You are on page 1of 10

Journal of Safety Research, Vol. 30, No. 2, pp.

113–122, 1999
Copyright © 1999 National Safety Council and Elsevier Science Ltd
Pergamon Printed in the USA. All rights reserved
0022–4375/99 $–see front matter

PII S0022-4375(99)00005-5

The Efficacy of Material Safety Data Sheets and


Worker Acceptability

Charles C. Phillips, Bill C. Wallace, Charles B. Hamilton, Robert T. Pursley,


Gregory C. Petty, and Charles K. Bayne

This study quantified how well the Material Safety Data Sheet (MSDS) dif-
fused information to workers, compared the efficacy of an OSHA type form
with the Chemical Manufacturer’s Association/American National Standards
Institute (CMA/ANSI Z400.1) structure and the International Chemical Safety
Card (ICSC), and determined worker acceptability and understanding of MS-
DSs at a large national laboratory. An inventory questionnaire and quantitative
testing were used to gather data from 160 union workers employed at the na-
tional laboratory. Most of the workers in the sample population studied re-
ported favorably on MSDS acceptability, accessibility, and understanding. All
three formats significantly diffused information to workers, although approxi-
mately one third of the information was not absorbed. The rank order of finish
between the three different formats was ICSC, OSHA type format; and CMA/
ANSI Z400.1 respectively; however, there were no significant differences in
scores. There were significant differences in how well each format answered
specific test questions. © 1999 National Safety Council and Elsevier Science Ltd

Keywords: Adoption diffusion, chemicals, chemical safety card, comprehensi-


bility, cognitive theory, efficacy, format, health belief model, laboratory, mate-
rial safety data sheets, worker education

INTRODUCTION reported in the literature (Meninger & Margolies,


1994). Further, employers may assume that sim-
Difficulty in the interpretation and use of the Ma- ply providing the MSDS to employees will en-
terial Safety Data Sheet (MSDS) has often been able them to understand all that is necessary for
protection and that such actions will meet Occu-
Charles C. Phillips, Ph.D., CIH, is a Senior Industrial Hy-
gienist in the Office of Safety and Health Protection, Oak member of the Tennessee Valley Chapter of the American In-
Ridge National Laboratory, Oak Ridge, Tennessee. He is Dip- dustrial Hygiene Association.
lomat in the American Academy of Industrial Hygiene and the Robert T. Pursley, Ph.D., is an Associate Professor in the
American Academy of Sanitarians. Department of Health, Leisure, and Safety, College of Human
Bill C. Wallace, Ed.D., is a Professor at the University of Ecology, University of Tennessee.
Tennessee in the Department of Health, Leisure, and Safety, Gregory C. Petty, Ph.D., is a Professor and Administrative
College of Human Ecology. He is a Certified Health Educa- Head of the Department of Human Resource Development,
tion Specialist. College of Human Ecology, University of Tennessee.
Charles B. Hamilton, Professor, Dr. P.H., is the Adminis- Charles K. Bayne, Ph.D., is a Statistical Consultant in the
trative Head of the Department of Health, Leisure, and Safety, Computer Science and Mathematics Division, Oak Ridge Na-
College of Human Ecology, University of Tennessee. He is a tional Laboratory, Oak Ridge, Tennessee.

Summer 1999/Volume 30/Number 2 113


pational Safety and Health Act (OSHA) require- The complexity of the MSDS varies with
ments. Others assume, as indicated by current the numerous different formats that are part of
formats in use, that the more information pro- the approximately four billion in circulation in the
vided on the MSDS form, the better for the United States (Van Santen, 1989). The Chemical
worker and employer (Samways, 1988). Either Manufacturers Association (CMA), which has
assumption could result in worker avoidance or over 200,000 chemical products in its inventory,
misunderstanding of content of the MSDS that in concert with ANSI, adopted a uniform MSDS
could lead to injury or death associated with un- format that was released as a voluntary standard
necessary exposure to toxic chemicals or their re- early in 1993. User friendliness of this format,
actions. No matter in what format the MSDS is which was expanded to 16 sections, and the
provided, the ability to understand its content or OSHA type format with 8–9 sections, needed to
message is of singular importance in protecting be further evaluated and compared in the worker
health and safety. population. The new MSDS adopted by CMA
Only a few studies on the efficacy of the MSDS and several European countries could cost mil-
and worker acceptance have been published since lions of dollars to implement (Van Santen, 1989;
the Hazard Communication Standard was imple- Campbell, 1992). Such a large outlay of resources
mented in 1983. Studies conducted by Nelkins & needed to be justified through supportive re-
Brown (1984), Hadden (1988), Kearney/Centaur search evidence that format is a significant prob-
Division of A.T. Kearney, Inc. (1991), Lu & lem associated with user understanding of cur-
Moure-Eraso (1993) were qualitative or descrip- rent OSHA type formats and hybrids.
tive in design and analysis. The perception given Further, the HAZCOM Standard, which in-
by the results of these studies was that MSDS cludes the MSDS requirement, has been cited by
formats were hard for workers to understand and employers as having no significant impact on re-
that content and readability needed improve- duction of workplace related injuries and illnesses
ment. A few of the studies suggested a more sim- (GAO, 1992). If a majority of employers believe
ple design be attempted, while others advocated this to be true (82–84%), the inability of workers
more information be added to service the needs in deciphering the MSDS could be a contributing
of a diverse user group. None of these studies de- factor (GAO, 1992). Employer difficulty in pro-
termined whether there was significant diffusion viding adequate training on a difficult to read
of information from any MSDS to the worker or MSDS could also influence this opinion (Phil-
whether there were significant differences in the lips, 1997).
efficacy of varying formats. The OSHA con- Considering the above, a study was designed
tracted Kearney/Centaur study entitled, Study on to determine the acceptability of MSDSs; assess
the Comprehensibility of MSDSs, in its final re- how well they diffuse information to workers;
port urged further research be conducted on the and provide a comparison of the efficacy of an
role format played in comprehension by workers OSHA type format, CMA/ANSI Z400.1 format,
(Kolp, Sattler, Blaymew, & Sherwood, 1993; Kear- and the International Chemical Safety Card (ICSC).
ney/Centaur Division of A.T. Kearney, Inc., 1991).
This study emerged to explore the speculation
that format might play a role in MSDS compre- METHODS
hension and as a result of a recommendation cited
in the aforementioned report. A further stimulus to The 160 workers studied were purposely selected
conduct the study was the need of a large national from a population of 824 members of the Atomic
laboratory in Tennessee to assess worker accept- Trades and Labor Council employed at a large
ability, understanding, and use of its MSDSs. national laboratory in Tennessee. Eighteen crafts
The need for the study was underscored by comprised the 160 workers. Instrumentation used
the large number of deaths, injuries, and illnesses in the study consisted of a 40-question inventory
that occur in U.S. industries every year (DHHS, on acceptability of the MSDS to the worker and a
1991). Over 482,000 occupational illnesses were 10-question MSDS comprehensibility test that
diagnosed in 1993 and 24,000 respiratory ill- was administered twice to the worker. The treat-
nesses were associated with toxic agents (NSC, ments or independent variables, three differently
1995). Of primary concern in reducing exposures formatted MSDSs, were controlled for level of
was employee understanding of protective infor- readability and specific chemical addressed. Two
mation including the MSDS (DHHS, 1991). different MSDSs and the ICSC were used as in-
114 Journal of Safety Research
dependent variables in a comprehensive test in- scores [mean 1 2 mean 2]). Parametric statistics
volving a randomized control group pretest/post- including one-tailed t-tests, ANOVAs, and multi-
test design (Campbell & Stanley, 1966). The ple comparison tests at a 0.05 level of signifi-
ICSC was not a typical MSDS; it was a fact sheet cance were used to identify any significant dif-
prepared by the International Program on Chemi- ferences between the three MSDS formats in
cal Safety at the World Health Organization of their ability to diffuse information to the workers
the United Nations. It was developed for the use studied. The data from the study were gathered
of international workers and for similar purposes from questions and testing administered in a class-
of the MSDS. For the purpose of the study, the room over a 1-hour period.
ICSC was considered analogous to an MSDS.
Each treatment format included differed as fol-
lows: the OSHA type form consisted of nine sec- RESULTS
tions; the new CMA/ANSI form had 16 sections;
and the ICSC had nine sections. All forms were The following section outlines the major find-
data sheets addressing the same chemical com- ings, first in terms of MSDS acceptability to the
pound. The treatments were introduced during union workers studied and then as related to MSDS
the post-test. Relationships between the variables efficacy.
of worker’s age, education level, years worked,
and the number of chemicals used were evalu-
ated for effect on the efficacy of the MSDS. Acceptability Inventory
Six research questions and 10 hypotheses were
developed for this study. Research questions, an- 1. The sample population consisted of 140 males
swered through use of the MSDS efficacy inven- and 20 females of which 80% were at or
tory, were focused to describe acceptability of above the age of forty. Sixty-nine percent of
the MSDS. The questions were used to identify this middle-aged workforce had achieved a
demographics, level of training on the MSDS, high school (9th to 12th grade) education
level of understanding of the MSDS, level of ac- and 30% had some level of college work.
cessibility and acceptability, and health belief The sample population was almost evenly
impact on MSDS usage. The results of the MSDS split in the number of years worked at the
efficacy inventory were analyzed descriptively. national laboratory with 49% having worked
The null hypotheses were developed to test 10 years or less, and 50% greater than 10
the efficacy of the MSDS in diffusing informa- years. There were 18 job crafts identified in
tion to workers; compare the three different for- the sample population, predominantly includ-
mats; and determine the efficacy of each in an- ing labor/pest controllers, carpenters, paint-
swering specific questions addressed on the MSDS ers, and pipefitters (Table 1). Ninety percent
comprehensibility test. of the worker population has worked with
In an experimental design, workers were ran- chemicals on a weekly basis requiring the use
domly assigned to four different test groups of 40 of MSDS sheets.
individuals. Using an MSDS comprehensibility 2. Ninety-five percent of the union workers
test, all groups were pretested on their knowledge studied had undergone training on MSDSs
about a specific chemical and then post-tested us- either on the job, in a classroom, or both.
ing the same test. Post-test questions were ran- The union workers agreed that training on
domized in comparison to the pretest. Each indi- the MSDS at the national laboratory met
vidual in the three groups randomly received one with their satisfaction. Eighty-eight percent
of the three differently formatted MSDSs to use of the sample population indicated that the
as a reference in answering questions on the post- training helped them better understand the
test. MSDSs were not provided during the pre- MSDS. As reflected by responses to the
test. A fourth group served as a control receiving MSDS inventory, a majority of the workers
the pretest and post-test without an MSDS refer- knew appropriate times to request an MSDS
ence sheet. The workers were not aware that they (e.g., during planning stages, safety brief-
were going to be tested on MSDSs nor of the ings, and prior to working with a chemical).
chemical to be addressed. Efficacy of the differ- 3. Workers responding were in disagreement as
ent forms was measured by gain scores (i.e., the to whether MSDSs were difficult to under-
difference between pretest scores and post-test stand, with 39% indicating they were and
Summer 1999/Volume 30/Number 2 115
Table 1. Job Categories of Sampled Workers the remainder rarely to almost never used
(N 5 160) them. Changing work habits to safer prac-
Job Category Frequency Percentage tices, as a measure of MSDS acceptability,
also ranked high in that three-fourths of the
A/C refrigeration 6 03.75 workers surveyed had implemented such
Boilermaker 4 02.5 changes. Most of the workers surveyed also
Carpenter 21 13.13 agreed with the use of health hazard ratings
Chemical worker 2 01.25 on MSDSs and did not consider them con-
Electrician 10 06.25 fusing.
Insulator 7 04.37
Laborer/Pest controller 21 13.13
6. Workers, in a consensus, agreed that chemi-
Machinist 8 05.00 cals were a threat to their health including
Millwright 9 05.63 those with which they currently worked. Work-
Painter 21 13.13 ers by a majority also indicated that the threat
Pipefitter 19 11.88 of chemicals was not exaggerated at the na-
Power equipment operator 3 01.88 tional laboratory.
Rigger/iron worker 3 01.88
Sheetmetal worker 9 05.63
Steamplant worker 3 01.88 Efficacy of Formats
Truck driver 5 03.13 7. The OSHA type MSDS, ICSC, and CMA/
Utility mechanic 6 03.75
Welder 3 01.88
ANSI Z400.1 forms each made a significant
difference in worker’s post-test scores com-
pared to pretest scores in a test of the work-
ers’ abilities to answer health and safety
46% disagreeing. Over half of the workers questions about a particular chemical. The
indicated the MSDS was easy to read and forms each significantly diffused informa-
one-third disagreed. Most workers did not tion to workers.
think MSDSs were too lengthy with 20% in 8. Considering the highest average total gain
disagreement. Almost half of the union work- scores achieved by each different MSDS form,
ers studied did not consider the MSDS for- there was a rank order of finish: ICSC (68.8,
mat to be confusing. However, one-third dis- 39.8); OSHA type form 174 (67.2, 35.8);
agreed. Workers also responded that symbols and CMA/ANSI Z400.1 (64.8, 28.6), respec-
found on MSDSs were easy to read. tively (Figure 1). However, there were no
4. The study revealed that MSDSs were acces- significant differences in these scores, thus,
sible to the workers studied at the laboratory in efficacy of the formats (Table 2). There
during the work shift and were provided by
management at the stages of work that work-
ers wanted to see them. MSDSs were pro- FIGURE 1 Average total scores and average total gain
vided in written format, on computers or scores for three different MSDS forms post-test.
verbally to the workers studied and in almost
half of the responses, all three media were
said to be available. Seventy percent of work-
ers inventoried indicated they had requested
and received MSDS information with 90%
of those having received a hard copy MSDS.
5. Worker acceptability of the MSDS, as mea-
sured by worker satisfaction with informa-
tion provided, was very high. Greater than
90% of workers responded that MSDSs were
satisfactory to very satisfactory in providing
protective information and answering ques-
tions. Acceptability as measured by worker
frequency of usage was low. A third of the
union workers studied responded that they
used MSDSs at least half to all of the time;
116 Journal of Safety Research
Table 2. ANOVA Summary Table for Experimental form were each significantly higher on three
Design Comparing Mean Gain Scores of Three different questions. (Figure 3).
Different MSDSs 12. The ICSC form was significantly better than
Degrees Sum the OSHA type form in answering questions
of of Mean on chronic and immediate health effects.
Source Freedom Squares Square F-Value a Both the ICSC and OSHA type forms were
significantly better than the CMA/ANSI
Between Z400.1 form in answering fire-related ques-
forms 2 31.82 15.91 2.89 tions. The OSHA type form was signifi-
Within
cantly better than the other two forms in an-
forms 117 644.697 5.51
Corrected swering spill response questions. It also was
total 119 676.51 significantly better than the CMA/ANSI form
in answering cancer-causing potential ques-
a F(2, 117, 0.05) 5 3.08. tions. The CMA/ANSI form was not signifi-
cantly better in any of the questions asked on
the MSDS comprehensibility test. Addition-
were significant differences in the scores of ally, the ICSC outscored the other two forms
each format and the control group (Tables 3 on questions pertaining to routes of entry
and 4). and body systems immediately affected. The
9. There were no significant differences on aver- ICSC also outscored the CMA/ANSI form
age gain scores among groupings on the num- in answering questions on chronic health ef-
ber of chemicals used, grade level achieved, fects. The OSHA type form outscored the
and years worked at the lab for each form other two forms in determining what personal
(Table 5). protective equipment to use. The CMA/
10. There were no significant differences on av- ANSI Z400.1 form narrowly outscored the
erage gain scores among the groupings on other two forms on the question of first aid
age for the ICSC and CMA/ANSI Z400.1 procedures and outscored the OSHA type
forms. Results on the OSHA type form re- form in determining chronic and immediate
vealed that the 40–49 age group scored sig- health effects. The CMA/ANSI also out-
nificantly better than the 50–59 age group. scored the ICSC format on questions per-
11. There were significant differences in average taining to spills and leaks. The OSHA type
gain scores achieved on five MSDS compre- form scored higher than the CMA/ANSI
hensibility test questions between the three Z400.1 form in answering the question on
different forms (Figure 2). The ICSC had the who to contact for further information.
highest average gain scores on five questions
followed by the OSHA type form with three
and the CMA/ANSI Z400.1 form with one SUMMARY OF FINDINGS
(Figure 3). Both the ICSC and OSHA type
Within the limitations of this study and to the ex-
tent to which the data and findings were valid and
Table 3. ANOVA Summary Table for Experimental reliable, the following conclusions are drawn by
Design Comparing Average Gain Scores of Three the authors.
Different MSDSs to a Control Group
Degrees Sum 1. Union workers studied were trained on
of of Mean MSDS usage, were satisfied with the train-
Source Freedom Squares Square F-Value a ing, and understood the MSDS.
2. Workers considered the MSDS to be easy to
Between read, not too lengthy or confusing, and easy
forms 3 521.57 173.91 37.62 to understand.
Within 3. There was ready access to the MSDS during
forms 156 721.11 4.63
the work shift or whenever requested by
Corrected
total 159 1242.83 union workers studied.
4. High acceptability of the MSDS was demon-
a F(3, 156, 0.05) 5 2.66. strated through reported satisfaction by the
Summer 1999/Volume 30/Number 2 117
Table 4. Dunnett’s Test Presented as 95% Confidence Intervals for an Experimental Design Comparing
Average Gain Scores of Each MSDS to the Control Group
Simultaneous Lower Difference Simultaneous Upper
Form Confidence Limit Between Means Conf idence Limit

ICSC—controla 3.60 4.60 5.60


OSHA TYPE—controla 3.16 4.16 5.16
CMA/ANSI—controla 2.34 3.36 4.36
ap , 0.05.

workers. There was, however, low usage of three different MSDS formats enabled work-
the forms. ers to answer specific questions on the MSDS
5. Workers inventoried perceived chemicals to comprehensibility test.
be a threat to their health.
6. Each MSDS evaluated significantly diffused
information to workers studied. DISCUSSION
7. There were no significant differences in the
efficacies of the three different MSDS formats. Most of the workers comprising the sample pop-
8. Education level, number of chemicals used, ulation studied reported favorably on MSDS ac-
and years worked at the national laboratory ceptability, accessibility, and understandability.
did not significantly affect the efficacy of the However, the finding that one-fifth to one-third of
MSDS. the workers did not agree that the MSDS was
9. Age did not influence the efficacy of the easy to read, understand, and was not too lengthy
MSDS with the exception of age groupings or confusing suggests further training, format
40–49 and 50–59 using an OSHA type format. and content changes might need to be consid-
10. There were significant differences, when ered. Further, although most workers reportedly
comparing scores, in how well each of the had requested to see an MSDS at some time dur-

Table 5. Results of Hypotheses Based on the MSDS Inventory and MSDS Comprehensibility Test
ANOVA
Hypotheses Probabilities Comment

There are no significant differences in the ICSC: 0.09 For each form, there are no significance
average test scores achieved on OSHA ANSI: 0.86 differences among the grouping on the
type form, CMA/ANSI, and ICSC forms OSHA: 0.66 number of chemicals used.
between groupings on the number of
chemicals used.
There are no significant differences in the ICSC: 0.69 For each form, there are no significance
average test scores achieved on OSHA ANSI: 0.77 differences among the grouping on grade
type form, CMA/ANSI, and ICSC forms OSHA: 0.16 level completed.
between groupings on grade level
completed.
There are no significant differences in the ICSC: 0.60 For each form, there are no significance
average test scores achieved on OSHA ANSI: 0.30 differences among the grouping on years
type form, CMA/ANSI, and ICSC forms OSHA: 0.81 worked at the national laboratory.
between groupings on years worked at
the national laboratory.
There are no significant differences in the ICSC: 0.11 There are no significant differences among the
average test scores achieved on OSHA ANSI: 0.70 grouping on age for ICSC and ANSI/CMA.
type form, CMA/ANSI and ICSC forms OSHA: 0.04 Note: The results on OSHA type form were
between groupings on age. signficant between age groups: 40–49 and
50–59.

118 Journal of Safety Research


FIGURE 2 ANOVA F-statistic values for comparing the average gain scores of the three forms for each question. Dash line
is the percentile value of the F-distribution at the 5% significance level.

ing their jobs, almost one third of those studied cepts of the Social Cognitive Theory and the
had never asked to see one. The reason for this Health Belief Model (Rogers, 1983; Rosenstock,
disparity needed further evaluation at the national Strecher, & Becker, 1988). A plausible explana-
laboratory. Considering the strong agreement tion for this difference in belief and usage might
found in the study between the number of work- have been the low number of chemicals each
ers that believed chemicals to be a hazard and the worker used on a weekly basis; familiarity with
number that had switched to safer work prac- chemicals; and the use of standard operating pro-
tices, there should have been a greater reported cedures at the laboratory, which included safety
use of MSDSs in the sample population. The and health guidance. MSDS format testing dem-
finding of low MSDS usage in the face of per- onstrated a significant difference in how well the
ceived threat was not in keeping with the pre- 40–49 age group scored as compared to the 50–

FIGURE 3 Average gain scores for each question on a form. Values connected by the same line are not significantly
different at the 5% significance level.

Summer 1999/Volume 30/Number 2 119


59 age group when using the OSHA type form as was pervasive with the total union work-
a reference. Further, evaluation as to why these force. Actions such as training, adjusting the
two age groups differed needs to be undertaken. average reading level, simplification of con-
Why the same age groups did not have a similar tent and format should be considered.
effect when using the other two formats needs to 4. Any future MSDS usage studies at the labo-
be studied. This finding, if not unique within the ratory should include researchers and other
sample population, may indicate the need to employees that may come in contact with
study the effects of age on MSDS diffusion effi- chemicals in their jobs.
cacy in older working populations, who are re- 5. Future research on the efficacy of the MSDS
maining longer in the workforce (Rosenstock et should include different ethnic groups as in-
al., 1988). dependent variables. Testing should be con-
Workers tested did not absorb one third of the ducted in work populations that use English
information provided on each of three different as a secondary language.
MSDSs studied, demonstrating weaknesses in 6. Similarly, as more senior citizens (ages greater
the formats’ ability to diffuse information to than 65) are returning to the work force, the
workers. This may suggest the MSDSs did not effects of aging on the efficacy of the MSDS
adhere to principles of simplicity found in Adop- needs in-depth study. Age related deprecia-
tion Diffusion Theory (Strecher, Devellis, Becker, tion on working memory and effect on MSDS
& Rosenstock, 1986). efficacy is one example of what might be re-
Multiple comparison of the scores achieved searched.
by each format on each question of the MSDS 7. This study should be extended by examining
comprehensibility test demonstrated significant on a national and international level the effi-
differences in efficacy. Considering that the level cacy of the three MSDS formats studied ap-
of readability and specific chemicals addressed plying the variables used in the study and
were analogous to all forms, it was plausible that those recommended for future research.
the manner in which a form presented informa- 8. More in-depth research needs to be con-
tion sought after might have affected the scores ducted to discover why the ICSC and OSHA
(e.g., length of explanation, jargon, sentence struc- type formats performed better than the CMA/
ture, fuzzy language, and phraseology; Hoch- ANSI Z400.1 structure in answering particu-
hauser, 1997). Further study needs to be con- lar questions on the MSDS comprehensibil-
ducted to discern the impact of these variables in ity test. Variables such as jargon, fuzzy lan-
enabling the three formats to answer particular guage, length of sentences, presentation of
questions on the MSDS comprehensibility test. material, phraseology, and use of symbols
should be tested for effect on MSDS efficacy.
9. Why one-third of the information provided
RECOMMENDATIONS on the three MSDSs was not diffused to the
union workers studied needs further research.
1. Further investigation should be conducted at 10. Self-efficacy of the worker in relation to ab-
the national laboratory to discern why two- sorbing such scientific information as found
thirds of the union sample population stud- on MSDS should be researched for effect on
ied rarely to almost never used MSDSs and the efficacy of the MSDS.
whether such poor usage is reflective of the 11. Learning theories such as the Adoption Dif-
entire union workforce. fusion Innovation Theory, Social Cognitive
2. More study should be conducted at the labo- Theory, and Value Expectancy Theory (Health
ratory to determine why the 40–49 year age Belief Model) should be considered in inter-
group scored significantly better using the preting the efficacy of different MSDS formats.
OSHA type form as a reference than did the 12. Until further quantitative research to evalu-
50–59 year age group. ate the efficacy of various MSDS formats is
3. Efforts should be undertaken at the labora- completed and considering the findings of
tory to address the needs of the one-fifth to this limited research, caution should be used
the one-third union workers who indicated in movement toward widespread adoption of
that MSDSs were difficult to understand and the new CMA/ANSI Z400.1 structure. Ef-
read, and confusing. Such efforts should in- forts should be made to further field test the
clude determining whether this percentage new format.
120 Journal of Safety Research
IMPLICATIONS 6. The two studies differed as to whether work-
ers felt the MSDS to be confusing with the
This study is similar to the study conducted by Kearney/Centaur study revealing two-thirds
Kearney/Centaur Division of A. T. Kearney, Inc., in of the workers indicating yes versus one-
1990–1991, under contract to OSHA, on the com- third for the national laboratory.
prehensibility of Material Safety Data Sheets. Al- 7. Education was cited as having influenced the
though the methodologies and many of the find- comprehensibility of MSDSs in the Kear-
ings of the two studies are different, there are ney/Centaur study. Level of education did
similarities. not influence the results of the laboratory
The Kearney/Centaur study was non-random study in that readability was controlled to
and descriptive in analysis. It involved testing the 9th grade level.
ninety-one union workers from 13 different man- 8. The average test score achieved on the
ufacturing plants and one local trade union located OSHA form and hybrids was 64 points and
in Maryland. All workers were open-book tested the ICSC format averaged 70.6 points in the
against four different randomly selected MSDSs Maryland study. Comparatively, total scores
averaging a twelfth grade readability. A ques- achieved in the national laboratory study on
tionnaire on demographics and other variables the OSHA type form MSDS was 67.2, ICSC
that might effect MSDS comprehension was ad- 68.8, and CMA/ANSI Z400.1 structure 64.8,
ministered. At the request of OSHA, the contrac- respectively. Although there was a rank or-
tor of the study, no sophisticated statistical analy- der of finish in each study, in comparing the
ses were conducted. No pretesting and few controls formats, both found that about one-third of
were used in the study. The MSDS comprehensi- the information available from the MSDS
bility test used in the Kearney/Centaur study was was not diffused to the worker. The national
also used in the national laboratory study. The laboratory study found that on the average
highest score achievable on the test was 100 points one-third of the information was known or
for both studies. The following are comparative guessed by the worker prior to the introduc-
findings of the two research studies: tion of the MSDS and about one-third of the
required information was not imparted to the
1. In both studies, 69% of the workers were user. Both studies found that workplace
30–49 years of age and had 69–70% of them characteristics did not appear to have an im-
completing 9–12th grade and 30% complet- pact on MSDS test scores.
ing 13–16th grade. 9. As found in the laboratory study, the highest
2. While 80% of those studied in the Kearney/ average MSDS comprehensibility test score
Centaur evaluation had seen an MSDS be- in the Maryland study was achieved by the
fore and 45% had been trained, 95% of the ICSC format. In the Maryland study the
workers at the national laboratory had seen ICSC ranged 6–23% higher as compared to
and been trained on MSDSs. 1–6% in the laboratory study.
3. Where two-thirds of the Maryland workers 10. Both research studies gave supportive evi-
had requested MSDSs, three-fourths of the dence, one descriptive and the other quanti-
laboratory sample population had done the tative, that format does play a role in the effi-
same. Also 90% of the workers at the na- cacy of the MSDS. However, as shown in the
tional laboratory in the Tennessee study who national laboratory work force study, there
had requested MSDSs received them in hard was not a significant difference in the scores
copy compared to 60% in the Maryland study. of the MSDS formats studied.
4. Similar to the findings of the Kearney/Cen- 11. This study is timely and of national and in-
taur report where 80% of the workers had ternational importance in that efforts are cur-
changed their work practices to safer meth- rently underway by OSHA, the International
ods as a result of the MSDS, the national Standards Organization, and others to try to
laboratory study revealed that 79% of their harmonize the MSDS format worldwide. Al-
study population had done likewise. though the study is limited to a small worker
5. In the Maryland study, almost half of the population in Tennessee, it produced sup-
surveyed workers reported the MSDS as portive evidence that more research and ef-
helpful while 90% of the national laboratory forts need to be directed toward developing
workers similarly responded. an MSDS format that diffuses all of its infor-
Summer 1999/Volume 30/Number 2 121
mation in a fully understandable manner. Kolp, P., Sattler, B., Blaymew, M., & Sherwood, T. (1993).
The formats and contents tested were only Comprehensibility of material safety data sheets. Ameri-
can Journal of Industrial Medicine, 23,135–141.
two-thirds of the way toward achieving such Lu, C.W., & Moure-Eraso, R. (1993). Quality of hazard
a goal. communication in the manufacturing industry: Evalua-
tion of material safety data sheets. Paper presented at the
American Industrial Hygiene Conference. New Orleans,
Louisiana.
REFERENCES Meninger, B., & Margolies, D. (1994). Safety data sheets
difficult to understand. Kansas City Business Journal, 13,
Campbell, D.T., & Stanley, J. C. (1966). Experimental and 55.
quasiexperimental designs for research. Chicago: Rand Nelkins, D., & Brown, M.S. (1984). Workers at risk. Chi-
McNally. cago: University of Chicago Press.
Campbell, S.L. (1992). A new look for the MSDS. Occupa- National Safety Council. (1995). Accident facts. Itasca, IL:
tional Health and Safety, 61(6), 62, 65–67. Author.
Department of Health and Human Services. (1991). Healthy Phillips, C.C. (1997). The efficacy of material safety data
people 2000 National Health Promotion and disease pre- sheets and worker acceptability. Dissertation Abstracts
vention objectives. DHHS Publication No. 91–50213. International, 58(9), B4742.
Washington, DC: U.S. Government Printing Office. Rogers, E.M. (1983). Diffusion of innovations. New York:
General Accounting Office. (1992). Occupational safety The Free Press.
and health employers’ experiences in complying with the Rosenstock, I.M., Strecher, V.J., & Becker, H.M. (1988).
hazard communication standard. GAO Publication Social learning theory and the health belief model. Health
HRD-92–63BP. Gaithersburg, MD: U.S. General Ac- Education Quarterly, 15 (2), 175–183.
counting Office Publications. Samways, M. (1988). Functionally illiterate worker also has
Hadden, S.G. (1988). Community right to know: Results of “right to understand.” Occupational Health and Safety,
four surveys. Working Paper 44. Austin, TX: Lyndon B. 57(1), 49–53.
Johnson School of Public Affairs. Strecher, V.J., Devellis, B.M., Becker, M.H., & Rosenstock,
Hochhauser, M. (1997). Why aren’t material safety data I.M. (1986). The role of self-efficacy in achieving health
sheets more effective? Occupational Health and Safety, behavior change. Health Education Quarterly, 13(1), 73–
66(8), 58–64. 91.
Kearney/Centaur Division of A.T. Kearney, Inc. (1991). Van Santen, R.A., (1989). Info Safe: Managing and provid-
The comprehensibility of material safety data sheets, fi- ing information on chemicals. Chemistry in Australia, 66,
nal report. Alexandria, VA: Author. 440–441.

122 Journal of Safety Research

You might also like