Assessing College Students' Risk Perceptions of Hazards in

Article Cite This: J. Chem. Educ. XXXX, XXX, XXX−XXX

pubs.acs.org/jchemeduc

Assessing College Students’ Risk Perceptions of Hazards in Chemistry Laboratories Clara Rosalía Á lvarez-Chav́ ez,*,† Luz S. Marín,‡ Karla Perez-Gamez,§ Mariona Portell,∥ Luis Velazquez,† and Francisca Munoz-Osuna†

Downloaded via UNIV PARIS-SUD on August 19, 2019 at 23:17:37 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

†

Department of Chemical-Biological Sciences, Sustainable Development Program, University of Sonora, Boulevard Luis Encinas y Rosales S/N, Colonia Centro, 83000 Hermosillo, Sonora, México ‡ Department of Safety Sciences, Indiana University of Pennsylvania, Indiana, Pennsylvania 15705, United States § Sustainable Development Program, University of Sonora, Boulevard Luis Encinas y Rosales S/N, Colonia Centro, 83000 Hermosillo, Sonora, México ∥ Department of Psychobiology and Methodology of Health Sciences, Autonomous University of Barcelona, Campus de Bellaterra, 08193 Cerdanyola del Vallès, Spain S Supporting Information *

ABSTRACT: College laboratories are generally perceived to be low-risk environments in comparison to industrial laboratories and plant operations. However, accidents in college chemistry laboratories have revealed the safety conditions to which both students and staff may be exposed. Improving the effectiveness of laboratory safety training programs and chemical safety education requires gaining an understanding of how undergraduate students may perceive the risk associated with chemistry laboratory settings. This study characterized risk perceptions of safety hazards in chemistry laboratories among college students. Undergraduate college students from the chemistry and biology department of a university in Mexico were surveyed. The Workers’ Risk Perception Dimensional Evaluation (EDRP-T) was used to characterize risk perceptions through nine dimensions and the overall perceived risk for three risk factors: laboratory work, chemical splashes, and chemicalsubstances inhalation. Perceived risk was characterized in a sample of 521 undergraduate students. Students felt confident in successfully dealing with the risk factors evaluated despite feelings of dread and vulnerability as well as concerns about the severity of the consequences of an injury. Their perceived ability to control and avoid these risks might have reflected the students’ self-efficacy. Discrepancies in characterizing risk perception as a multidimensional construct or a direct, measurable characteristic were identified. Gaining an understanding about what undergraduate students do and do not perceive as hazardous is a valuable input to develop risk management and communication strategies with the potential to influence students’ decision-making process that can result in safer behaviors. Successful design and implementation of chemical education programs requires recognizing gaps at all levels. KEYWORDS: First-Year Undergraduate/General, Second-Year Undergraduate, Upper-Division Undergraduate, Chemical Education Research, Curriculum, Laboratory Instruction, Safety/Hazards FEATURE: Chemical Education Research

■

INTRODUCTION College laboratories are generally perceived to be low-risk environments in comparison to industrial laboratories and plant operations.1,2 However, safety in academia is receiving increasing attention.3,4 Accidents in chemistry laboratories, in particular, have revealed the safety conditions to which both students and staff may be exposed. A study conducted at Iowa State University revealed that, from 2001 to 2014, a total of 1497 lab-related accidents were reported and 33.3% of these accidents involved students.5 Stuart and Toreki reported that, from 5381 hazmat events recorded for 3 years (2010−2013), © XXXX American Chemical Society and Division of Chemical Education, Inc.

10% were related to laboratory settings and 4% to an educational facility outside laboratories.6 The U.S. Chemical Safety Board (CSB), which typically investigates chemical accidents to protect workers, has also observed safety gaps in academic laboratories.3,7−9 Between 2001 and 2011, the CSB collected preliminary data on 120 explosions, fires, and chemical releases at university laboratories and other research Received: November 6, 2018 Revised: July 30, 2019

A

DOI: 10.1021/acs.jchemed.8b00891 J. Chem. Educ. XXXX, XXX, XXX−XXX

Journal of Chemical Education

Article

individual characteristics (e.g., age, gender, experience, and education) and social and cultural factors (e.g., culture, habits, political orientation, race, beliefs, and values).21−24 In light of this concept, risk perception may also be influenced by the sense of vulnerability and by a person’s knowledge of and ability to respond to hazards.25 This explains why similar conditions are perceived differently at individual and group levels.26 For example, differences in perception of risk and particularly discrepancies between experts and the general public are common. Judgments of potential risk may derive from an inaccurate evaluation of the prevalence and harmfulness of hazards in the environment27 and may influence the way individuals react to risks. Some studies have identified positive correlations between perceived risk and risky behavior, accidents, and near-misses.23,26,28−30 The psychometric paradigm, which defines risk perception from a multidimensional perspective, is one of the most relevant research approaches to the perceived risk construct.31 According to this approach, risk perception is quantifiable and predictable. This paradigm is suggested as a useful tool for identifying similarities and differences in risk perception among individuals and groups.32 The first psychometric studies investigated nine qualitative dimensions of risk perception:33

facilities that occurred around the U.S. These accidents vary in severity from minor to severe injuries, fatalities, and extensive property damage.10 The CSB identified the need to examine internal safety policies and procedures for research laboratories to ensure that adequate safety systems and an active safety culture are in place. Given that chemical substances must be handled under specific conditions of pressure and temperature to facilitate chemical reactions, the chemical laboratory environment may expose students to a variety of hazards which, along with unsafe processes and behaviors, have the potential to contribute to minor and even to severe accidents.11 The presence of hazards such as chemical substances and experimental conditions are intrinsic to the nature of chemical laboratory work (academic, research, or industrial), and therefore the most effective methods of control (elimination and substitution) might not be pertinent in all cases. Thus, to protect lab-related personnel, controls must also be oriented to reduce risk levels. Then, a conceptual distinction must be made between hazard and risk. From the safety perspective, a hazard is defined as any source with a potential to cause damage, harm, or adverse health effects, while risk refers the combination of the likelihood of the occurrence of harm and the severity of that harm.12 The likelihood can be determined objectively (probability) or subjectively (perceptions). Risk perception is comprised of individuals’ judgments of risk and plays a fundamental role in designing safety interventions.13 Hazard recognition and knowledge to assess the risk associated with an experiment or laboratory operation are essential skills for maintaining safe conditions in laboratory environments.14 Controlling risks in a university chemistry laboratory may be complex, given that, in order to minimize lab-related risk levels, it would additionally be necessary to modify factors related to both work organization and longstanding practices. Factors such as researchers (i.e., principal investigators and graduate and undergraduate students) working alone or without proper supervision, time restrictions and long shifts (day/night), and poor attitudes toward safety may exacerbate risk levels already present in the laboratory environment. As the CSB reported, achieving change in this environment requires improving research-specific hazard recognition and evaluation, developing properly written protocols and chemical safety education, facilitating hazard reporting, and documenting incidents and near-misses.10 Multiple factors influence the presence of hazards, their subsequent recognition, and their prevention and control. The social and organizational environment, along with safety factors and individual characteristics, interact to influence the occurrence of undesired safety events.15−17 Injury prevention requires an understanding of precisely how particular factors may contribute to injuries.18 Accurate identification and perception of risks are fundamental to the success of any safety training program and chemical safety education. When hazards remain unrecognized, or the associated safety risk is underestimated, the likelihood of catastrophic and unexpected injuries, property damage, and experiment failure may increase.19 Risk perception refers to the subjective assessment of the probability that a specified type of event may happen and how concerned subjects are with the consequences.20 Sjoberg et al. referred to risk perception as an individual, social, and cultural construct reflecting values, symbols, history, and the ideology of human social existence. Thus, risk perception is related to

1. 2. 3. 4. 5. 6. 7. 8. 9.

voluntariness of exposure to the source of the risk; immediacy of its effect; control over it; severity of its consequences; knowledge on the part of those who are exposed to the risk; knowledge attributed to the experts; novelty; catastrophic potential; fear generated by the risk.

Subsequent studies introduced modifications to this list of characteristics in order to adapt it to different contexts. Portell and Solé34 excluded voluntariness and novelty and included two new dimensionsavoidance, and perceived vulnerability to riskwhich had already been defined in previous studies of the psychometric paradigm.35,36 The most general conclusion of this type of study is that the aforementioned qualitative characteristics of risk can be condensed into a small number of dimensions, with fear and ignorance being the most important and stable of them.37 The psychometric paradigm states that laypeople judge risk based on multiple attributes beyond the traditional severity and exposure used in risk assessment. Research on risk perception has been focused on identifying the mechanisms through which different risk perceptions may contribute to the injury prevention causal pathway. Adoption of safety behaviors such as wearing personal protection has been associated with perceived risk.38 Risk perceptions can also help to identify gaps in knowledge or attitudes toward risk that must be bridged in order to improve the efficacy of relevant chemical safety education programs such as hazard recognition, training, and staff safety commitment. A previous study in university laboratories found that more than half of the students had experienced at least one incident in the laboratory; some of these incidents had a contributing cause, such as the lack of verification of material and equipment conditions and the failure to wear appropriate personal protective equipment.39 B



■

Article

RESEARCH QUESTIONS In academia, chemistry laboratories support faculty research or are associated with chemistry courses for those students majoring in chemistry or applied science majors. Research chemistry laboratories often involve more specific chemical substances and equipment to conduct experiments as well as graduate students who due to their individual education progress from undergraduate to a graduate level may have a better knowledge of the hazards and risk levels associated with chemistry laboratory experiments. In contrast, the primary users of laboratories associated with chemistry courses are undergraduate students who may be exposed to less risk since, by design, these experiments require smaller volumes of chemical substances and are conducted under more controlled conditions. Providing undergraduate students with the technical skills to function in demanding job environments such as research and industrial laboratories is a priority at the undergraduate level. However, lab chemistry courses are also an opportunity to provide formal safety education that instills safety skills that will be required later during their professional career. To date, no comprehensive studies exist on the factors influencing risk perceptions among college students enrolled in laboratory classes. This study was aimed at characterizing college students’ risk perceptions of safety hazards in chemistry laboratories. In the first phase of this study, the most common risk factors present during laboratory practices were identified and perceived risk was characterized using both multidimensional risk constructs and overall risk perception. Then, the association between risk dimensions and overall risk was examined. From this aim, the research questions are as follows: 1. What are the risk perception dimensions associated with the exposure to chemistry laboratory work among undergraduate students? 2. How do undergraduate students perceive risk when they work in chemistry laboratories?

principal investigator sent a communication via e-mail to faculty members and laboratory technicians explaining the study and the scope of their participation. The Delphi method was conducted in two runs. In the first run, individual face-toface interviews consisting of open-ended questions about safety conditions and the main risk factors in the university chemistry laboratories were conducted with the faculty members. The following is a typical example of the type of question posed: “In your opinion, what are the conditions that most frequently cause or may cause incidents/accidents to students when they are engaged in laboratory tasks?” First-round responses regarding risk factors present in the chemical laboratories were analyzed through a qualitative thematic analysis process in order to list the most common risk factors. Then, a list of risk factors was elaborated on the basis the interviews. To determine the perceived order of importance of the identified risk factors, on the second run, the same group of experts was consulted regarding the probability that an incident or accident could occur due to the risk factors included in the list. For example, they were asked, “How would you rank the probability that an accident/ incident might occur due to the factors included in the list?” The probability was ranked on a four-point Likert scale with the following choices: 1. highly probable; 2. very likely; 3. moderately likely; 4. almost never. The three risk factors that ranked the highest were selected to assess perceived risk among undergraduate students. In general, the Delphi method uses systematic procedures to establish priorities and determine the amount of consensus in the group regarding priorities. In this study, faculty and staff responsible for designing experimental protocols were selected as the group with the knowledge and experience to identify dangerous situations/conditions (hazards) that may have affected students or with the potential to harm students during the laboratory experiments. Due to their background, we relied on this group of experts to identify and prioritize the risk factors that students face during their chemistry lab courses.

■

METHODS This study had two phases. The first phase consisted of a qualitative study involving Mexican university academic staff members who were recruited with the aim of identifying the main risk factors in the university’s chemistry laboratories. Risk factors identified in this phase were used in the next phase of this study to assess the perceived risk. The second phase consisted of surveying undergraduate students enrolled in chemistry laboratory courses, to measure their risk perceptions of risk factors in the laboratories. The study took place from March 2015 to May 2016 on the university campus. The Bioethics Committee of the university approved the study protocols and materials.

Risk Perception Tool

The Workers’ Risk Perception Dimensional Evaluation (EDRP-T [acronym in Spanish]), a 10-item tool,31,34 was used to characterize students’ perceptions of risk in the university chemistry laboratories (see Table S1 in the Supporting Information). The EDRP-T questionnaire is based on the psychometric paradigm and, according to its developers, can be used in different contexts and for several risk factors.42 This instrument is part of the Technical Note on Prevention (NTP [acronym in Spanish]) 578 published by the Spanish National Institute of Safety and Health at Work (INSHT [acronym in Spanish]) in 2001. The EDRP-T questionnaire is comprised of nine questions related to dimensions of risk perception: 1. personal knowledge; 2. expert knowledge; 3. dread; 4. vulnerability; 5. severity of consequences;

Risk Factors Selection

This first phase of this study used the Delphi technique to identify and prioritize risk factors that would best represent the safety conditions in the university chemistry laboratories. In general, the Delphi method uses systematic procedures to establish priorities and determine the amount of consensus in the group regarding those priorities.40,41 A total of 24 faculty members and technicians from the Department of Chemistry and Biology Sciences (DCBS) who were currently teaching or have taught chemistry laboratory courses in the last year were invited to participate. The C


Journal of Chemical Education 6. 7. 8. 9.

Article

noise impedes the effective communication of instructions and disturbs both students and professors.

avoidability; controllability; catastrophic potential; immediacy.

Survey Questionnaire Pilot-Testing

The EDRP-T risk perception questionnaire was pilot-tested using cognitive interviews to gather information regarding the readability of the questions and to anticipate logistics during the data collection stage. Moreover, the pilot-testing was used to gather content validity evidence based on the response process. It determined the extent to which the thought processes of participants demonstrated that they understood each question in the same way that it was defined by the instrument developers.43−45 First, the survey questionnaire was administered individually and in-person to 15 students to evaluate the language, readability, and interpretation of each question. Students were asked to read each item aloud, give their interpretation about what was being asked, and discuss what information they retrieved to give a response on the seven-point Likert scale. No prompts were provided to facilitate students’ explanation of their thought process. The questionnaire was then pilot-tested again with the aim of evaluating the logistics of a self-administered online survey. The survey was sent out through SurveyMonkey to a convenience sample of 160 students enrolled in the nutritional sciences, clinical chemical biology, or food chemistry undergraduate study programs. The survey was available for 1 month, and after that period the data were collected and analyzed.

Question 10 was comprehensive and was designed to obtain an estimate of the overall magnitude of the perceived risk. Item responses for questions 1−9 are based on a seven-point response format scale with only the end categories labeled (very low risk (1) and very high risk (7)) and numerical values (2, 3, 4, 5, and 6) in between. The last question measures the overall risk perceived, on a scale from 0 to 100 with anchoring end points labeled as very small risk and very high risk). The scale is suggested as a flexible evaluation tool that can be adapted to different types of risks; therefore, the questionnaire was used to evaluate students’ perceptions of risk conditions in university laboratories. The survey questionnaire was adapted to respond to the study goals by including demographic questions such as age, gender, current year of study (semester), and degree, as well as lab-related injury questions such as whether the student had experienced an incident/accident in a laboratory course. Additionally, questions about expert knowledge were adapted to measure perceptions regarding the safety knowledge exhibited by laboratory professors and technicians. Question number 10 (regarding overall risk) was modified to be more relevant to the environment of laboratories; thus, instead of asking about “risk of death”, the question inquired about the “risk of a severe accident or severe illness”.

Participant Recruitment

The study was conducted with undergraduate students from the DCBS at the university campus. Selection criteria for participating students included being a student actively enrolled in nutritional sciences, clinical chemical biology, or food chemistry and being enrolled in a chemistry laboratory course at the time of completing the survey. Researchers visited the classroom to inform potential participants about the study objectives, methodology, and benefits, as well as to invite them to support the research by answering the risk perception survey. The students were informed that their participation was voluntary and confidential; thus, their identity would not be disclosed at any stage of the study. Students who agreed to participate signed an informed consent form and provided an e-mail address where they wanted to receive the link to fill out the questionnaire. No compensation was provided for participating in the study.

Sample Background

The study targeted students majoring in three areas: (i) food chemistry; (ii) clinical biochemistry sciences; and (iii) nutrition sciences. Although these majors demand fewer chemistry courses than what is required for a chemistry major, they require a significant number of core chemistry courses. During the first half of a 5 year study plan, students in food chemistry and clinical biochemistry sciences are required to take 13 chemistry core courses (lecture and laboratory), including general chemistry (one course), inorganic chemistry (one course), organic chemistry (three courses), analytical chemistry (three courses), biochemistry (two courses), and physical chemistry (three courses). Students from nutritional sciences are required to take five core courses from the 13 courses previously listed. In the second half, core courses vary by major but include at least the following: food analysis (two courses) and clinical chemistry (two courses). The curriculum also includes a first-semester course in Health and Safety in Laboratory Environments that is optional for students in nutritional sciences but mandatory for students in the other two majors. The use of volatile solvents is widely extended during the laboratory practices associated with chemistry classes described above. For example, alcohols, inorganic acids, and organic and halogenated solvents are commonly used for synthesis, solubility tests, and extraction processes. Hexane or petroleum ether is used to extract fat from food samples. To date, the use of microscale kits has not been implemented except for kit analysis in clinical chemistry classes. All of the laboratories have extraction hoods; however, they have been in use for several years without regular maintenance resulting in high noise levels during their operation. Therefore, some professors may prefer to turn them off arguing that the

Data Collection and Analysis

The questionnaire was sent by e-mail via SurveyMonkey to students who signed the informed consent form. The survey was available for participants during a 4 month period corresponding with the academic term, and then the data were downloaded to be analyzed. Parametric analytical approaches are acceptable for Likert-type response format when some specific criteria are met.46−48 Individual rating items with numerical response formats at least five categories in length may generally be treated as continuous data. Numerical scales in the response set with equally spaced integers can suggest an interval level of measurement from a statistical standpoint and thus allow the use of parametric approaches.46 These conditions were applied for the statistical analysis of the EDRP-T data; however, all of the assumptions (normality, equal variances, and sample size) were tested, and when the data did not support them, nonparametric tests (i.e., D



Article

Survey Questionnaire Pilot-Testing

Friedman test) were used. Descriptive statistics, including frequencies and percentages, were generated for the demographic variables. Correlation analyses using Spearman’s correlation coefficient was completed to assess the relationship among EDRP-T 9 dimensions. The Friedman nonparametric for repeated measures test was used to assess for differences in risk perception among risk factors evaluated. The Friedman test is appropriate when the variable is measured at the ordinal or continuous level, and it does not need to be normally distributed. To supplement p-values, effect sizes in terms of Cohen’s d were estimated based on nonparametric tests to quantify the size of mean differences. Then, to determine the factor structure of risk perception dimensions for each risk factor evaluated, an exploratory factor analysis (EFA) with a varimax rotation was conducted. We explored the latent factors under the nine dimensions of perceived risk proposed in the EDRP-T survey. The eigenvalues-greater-than-one rule was used, and a minimum of 50% of variance accounted for was required to select a final model. Due to the lack of a common factor structure across the three risk factors evaluated, the nine EDRP-T dimensions were used to predict the overall perceived risk. The association between the perceived-risk dimensions (questions 1−9) and the overall risk perception (question 10) was tested using generalized linear models with demographic variables integrated step by step as control variables. The generalized linear regression allows for the dependent variable to have a non-normal distribution. The nine perceived-risk dimensions were the independent variables, and the overall risk perception was the dependent variable. The analyses were performed using the IBM Statistical Package for the Social Sciences (SPSS) v. 23.

■

The first step of the instrument pilot-testing (n = 15) resulted in minor wording changes in items A2 and A8. Item A2 was reworded to clarify that the item referred to the laboratory safety responsible(s). Item A8 was modified to clarify that this question inquired about safety-related situations that could affect other students inside or outside of the classroom. The response process validity provided evidence that students consistently understood most of the concepts assessed in each item. Items A1 and A2 were identified as inquiring about knowledge levels regarding hazards associated with conditions and chemical substances handled in the laboratory. Item A3 was clearly linked to fear, dread, and distress of incurring personal harm. Item 4 was associated with the possibility of suffering personal harm due to the exposure to the risk factor, while item 5 was linked to the extent of the consequences. Items A8, A9, and G1 provided unambiguous interpretations. Item 6 (avoidability) and 7 (controllability) required rewording because the items were initially perceived as measuring the same concept. Of the 160 students who had agreed to participate, only 102 responded to the online self-administered surveya response rate of 64%. The pilot tests showed that 95% of the participants answered the questionnaire in a range of 7.5−12 min; inattentive or careless responses were identified through repetitive answer patterns in 1% of the surveys. These findings made it possible to set criteria for the large study survey administration and data management. Perceived Risk Dimensions

A total of 1382 students from a population of 1648 enrolled in the DCBS agreed to participate in the risk perception survey. Of those potential participants, 617 responded to the survey a response rate of 45%. Of these 617 surveys, 95 were discarded because fewer than 80% of the questions had been completed. The final sample was comprised of 521 students corresponding to 32% of the total population of DCBS students. By the study program, the sample included 34% of Clinical Bio-Chemistry Sciences students, 32% of Food Chemistry students, and 26.5% of Nutrition Sciences. Participants were on average 20 years old (SD = 2.87; range of 17−50) and mainly women (65% female) (Table 1). Risk perception scores across the nine qualitative dimensions varied from 3.1 to 6.3, with an average of 4.7, on a scale of 1−7 (very low to very high) (Figure 1). Across the ninedimension scale, inhaling chemical substances was the risk factor with the highest average perceived risk (4.9). However, the overall risk perception (0−100 scale) for the three risk factors (laboratory work, chemical-substance splashes and spills, and inhaling chemical substances) varied from a score of 43 to a score of 49, with laboratory work being the risk factor with the highest perceived risk at an average score of 49. While the average of the nine-dimension instrument was above the scale mean (3.5), the average for the overall risk-perception question was below the scale’s mean (50) for the three risk factors evaluated. Friedman’s test was performed to determine the differences in risk-perception dimensions and overall perceived risk among students for each risk factor evaluated (Table 2). Significant differences were identified across the risk dimensions and overall perceived risk except for the expert knowledge dimension. Effect sizes in terms of Cohen’s d for nonparametric approaches indicated mainly small effects (0.2− 0.4), intermediate effects (0.5−0.7), and a few large effects

RESULTS

Delphi Method

A total of 12 out of 24 faculty members and technicians responded to the invitation to participate in the Delphi evaluation at the time of this study, three participants reported to have less than 3 years of experience teaching in the university chemistry laboratory and three reported 3−18 years of experience, while the remaining six reported between 18 and 24 years of experience. In face-to-face interviews, the 12 members identified dangerous situations in the university chemistry laboratories. A list of 13 risk factors was identified in this first round. The most common hazardous situations that emerged at least once during the interviews with different faculty/staff members were these: • the use of chemicals as a risk factor due to splashing on the skin and eyes (7 times); • the use of materials and hot substances causing burns (4 times); • the use of glass and sharp objects causing cuts (4 times); • inhalation of chemical substances (3 times). During the second round, laboratory material handling and hot substances causing burns ranked first, followed by factors leading to the occurrence of cuts. Inhalation of chemicals ranked third, and then chemical-substance splashes. From these results, three risk factors were identified to be included in the perceived-risk survey: (i) laboratory work, which included the handling of hot materials and sharp objects; (ii) splashes from chemical substances; and (iii) inhalation of chemical substances. E



Article

the expert knowledge dimension (6.3), the personal knowledge dimension (5.3−5.4), and the avoidability dimension (5.0− 5.4), which measured the extent to which the participant believed he or she was able to prevent the risk factor from triggering a dangerous situation. Students perceived themselves to be less vulnerable to injuries caused by laboratory work than to those caused by inhaling chemical substances as they perceived themselves to be more able to avoid consequences derived from exposure to laboratory work than from inhaling chemical substances. Aligned with these findings, the controllability dimension, which measured perceived individual ability to control any damage caused by a risk factor, was scored higher for laboratory work than for inhaling chemical substances.

Table 1. Demographics of Participating Students in the University Laboratories Risk Perception Survey Variables Gender Male Female Age in years 18−20 21−24 25−50 Study program Clinical Biochemistry Sciences Nutrition Sciences Food Chemistry Academic yeara First Second Third Fourth Fifth Laboratory-related accidents

Participants (Total n = 521)

Percentage

182 339

35 65

360 145 14

69 28 3

300 122 99

57 24 19

203 97 123 73 24 124

39 19 24 14 5 24

Factor Structure of Risk Perception for Risk Factor

The correlation analysis showed significant correlations among some EDRP-T risk dimensions. Subsequently, for each of the three risk factors under study, the nine risk dimensions were subjected to EFA, using a maximum likelihood estimation method with Varimax rotation. The Kaiser−Meyer−Olkin (KMO) test statistic was ≥0.7, which is considered moderate, and Bartlett’s test of sphericity was significant (p < 0.001), indicating an acceptable solution for each one of the three risk factors. There were three components with an eigenvalue > 1.0. In each component, only items with loading higher than 0.4 were retained. The three-component structure explained at least 54% of the variance (inhaling chemical substances, 60%; chemical-substance splashes and spills, 57%; laboratory work, 54%). Results from EFA indicated that the EDRP-T risk dimensions could be condensed to three latent factors (components) (Table 3). In light of Slovic’s37 work, component 1 could be labeled as “dread risk”. Component 1 was defined by four dimensions: dread, vulnerability, severity of the consequences, and catastrophic potential for all three risk factors evaluated. Dread risk characterizes a person’s emotional response to dread, estimation of the personal

a

These are 5 year programs; each academic calendar year includes two terms.

(0.8

Assessing College Students' Risk Perceptions of Hazards in

Recommend Documents