effect of emotional valence on the perception level of
safety hazard.

Chong et al. Emotional States Impact Recognition Ability
Frontiers in Psychology | www.frontiersin.org 14 June 2022 | Volume 13 | Article 895929
emotional valence shi by one unit from neutral emotional state.
Workers with high emotional valence have a more relaxed and
pleasurable state, with increased reaction speed but reduced ability
to judge and perceive safety hazards due to inattentiveness.
Construction workers also need to avoid negative emotions such
as excessive sadness and grief, for each unit decrease in emotional
state, the feedback time for safety hazards recognition increases
by 0.556 units. Meanwhile the identication accuracy of safety
hazards decreases by 1.308 units and the perception level of safety
hazards decreases by 1.772 units. When workers are immersed
in a state of loss and frustration, the attention allocated to safety
hazards identication decreases, prolonging their own judgment
time, with a concomitant decrease in the accuracy of safety hazards
identication and the level of safety hazards perception. erefore,
neutral emotions are the safest emotions.
CONCLUSION
Behavioral experiment revealed that the support vector machine
(SVM) algorithm was eective in classifying galvanic skin
response signals to identify emotional states. e reaction time
to recognition ability of safety hazards of construction workers
under negative emotion is longer than neutral and positive
emotions, and the identication accuracy of safety hazards
and the perception level of safety hazards are lower, so the
general recognition ability of safety hazards of construction
workers under negative emotion is poorer. e reaction time
to safety hazards identication is shorter for construction
workers in positive emotions, but the accuracy of safety hazards
identication and the level of safety hazards perception are
lower, and the accuracy of safety hazards identication and
the level of safety hazards perception are higher for construction
workers in neutral emotions than in negative and positive
emotions. For construction workers with more than 20 years
of experience, work experience can eectively reduce the impact
of emotional uctuations on the accuracy of safety hazards
evaluation. Emotion predicted the recognition ability of safety
hazards of construction workers, with a moderate negative
correlation between reaction time to safety hazards and emotional
valence, and a low relationship between accuracy of safety
hazards identication and perception level of safety hazards
and emotional valence shaped an inverted “U.” Compared to
positive emotions and negative emotions, construction workers
in neutral emotions have the highest level of accuracy of safety
hazards identication and perception of safety hazards, making
neutral emotions deemed to be the safest emotion.
e complexity and dynamics of the construction site require
workers to identify the safety hazards present on the site timely
and accurately, and keeping their emotional state stable is
benecial to improving construction workers’ ability to identify
safety hazards and keep themselves safe. Currently, China’s
construction workers are generally poorly educated, lack
continuous psychological training and have weak emotional
control, while safety training in construction companies tend
to focus on the operational specications, unsafe behaviors,
the requirements for wearing safety gear and the main
prohibitions of safe production, and rarely include emotional
management and requirements in safety education and training.
Construction companies should pay more attention to the
emotional health of construction workers and keep their
emotional state stable through psychological training, to improve
workers’ awareness of their emotion and emergency handling
ability, therefore to reduce the probability of safety accidents
and improve the safety management of construction sites.
ere are some limitations to the present study which may
be relevant for future research. First, the study used galvanic
skin response to monitor emotions, while scientic and
technological advances have led to increasingly sophisticated
techniques for monitoring physiological signals. Some studies
focus on collecting EEG signals through EEG devices (Takehara
etal., 2020; Long etal., 2021), exploring brain activity in dierent
emotional states. Eye-tracking devices were used to collect
construction workers’ eye-movement signals, and analyses how
eye-movement signals reect the emotions of construction workers
(Soleymani et al., 2012). Future research could investigate the
impact of emotions on an individual’s physiological signals, as
well as cognitive abilities, in a multidimensional approach through
the applications of novel devices. Second, the participants in this
study were mainly construction workers, and this paper explored
the eect of work experience on construction workers’ emotions.
Future research could further focus on the variability within
groups of construction workers, such as dierent personality traits
(Sugi et al., 2020; Maier et al., 2021), dierent populations
(Gutiérrez-Cobo et al., 2017) or gender (Lischke et al., 2020),
to explore dierences in the inuence of emotions between groups
of construction workers. At last, the study categorized emotions
into the three most basic types, negative, neutral and positive
by emotional valence. In fact, there are more varieties of emotions
and even research paradigms, research on dierent negative
emotions has gained extensive attention (Pittig et al., 2014;
Topolinski and Strack, 2015), and subsequent research can
beconducted from these perspectives and bestudied in more detail.
DATA AVAILABILITY STATEMENT
e raw data supporting the conclusions of this article will
be made available by the authors, without undue reservation.
ETHICS STATEMENT
e studies involving human participants were reviewed and
approved by the Ethics Committee of Shanghai University.
AUTHOR CONTRIBUTIONS
DC, AY, and HS contributed to conceptualization, writing—
review and editing, formal analysis, methodology, and original
dra. HS and DC contributed to investigation. DC and YZ
contributed to supervision. All authors contributed to the article
and approved the submitted version.

Chong et al. Emotional States Impact Recognition Ability
Frontiers in Psychology | www.frontiersin.org 15 June 2022 | Volume 13 | Article 895929
FUNDING
is study was funded by the National Natural Science Foundation
of China (grant no. 71901139) and Science and Technology
Commission of Shanghai Municipality (grant nos. 19DZ1204203
and 21692195100).
ACKNOWLEDGMENTS
e authors would like to specially thank Shanghai Road &
Bridge (Group) Co., Ltd., Shanghai Construction Group (SCG),
and China State Construction for providing generous help on
data collection.
REFERENCES
Abdelhamid, T. S., and Everett, J. G. (2000). Identifying root causes of
construction accidents. J. Constr. Eng. Manag. 126, 52–60. doi: 10.1061/
(ASCE)0733-9364(2000)126:1(52)
Al-Bayati, A. J., Albert, A., and Ford, G. (2019). Construction safety culture
and climate: satisfying necessity for an industry framework. Pract. Period.
Struct. Des. Constr. 24:04019028. doi: 10.1061/(ASCE)SC.1943-5576.0000452
Alizadeh, S. S., Mortazavi, S. B., and Sepehri, M. M. (2015). Assessment of
accident severity in the construction industry using the Bayesian theorem.
Int. J. Occup. Saf. Ergon. 21, 551–557. doi: 10.1080/10803548.2015.1095546
Chen, S., Zhang, L., Jiang, F., Chen, W., and Chen, H. (2020). Emotion recognition
based on multiple physiological signals. Chin. J. Med. Instrum. 44, 283–287.
doi: 10.3969/j.issn.1671-7104.2020.04.001
China Construction Industry Association (2020). Statistical analysis of the
development of the construction industry in 2020 [EB/OL]. Available at:
https://mp.weixin.qq.com/s/EAO_iMuFO_rJ7ii6MQO1kg
China National Bureau of Statistics (2020). Statistical Bulletin on National
Economic and Social Development of the People’s Republic of China 2020
[EB/OL]. Available at: http://www.stats.gov.cn/ztjc/zthd/lhfw/2021/lh_
hgjj/202103/t20210301_1814216.html
Dougherty, F. E., Bartlett, E. S., and Izard, C. E. (1974). Responses of schizophrenics
to expressions of the fundamental emotions. J. Clin. Psychol. 30, 243–246.
doi: 10.1002/1097-4679(197407)30:3<243::AID-JCLP2270300304>3.0.CO;2-0
Dzedzickis, A., Kaklauskas, A., and Bucinskas, V. (2020). Human emotion recognition:
review of sensors and methods. Sensors 20:592. doi: 10.3390/s20030592
Ekman, P. (1992a). Facial expressions of emotion: an old controversy and new
ndings. Philosophical transactions of the Royal Society of London. Series
B: Biol. Sci. 335, 63–69. doi: 10.1098/rstb.1992.0008
Ekman, P. (1992b). An argument for basic emotions. Cognit. Emot. 6, 169–200.
doi: 10.1080/02699939208411068
Farmer, E., and Chambers, E. G. (1929). A Study of Personal Qualities in
Accident Proneness and Prociency. London: H.M. Stationery O.
Ferguson, C. J. (2009). An eect size primer: a guide for clinicians and researchers.
Prof. Psychol. Res. Pract. 40, 532–538. doi: 10.1037/a0015808
Fredrickson, B. L. (1998). What good are positive emotions? Rev. Gen. Psychol.
2, 300–319. doi: 10.1037/1089-2680.2.3.300
Fredrickson, B. L., and Branigan, C. (2005). Positive emotions broaden the
scope of attention and thought-action repertoires. Cognit. Emot. 19, 313–332.
doi: 10.1080/02699930441000238
Gerdes, A., Wieser, M. J., and Alpers, G. W. (2014). Emotional pictures and
sounds: a review of multimodal interactions of emotion cues in multiple
domains. Front. Psychol. 5:1351. doi: 10.3389/fpsyg.2014.01351
Golparvar, M. (2016). Unconventional functions of deviant behaviors in the
relationship between job stress and emotional exhaustion: three study ndings.
Curr. Psychol. 35, 269–284. doi: 10.1007/s12144-014-9292-8
Gutiérrez-Cobo, M. J., Cabello, R., and Fernández-Berrocal, P. (2017). e
three models of emotional intelligence and performance in a hot and cool
go/no-go task in undergraduate students. Front. Behav. Neurosci. 11:33. doi:
10.3389/fnbeh.2017.00033
Hallowell, M. (2010). Safety risk perception in construction companies in the
Pacic northwest of the USA. Constr. Manag. Econ. 28, 403–413. doi:
10.1080/01446191003587752
Haslam, R. A., Hide, S. A., Gibb, A. G. F., Gyi, D. E., Pavitt, T., Atkinson, S.,
et al. (2005). Contributing factors in construction accidents. Appl. Ergon.
36, 401–415. doi: 10.1016/j.apergo.2004.12.002
Heinrich, H. W. (1941). Industrial Accident Prevention. A Scientic Approach.
Industrial Accident Prevention. A Scientic Approach. (Second Edition).
Isen, A. M., and Patrick, R. (1983). e eect of positive feelings on risk
taking: when the chips are down. Organ. Behav. Hum. Perform. 31, 194–202.
doi: 10.1016/0030-5073(83)90120-4
Jeon, M., Walker, B. N., and Yim, J.-B. (2014). Eects of specic emotions on
subjective judgment, driving performance, and perceived workload. Transport.
Res. F: Trac Psychol. Behav. 24, 197–209. doi: 10.1016/j.trf.2014.04.003
Johnson, E. J., and Tversky, A. (1983). Aect, generalization, and the perception
of risk. J. Pers. Soc. Psychol. 45, 20–31.
Kajiwara, Y., Toshihiko, S., and Haruhiko, K. (2019). Predicting emotion and
engagement of workers in order picking based on behavior and pulse waves
acquired by wearable devices. Sensors 19:165. doi: 10.3390/s19010165
Kappes, C., Streubel, B., Droste, K. L., and Folta-Schoofs, K. (2017). Linking
the positivity eect in attention with aective outcomes: age group dierences
and the role of arousal. Front. Psychol. 8:1877. doi: 10.3389/fpsyg.2017.01877
Keltner, D., and Lerner, J. S. (2010). “Emotion,” in Handbook of Social Psychology.
eds. S. T. Fiske, D. T. Gilbert and G. Lindzey (Hoboken. NJ: John Wiley
& Sons), 317–352.
Kim, J., and André, E. (2008). Emotion recognition based on physiological
changes in music listening. IEEE Trans. Pattern Anal. Mach. Intell. 30,
2067–2083. doi: 10.1109/TPAMI.2008.26
Kim, M.-K., Kim, M., Eunmi, O., and Kim, S.-P. (2013). A review on the
computational methods for emotional state estimation from the human EEG.
Comput. Math. Methods Med. 2013:573734, 1–13. doi: 10.1155/2013/573734
Lang, P. J. (1995). e emotion probe: studies of motivation and attention.
Am. Psychol. 50, 372–385. doi: 10.1037/0003-066X.50.5.372
Lang, Peter J., Bradley, M. M., and Cuthbert, B. (1988). e International
Aective Picture System. Gainesville, USA: Center for Research in
Psychophysiology, University of Florida.
Lerner, J. S., and Tiedens, L. Z. (2006). Portrait of the angry decision maker:
how appraisal tendencies shape anger’s inuence on cognition. J. Behav.
Decis. Mak. 19, 115–137. doi: 10.1002/bdm.515
Lingard, H. (2013). Occupational health and safety in the construction industry.
Constr. Manag. Econ. 31, 505–514. doi: 10.1080/01446193.2013.816435
Lischke, A., Pahnke, R., Mau-Moeller, A., Jacksteit, R., and Weippert, M. (2020).
Sex-specic relationships between interoceptive accuracy and Emotion
regulation. Front. Behav. Neurosci. 14:67. doi: 10.3389/fnbeh.2020.00067
Liu, D., Jin, Z. Y., and Gambatese, J. A. (2020). Scenarios for integrating IPS-
IMU system with bim technology in construction safety control. Pract.
Period. Struct. Des. Constr. 25:05019007. doi: 10.1061/(ASCE)SC.1943-
5576.0000465
Loewenstein, G., and Lerner, J. S. (2003). e role of aect in decision making.
Prog Brain Res. 202, 37–53. doi: 10.1016/B978-0-444-62604-2.00003-4
Long, F. F. S. G., Zhao, X., Wei, S. C., Ng, X. L., Ni, A. P., Chi, P., et al.
(2021). Positive and negative emotion classication based on multi-channel.
Front. Behav. Neurosci. 15:720451. doi: 10.3389/fnbeh.2021.720451
MacLeod, C., Mathews, A., and Tata, P. (1986). Attentional bias in emotional
disorders. J. Abnorm. Psychol. 95:15, –20. doi: 10.1037/0021-843X.95.1.15
Maier, M. J., Schiel, J. E., Rosenbaum, D., Hautzinger, M., Fallgatter, A. J., and
Ehlis, A.-C. (2021). To regulate or not to regulate: emotion regulation in
participants with low and high impulsivity. Front. Behav. Neurosci. 15:645052.
doi: 10.3389/fnbeh.2021.645052
Manzoor, A., and Treur, J. (2015). An agent-based model for integrated emotion
regulation and contagion in socially aected decision making. Biol. Inspired
Cognit. Archit. 12, 105–120. doi: 10.1016/j.bica.2015.04.005
Ministry of Housing and Urban–Rural Development of China (2019). Circular
of the general oce of the ministry of housing and urban-rural development
on the production and safety accidents of housing and municipal engineering
in 2019 [EB/OL]. Available at: http://www.mohurd.gov.cn/wj/202006/
t20200624_246031.html

Chong et al. Emotional States Impact Recognition Ability
Frontiers in Psychology | www.frontiersin.org 16 June 2022 | Volume 13 | Article 895929
Muller, A. R., Pfarrer, M. D., and Little, L. M. (2014). A theory of collective
empathy in corporate philanthropy decisions. Acad. Manag. Rev. 39, 1–21.
doi: 10.5465/amr.2012.0031
Namian, M., Zuluaga, C. M., and Albert, A. (2016). “Critical factors that impact
construction workers’ hazard recognition performance.” Construction Research
Congress 2016, 2762–2772.
Pandit, B., Albert, A., Patil, Y., and Al-Bayati, A. J. (2019). Impact of safety
climate on hazard recognition and safety risk perception. Saf. Sci. 113,
44–53. doi: 10.1016/j.ssci.2018.11.020
Perlman, A., Sacks, R., and Barak, R. (2014). Hazard recognition and risk
perception in construction. Saf. Sci. 64, 22–31. doi: 10.1016/j.ssci.2013.11.019
Pittig, A., Pawlikowski, M., Craske, M. G., and Alpers, G. W. (2014). Avoidant
decision making in social anxiety: the interaction of angry faces and emotional
responses. Front. Psychol. 5:1050. doi: 10.3389/fpsyg.2014.01050
Pool, E., Brosch, T., Delplanque, S., and Sander, D. (2016). Attentional bias
for positive emotional stimuli: a meta-analytic investigation. Psychol. Bull.
142, 79–106. doi: 10.1037/bul0000026
Rachlin, H. (2003). Bounded rationality: e adaptive toolbox. J. Exp. Anal.
Behav. 79, 409–412. doi: 10.1901/jeab.2003.79-409
Radenhausen, R. A., and Anker, J. M. (1988). Eects of depressed mood
induction on reasoning performance. Percept. Mot. Skills 66, 855–860. doi:
10.2466/pms.1988.66.3.855
Salminen, S., and Tallberg, T. (1996). Human errors in fatal and serious
occupational accidents in Finland. Ergonomics 39, 980–988. doi:
10.1080/00140139608964518
Schwarz, N., and Clore, G. L. (1981). Mood, misattribution, and judgments
of well-being: informative and directive functions of aective states. J. Pers.
Soc. Psychol. 45, 513–523.
Soleymani, M., Pantic, M., and Pun, T. (2012). Multimodal Emotion Recognition
in Response to Videos. IEEE Trans. Aect. Comput. 3, 211–223. doi:
10.1109/T-AFFC.2011.37
Sugi, M., Sakuraba, S., Saito, H., Miyazaki, M., Yoshida, S., Kamada, T., et al.
(2020). Personality traits modulate the impact of emotional stimuli during
a working memory task: a near-infrared spectroscopy study. Front. Behav.
Neurosci. 14:514414. doi: 10.3389/fnbeh.2020.514414
Takehara, H., Shigekazu, I., and Tatsuya, I. (2020). Comparison between facilitating
and suppressing facial emotional expressions using frontal EEG asymmetry.
Front. Behav. Neurosci. 14:554147. doi: 10.3389/fnbeh.2020.554147
Tixier, A. J.-P., Hallowell, M. R., Albert, A., van Boven, L., and Kleiner, B. M.
(2014). Psychological antecedents of risk-taking behavior in construction. J.
Constr. Eng. Manag. 140:04014052. doi: 10.1061/(ASCE)CO.1943-7862.0000894
Topolinski, S., and Strack, F. (2015). Corrugator activity conrms immediate
negative aect in surprise. Front. Psychol. 6:134. doi: 10.3389/fpsyg.2015.00134
Watson, D., Clark, L. A., and Tellegen, A. (1988). Development and validation
of brief measures of positive and negative aect: the PANAS scales. J. Pers.
Soc. Psychol. 54, 1063–1070. doi: 10.1037/0022-3514.54.6.1063
Weiss, H. M., and Cropanzano, R. (1996). Aective events theory: A theoretical
discussion of the aective experiences and job satisfaction and variations
in aective experience over time. December Organ. Behav. Hum 78, 1–24.
Yang, X., Tian, Y., Feng, K., Yang, J., Zhang, S.-h., and Wang, S. (2020). Signal
game analysis on the eectiveness of coal mine safety supervision based
on the aective events theory. Complexity 2020, 1–9. doi: 10.1155/2020/5710419
Zelenski, J. M., Santoro, M. S., and Whelan, D. C. (2012). Would introverts
be better o if they acted more like extraverts? Exploring emotional and
cognitive consequences of counterdispositional behavior. Emotion 12, 290–303.
doi: 10.1037/a0025169
Zhang, X., Chao, X., Xue, W., Jing, H., He, Y., and Gao, M. (2018). Emotion
recognition based on multichannel physiological signals with comprehensive
nonlinear processing. Sensors 18:3886. doi: 10.3390/s18113886
Zhao, B., Wang, Z., Yu, Z., and Guo, B. (2018). “EmotionSense: Emotion Recognition
Based on Wearable Wristband.” in 2018 IEEE SmartWorld, Ubiquitous Intelligence
& Computing, Advanced & Trusted Computing, Scalable Computing &
Communications, Cloud & Big Data Computing, Internet of People and Smart
City Innovation. Oct 8–12, 2018; 2018 IEEE SmartWorld.
Zhou, Q., Fang, D., and Wang, X. (2008). A method to identify strategies for
the improvement of human safety behavior by considering safety climate
and personal experience. Saf. Sci. 46, 1406–1419. doi: 10.1016/j.ssci.2007.10.005
Conict of Interest: YZ is employed by Shanghai Road & Bridge (Group) Co., Ltd.
e remaining authors declare that the research was conducted in the absence
of any commercial or nancial relationships that could beconstrued as a potential
conict of interest.
Publisher’s Note: All claims expressed in this article are solely those of the
authors and do not necessarily represent those of their aliated organizations,
or those of the publisher, the editors and the reviewers. Any product that may
be evaluated in this article, or claim that may be made by its manufacturer, is
not guaranteed or endorsed by the publisher.
Copyright © 2022 Chong, Yu, Su and Zhou. is is an open-access article distributed
under the terms of the Creative Commons Attribution License (CC BY). e use,
distribution or reproduction in other forums is permitted, provided the original
author(s) and the copyright owner(s) are credited and that the original publication
in this journal is cited, in accordance with accepted academic practice. No use,
distribution or reproduction is permitted which does not comply with these terms.