Learning Center
Plans & pricing Sign in
Sign Out

Developing A Safety Climate Scale


									                   Developing A Safety Climate Scale

                                     Tsung-Chih Wu
                       Department of Industrial Safety & Health
                           Hungkuang Institute of Technology

     After many years of developing and testing, it has been suggested that safety
climate surveys are a much better predictor of an organizational safety performance
than other indicators tested. This presentation focuses on psychometric integrity of
safety climate measurement and important dimensions of safety perceptions in lab
workplace in colleges. First, basing on an extensive review of the literature, the
researcher developed an initial pool of items, convened an experts conference, and
then tried out the original questionnaire. Finally, item analysis, validity analysis, and
reliability analysis were carried out. The principal-component factor analysis of the
safety climate scale resulted in five factors, top management’s commitment and
actions, managers’ commitment and actions, personal commitment to safety,
emergency response, and perceived risk, explained 66.18% of the total variance.
Moreover, the internal consistency of the scale reached a Cronbach α value of .9516.
It is concluded that the scale is a valid and reliable instrument.
Keywords: Safety management, Safety culture, Safety climate.

1. Introduction

1.1 Background

     A college is the place to foster high-level persons, and may be also the place to
kill the persons. Firstly, a graduate student Maa caught his body between the pillars of
reinforced concrete when he practiced inside a lab in National Chiao-Tung University
on August 1996, he became disabled (Jiang, 1997). Secondly, an undergraduate
student Ke was shocked by electricity when he was practising with wires inside a lab
in National Hu-Wei Institute of Technology on October 1998, the result, death (Gau,
1998). Thirdly, an associate professor Jang touched a naked wire when he supervised
construction in National Kao-Hsing Normal University, his fate death, too (Wang &
Lan, 2000). Even though, these disasters were a part of accidents, many other
accidents may not be uncovered. After all, those who frequent labs have the right to

prevent themselves from being harmed, and the persons in charge have the
responsibilities to protect them from dangers in the workplace.
     After inspected by Labor Inspection Agencies in 1997, many labs in 33 colleges
in Taiwan were found to have a lot of weaknesses in which (Chen, 1998; Shyu, 1998),
listed in Table 1. From this table, particularly, we can see that there are many
imperfections in safety and health management. Hence, Colleges should enforce
perfect actions in order to prevent them from being punished, and protecting the
safety and health of workers and students, and improving their safety performance.
Eventually, a survey for safety climate may assess and cognize the potential hazards
within workplaces (Dedobbeleer & Beland, 1991), and may provide the frame of
reference for improving safety.

      Table 1 Results of inspecting labs etc. in colleges 1997
OSHAct        Content                               Frequencies  Percent            Rank
Article 5     Safety & health installation                   36     58.1               3
Article 6     Machinery & tools safeguarding                   5     8.1              10
Article 7     Operation area monitoring &                    29     46.8               6
              dangerous materials labeling
Article 8     Dangerous machinery or                           6     9.7                9
              equipment inspecting
Article 12 Physical & health examination                     30     40.4                5
Article 14 Safety & health structure                         51     82.3                1
Article 15 Certified operator for dangerous                    9    14.5                7
              machinery or equipment
Article 17 Notifying the contractor of the                     1     1.6               12
              work environment etc. in advance
Article 23 Safety & health training                          32     51.6                4
Article 24 Disseminating OSHAct                                2     3.2               11
Article 25 Preparing safety & health work                    42     67.7                2
Article 29 Completing monthly reports on                       8    12.9                8
              the statistics of occupational
                    Source: Labor Inspection Department, CLAEY(1998)

1.2 Safety climate

     Petersen (1996) indicated the safety culture determines organizational
effectiveness. One of the biggest difficulties of constructing an active safety culture is
to persuade employees’ hearts and minds to the safety improvement cause. By safety
climate surveying, labors can be persuaded that the organization is very serious in
striving to create an active safety culture which in turn will help to win over their
hearts and minds to the safety performance (Cooper, 1998).

     Although safety culture and safety climate are often regarded as important
concepts, not much consensus has been reached on the cause, the content and the
consequences of safety culture and climate in the past twenty years. Moreover, there
is an overall lack of models recounting the relationship of both concepts with safety
performance (Guldenmund, 2000). Safety climate is generally acknowledged to be the
perception of safety culture. Cooper (1998) proclaimed safety climate is a perceived
image that is mostly concerned with labors’ perceptions of the importance of safety
and how it is enforced within the organization. Safety climate is defined as a set of
molar perceptions of safety culture, shared by individuals, which is affected by
organizational factors and personal factors, and influences the safety behaviors of
employees. Reviewing the related safety literature (Zohar, 1980; Brown & Holmes,
1986; Dedobbeleer & Beland, 1991; Seppala, 1992; Niskanen, 1994; Coyle et al.,
1995; Williamson et al., 1997; Diaz & Cabrera; 1997; Cooper, 1998; Hayes et al.,
1998; Wu, 2001), this author points to five principal areas within safety climate. That
is, top management’s commitment and action, managers’ commitment and action,
personal commitment to safety, perceived risk, and emergency response.

1.3 Objective

     The aim of this paper is to construct the psychometric integrity of safety climate
measurement and important dimensions of safety perceptions in lab workplaces in
colleges. That is to say, the author attempts to develop a reliable and valid safety
climate scale as a tool to measure employees’ perception of lab safety within colleges
in Taiwan.

2. Method

2.1 Population and subject

     There were 150 universities and colleges, including 53 public and 97 private
schools, in Taiwan by September 2000 (MOE, 2000). It was estimated that within the
colleges at the time of the study, there were 45,000 workers. The teachers and other
workers frequent labs, testing rooms, practice shops, and testing shops is the subject
of this study. However, the total of these workers is unknown.

2.2 Questionnaire design

     A questionnaire was used to collect the data. A first draft of the questionnaire

was developed on the basis of a review of the literature regarding safety climate (Wu,
2001; Diaz & Cabrera, 1997; Hayes, et al. 1998; Williamson et al., 1997; Cooper,
1998; Coyle et al., 1995). The content of the draft questionnaire was discussed with
two senior safety managers from two colleges involved in the study. Each item of the
questionnaire was considered. The amended questionnaire was then sent to 3 experts
and 2 scholars who were asked to review the questionnaire and give their feedback.
The final draft version of the questionnaire contained a total of 62 questions, it was
four pages long with a cover page of instructions and contained closed format
questions. Closed questions required the respondents to rate their opinions on issues
relating to general information and safety climate. Respondents were not asked for
their names or the names of their college. They were asked to give a frank and honest
account of their opinions.

2.3 Questionnaire content

        The questionnaire used in this study contains general information and safety
climate scale. Questionnaire about general information and safety climate is described
as follows.

2.3.1     General information
        A 12-item questionnaire was prepared asking for information about
organizational factors and personal factors. The former consisted of size, nature,
staffing safety manager, establishing safety committee, and location. The latter
contained sex, age, the length of time in position, position level, accident experience,
safety training, and department.

2.3.2       Safety climate scale
        A total of 50 items referring to five dimensions were ready, with five-point
Likert scales (highly disagree, disagree, not sure, agree, highly agree) in order to
evaluate the subject’s agreement with them. These dimensions were top
management’s commitment and action, managers’ commitment and action, personal
commitment to safety, perceived risk, and emergency response. Each dimension of the
climate included ten items.

2.4 Procedure

        After judgement sampling, packs containing 50 questionnaires were posted to
each of nine colleges on October 2000. Return addressed envelopes and souvenirs

were supplied. A letter highlighted the aims of the survey accompanied the
questionnaires, that it was supported by the National Science Council. Two weeks
later, follow-up telephone calls were made to the safety department of each college,
asking them to remind their employees to complete and return their questionnaires if
they hadn’t already done so.

2.5 Data analysis

      Data from 374 managers and workers were used in the analysis. Item analysis
was used to screen suitable items of safety climate scale. Exploratory factor analysis
examined the construct validity of the scale. The Cronbachα coefficient was used to
test the internal consistency of the scale.

3. Results

3.1 Item analysis

      From the obtained data, all 50 items there were with a correlation with the scale
total over .30. The internal consistency of the scale reached a Cronbachα value
of .9534, and the Cronbachα value reached .9535 if item # 34 is deleted. Thereupon,
the item was omitted.

3.2 Validity analysis and reliability analysis

      The climate data, then, were factor analyzed using a principal-component factor
analysis with varimax rotation. This procedure resulted in six factors with an
eigen-value of >1 (KMO = .929; Bartlett = 15078.758; df = 1176; p = .000), which
explained 68.18% of the total variance, and all 49 factor loading coefficients were
greater than .50. Next, the internal consistency of the scale reached a Cronbachα
value of .9535. It included 49 items.
     The Factor 6, however, reached a Cronbachα value of .9164 if item # 48 is
deleted. This factor merely contained two items, item # 49 and item # 50, after item #
48 was deleted, and then this factor was omitted because it covered little valuable
      The remained climate data were factor analyzed again. This procedure resulted
in five factors, named as top management’s commitment and action, managers’
commitment and action, personal commitment to safety, emergency response, and
perceived risk (Table 2), explained 66.18% of the total variance, and all 46 factor

loading coefficients were greater than .50. Then, the factor scree plot showed as
Figure 1. Furthermore, the internal consistency of the scale reached a Cronbachα
value of .9516. It included 46 items.

       Table 2 Eigen-values, variances, number of items, and Cronbach α value for
                                  each safety climate factor
                                                                  Number of    Cronbachα
             Names                 Eigen-values    Variances
                                                                    items         value
Top management’s
                                          7.530       16.37%              10         .9612
commitment and action
Managers’ commitment and
                                          7.240       15.74%              10         .9524
Personal commitment to
                                          6.782       14.74%              10         .9358
Emergency response                        4.732       10.29%               7         .9222
Perceived risk                            4.160        9.04%               9         .8492
Total                                                 66.18%              46         .9516


         e      10

         l      4
                     1   4   7   10 13 16 19 22 25 28 31 34 37 40 43            46

                     Factor number

                                     Figure 1 Factor scree plot

4. Discussion

     Generally scales possessing the psychometric integrity must have a good
validity and reliability. Validity is the degree to which the measurement what it is
supposed to measure, and reliability is the degree to which the measurement
consistently measures whatever it measures (Gay, 1992). The instrument of safety

climate measures should have good validity and reliability, and then the outcomes of
measures could be explained properly and trustworthily.
     For deciding on how many components to retain, the most widely used criterion
may be Kaiser (1960). Kaiser proposed retain only those components whose
eigenvalues are greater than 1. Moreover, Cattell (1966) suggested a graphical method
called the scree test. In this method, what happens is, generally, that the magnitude of
successive eigenvalues drops off and then tends to level off. The suggestion is to
retain all eigenvalues in the sharp descent before the first one on the line where they
start to level off (Stevens, 1992). On the other hand, the Cronbach α value is
below .70 should never be considered suitable, nor should that cannot differentiate
between different levels of safety climate (Cooper, 1998).
     The result of this study indicates those five components whose eigenvalues were
greater than 4 were retained, and scree plot also showed that five components may be
retained. Additionally, the internal consistency of each sub-scale was greater than .80,
and total-scale was greater than .95. Overall, this safety climate scale in the study is
quite valid and reliable.


     The National Science Council of the Republic of China under grant NSC
89-2511-S-241-001 supported this research. The support received was greatly


Brown, R. L. & Holmes, H. (1986). The use of a factor-analytic procedure for
     assessing the validity of an employee safety climate model. Accident Analysis &
     Prevention, 18(6), 455-470.
Cattell, R. B. (1966). The scree test for the number of factor. Multivariate Behavioral
     Research, 1, 245-276.
Chen, S.-S. (1998). Preparing for safety and health inspection in colleges. Labor
     Inspection Department, Council of Labor Affairs Executive Yuan. [In Chinese]
Cooper, D. (1998). Improving safety culture: A practical guide. England: John Wiley
    & Sons.
Coyle, I. R., Sleeman, S. D., & Adams, N. (1995). Safety climate. Journal of Safety
    Research, 26(4), 247-254.
Dedobbeleer, N. & Beland, F. (1991). A safety climate measure for construction sites.
     Journal of Safety Research, 22(2), 97-103.

Diaz, R. I., & Cabrera, D. D. (1997). Safety climate and attitude as evaluation
     measures of organization safety. Accident Analysis & Prevention, 29(5),
Gau, S.-R. (1998). A student killed by electric shock. Central Daily, October 9, 1998,
     7. [In Chinese]
Gay, L. R. (1992). Educational research: Competencies for analysis and application.
     New York: Maxwell Macmillan International.
Guldenmund, F. W. (2000). The nature of safety culture: A review of theory and
    research. Safety Science, 34, 215-257.
Hayes, B. E., Perander, J., Smecko, T., & Trask, J. (1998). Measuring perceptions of
     workplace safety: Development and validation of the work safety scale. Journal
      of Safety Research, 29(3), 145-161.
Jiang, J.-C. (1997). Maa is singing the song of hope. China Times, January 23, 1997,
     5. [In Chinese]
Kaiser, H. F. (1960). The application of electronic computers to factor analysis.
     Educational and Psychological Measurement, 20, 141-151.
Ministry Of Education (2000). List of universities and colleges in Taiwan.
Niskanen, T. (1994). Safety climate in the road administration. Safety Science, 17,
Petersen, D. (1996). Safety by objectives: What gets measured and rewarded gets
     done. New York: Van Nostrand Rrinhold.
Seppala, A. (1992). Evaluation of safety measures, their improvement and
     connections    to   occupational accidents.     (FT Dissertation,    Institute   of
     Occupational Health, Finland, 1992). ProQuest - Dissertation Abstracts, Not
     available from UMI.
Shyu, G.-G. (1998). Illegal safety and health in labs in colleges. United Daily News,
     October 7, 1998, 6. [In Chinese]
Stevens, J. (1992). Applied multivariate statistics for the social sciences. New Jersey:
     Lawrence Erlbaum.
Wang, R.-L. & Lan, K.-C. (2000). An associate professor with his son shocked by
     electric in National Kao-Hsing Normal University. United Daily News, March
     26, 2000, 11. [In Chinese]
Williamson, A. M., Feyer, A.-M., Cairns, D., & Biancotti, D. (1997). The
     development of a measure of safety climate: The role of safety perceptions and
     attitudes. Safety Science, 25, 15-27.
Wu, T.-C. (2001). The correlational study between safety climate and safety
     performance in four categories of manufacturing industries in the Central

     Taiwan. Dissertation, National Changhua University of Education, Taiwan,
     Republic of China, 2001. [In Chinese]
Zohar, D. (1980). Safety climate in industrial organizations: Theoretical and applied
     implications. Journal of Applied Psychology, 65(1), 96-102.


To top