Developing a Questionnaire to Assess Safety Climate at Universities of Medical Sciences

An effective way to achieve the goal of identifying hazards before the occurrence of events is to measure the overall state of safety, A-R-T-I-C-L-EI-N-F-O A-B-S-T-R-A-C-T Article Notes: Received: May 29, 2019 Received in revised form: Jun 10, 2019 Accepted: Jun 11, 2019 Available Online: Aug 3, 2019 Background & Aims of the Study: The implementation of safety principles in work environments can directly benefit both employees and the organization by reducing the mental and physical strain, reducing the risk of work-related injuries, and performance improvement. Thus, with the consideration and implementation of safety principles, and the creation of a positive safety climate, organizations can improve their performance and the welfare and safety of their employees. The purpose of this study was to design and provide a domestic questionnaire of safety climate assessment at universities of Health Ministry in 2018-19. Materials and Methods: This descriptive-analytical study was carried out with the participation of health and safety specialists, methodologists and personnel of the universities of medical sciences. Universities and people were selected randomly. The validity of the questionnaire was measured by CVI and CVR analysis and then the structural validity was studied by factor analysis using R ،MPLUS and SmartPLS softwares.

as well as safety climate and culture (1). It has to be said that extensive activities have been carried out in various industries in the field of safety climate assessment, but so far no substantial research has been conducted on the safety of the educational centers especially in the universities, and this is the importance and novelty of the study. Obviously, the human resources of the educational centers, if they take into account the issues of safety and constructive thinking in this field, can play a special role as an important contributing factor in promoting the safety of society through influencing learners. Based on this, the researchers believe that by determining the safety status in these organizations, the culture of safety can be greatly improved. Since each change requires its own infrastructure, as well as the acquisition of the necessary infrastructure also needs its own tool; by providing a valid questionnaire for educational environments, safety-centered plans can be developed by comprehensive data at national level. The subject of this study is development of a questionnaire on the study of safety in universities as an educational indicator. As a complex and diverse environment, the university can be the focal point for events. The existence of different colleges brings to mind the range of activities from administrative to laboratory and technical activities. Activity in the lab creates a range of ergonomic, chemical, and even biological hazards. How people look at these issues is critical and directly affects the way in which operational programs are planned and how they are managed. Several questionnaires have been provided by various researchers regarding the status of occupational/safety status and the type of systems thinking in the organization (2). One of the most important factors in determining these factors is the type of organization's activity. Undoubtedly, scientific and educational environments like universities are different from industrial environments such as process, automotive or nuclear industries. Therefore, we need to use a questionnaire that assesses items related to activities in these environments. Since safety climate is in fact the attitude of individuals towards the safety and the existing conditions, it also affects the safety behavior of individuals. Hence, by assessing and improving the level of safety in the organization, it is possible to improve safety behaviors (2). Considering the existing deficiencies, this research can open up a number of issues in the area of safety in educational centers and provide a basis for future research in this area. In fact, safety climate is one of the tools that can be used to examine the management's performance in terms of safety, and ultimately, based on the results try to planning and correction (2). Improving safety climate is one of the most important factors in preventing dangerous conditions and incidents in the universities, and encourages the managers and employees to adhere to the safety standards.

Aims of the study:
The aim of present study was to develop a safety climate assessment questionnaire in universities affiliated to the Ministry of Health in Iran, 2018-19.

Study area and sampling points
This analytical and cross-sectional study was carried out among the staff and faculty members of nine medical universities in Iran, according to the territorial planning map. The design of the instrument to assess the safety climate was carried out in three steps. In the first stage, using the literature review and also studies on the safety climate in other organizations, with the advice of experts in the field of occupational health and safety, the primary pattern and content scope of the questionnaire were determined (3,4). Then, analysis was carried out using a focused group approach and an interview in order to ensure greater consistency and comprehensiveness of

Materials & Methods
the domains as well as the maximum fit with the working environment. The analysis team included health and safety experts and methodologists. In order to better match, the views of some people from the target community were gathered through interviews. Then the initial version of the questionnaire was prepared and reviewed according to the general principles of designing the questionnaire presented in various sources (4-7). Then, face and content validity (12 experts) of the designed questionnaire was studied. The validity of the questionnaire was assessed in four items: simplicity, relevance, clarity and necessity, and analyzed by CVI and CVR indices. Totally, considering that the number of questionnaires in the preparatory phase was 37, 315 questionnaires were distributed among the universities and sample members (faculty members, staff at the faculty and university central part and in units such as laboratories, libraries, facilities, etc.) from the university(s) selected randomly. People with one year or more work experience who were better informed about their organization's circumstances were included in the study. Meanwhile, participants had the discretion to leave the study at any stage if they did not want to continue. Reliability of the questionnaire was assessed by test-retest within one week among the subjects and the internal consistency analysis of the instrument through the Cronbach's alpha. After confirming the validity and reliability of the questionnaire, Exploratory Factor Analysis was used to investigate structural validity in R software and Confirmatory Factor Analysis using MPLUS and SmartPLS softwares and final version of the questionnaire was provided. In addition, analytical tests of T, and Mann-Whitney were used in SPSS V20. It should be noted that answers were in five-point Likert form (very agree, agree, no opinion, disagree, very disagree) and rated in 1-5. Furthermore, a demographic questionnaire including age, work experience and number of completed training courses was used.
After reviewing the texts, 37 questions were obtained and examined by a panel of 12 specialists. The validity of the questionnaire showed that some of the questions are not favorable. Therefore, the questionnaire was revised and relevant questions were rewritten or modified. Subsequently, face and content validity of the re-evaluated instrument was tested and conditions were acceptable at this stage. The average agreement rate of people in relation to the simplicity the questionnaire was 0.88. The minimum agreement rate was 0.58. The average agreement rate of experts regarding necessity of questions was 0.99, while the minimum agreement rate was 0.92. The ratio of agreement in each question based on simplicity, clarity and relevance is presented in table 1. According to the obtained results from CVR and CVI, the questions will remain in the questionnaire. Then, the reliability of the questionnaire was measure by test-retest in a university within one week. Correlation coefficient within the group was equal to 0.96, indicates that the questionnaire has stability (repeatability power).

Demographic factors
After co-ordination at different universities in the amateur areas, the participation of universities in the two regions 1 and 4 was not possible. On the other hand, the staff of the two universities in the 3rd district cooperated in the study (table 2). Finally, 265 cases out of 315 distributed questionnaires, were returned and used in analyzes. Of these, 147 people (55.5%) were male and the rest were women. Also, 204 (77%) were married. Regarding the level of education of the participants, most of them had a doctorate or higher degree (43%), while those with associate degree or less was 6.4%. Table 3 shows additional information in this regard.

Exploratory Factor Analysis
Given that the KMO index is higher than 0.05 (0.792), it can be said that the sample size was sufficient. Also, the probability of testing the Bartlett symmetry index is 0.000, which is lower than 0.05 so, this model is not faced with the problem of linear multiplicity. The exploratory factor analysis revealed that the software extracts 9 factors for 37 questions. Some questions were in two or three factors; given their nature, matched questions fell into one category. After this stage, two questions 3 and 19, which were based on software analysis in only one factor, remained alone, and since • Developing a questionnaire to assess safety climate at.. one question for a factor could not be sufficient; these two questions were left out of the questionnaire. Seven remained factors were identified under the heading of management commitment and performance in the field of safety, training, knowledge, workplace and conditions, personnel commitment and participation in the field of safety, emergency response and emergency preparedness. Table 5 lists the questions for each factor. By deleting questions 3 and 19, the final version of the questionnaire has 35 questions with total score of 35 to 175.

Confirmatory Factor Analysis
The output from exploratory factor analysis was verified by confirmatory factor analysis and was re-evaluated by WLSMV estimation method.

Convergent Validity
This kind of validity indicates that the indices of a structure how much are contributing in explaining the common variance. To assess convergent validity, two criteria are considered simultaneously: factor loading and Average Variance Extracted (AVE), Composite Reliability (CR) can also be another criterion in this field, which is presented in table 6 for this questionnaire. Also, in the present study, the proportionality indices of Standardized Root Mean squared Residual (SRMR) and Normed Fit Index (NFI) were used and were • Developing a questionnaire to assess safety climate at..

Instrument reliability
Reliability of the questionnaire was analyzed by Cronbach's Alpha and the alpha value was 0.909. The alpha for each of the seven factors is also given in table 6.

Predictive validity
With regard to the close relationship between safety and occupational accidents both in the industry (8) and in academic settings (9,10), the testing differences of safety climate score and its sub-factors between the two groups of people who have experienced and not experienced accidents in the workplace used to assess the predictive validity of the tool, which is commonplace in this field (8). After evaluating the normality of the data by Kolmogorov-Smirnov test it was depicted that only the total safety climate was normal (P>0.05) and its sub-scales were not normal (P<0.05). Therefore, independent t-test was used for total score and Mann-Whitney test was used in the remaining cases. T-test did not illustrate a significant difference between two groups (P=0.069). However, two factors of management commitment in the field of safety, and workplace and conditions had a significant difference (P<0.05). Those who had accident experience in the work environment had a lower average score in the management commitment and performance factor (41.79 vs. 46.33). On the other hand, this group of people had a lower score on participation in the field of safety (14.53) compared to those who did no accident (15.39). In addition, although the difference in total safety score between these two groups was not significant, the safety climate among participants who did not experience accident was better (with an average of 105.82 versus 99.80). Taking into account the principles of safety in the workplace is critical to achieving the goals of the organization and its productivity growth. Organizational subcultures such as safety culture can play a major role through better involvement of personnel. A definite definition of safety culture has not been stated, but it can be considered as an organizational subculture that affects safety behavior of individuals. It is not unreasonable to consider Dominic Cooper's safety culture model for university safety. This model, presented in 2000 considers culture in three parts: employees: climate, behavior, management (11). In the case of the first element, employees' attitudes will be assessed, it is a perceptual audit that questionnaire is the most important and most used tool in this regard. The second element is employees' behavior, which is evaluated using behavioral sampling. The third component is management of the organization, which reviews the activities of organization's management. The most important of these actions are laws, guidelines, resource allocation, communication, planning and control. The evaluation of this element is externally/objective and the people evaluate the various factors objectively. Hoffmeister et al., quoted from Schneider and Richards, that organizational climate theory states that the climate derived from employees' efforts to understand their work environment so that they can estimate what types actions are supported at work (12). In a series of steps, the safety climate assessment questionnaire in universities in Iran was presented in this study. Then a trial was done and reviewed to find evidence of the initial validity of the tool. The factor analysis confirmed the seven-factor model (management commitment and performance in the field of safety, personnel participation and commitment in the field of safety, training, knowledge, workplace and conditions, emergency preparedness and emergency response). Utilizing the tool in universities and analyzing data showed that it has a total reliability of 0.92 which is acceptable compared with 0.7, which is considered as a desirable reliability level (13). Also, its subscales were in acceptable condition.
The findings indicated that the questionnaire was acceptable. Since human resources are the main and vital part of organizations, recognizing the status of the organization from their point of view is important. In the area of safety, it is also necessary to have an instrument to assess its climate. This tool can help managers to plan for the future. It is worth noting that other factors and places can also be investigated in future studies.