A valid and reliable basic first aid and emergency care knowledge test for use at the
college/university level was developed. The First Aid and Emergency Care Knowledge Test for
College and University Students provides a viable alternative to the existing testing mechanisms
and affords the
classroom instructor the opportunity to compare the performance
of students in his/her respective class to others across the nation. This test also has application as
a pretest - posttest to guide classroom instruction.
Unintentional injuries are currently the leading cause of death in individuals 1 - 44 years of age
(Ventura, Peters, Martin & Maurer, 1997). It has long been assumed
that training in first aid and emergency care has the potential to mitigate the results of such
unintentional injuries and to raise safety and health awareness. As a
result, basic first aid and safety has become an integral segment of the high school and college
curricula to provide instruction to populations at high risk.
Poole & Ludwig (1960), conducted an
investigation of colleges and universities concerning the offering of first aid courses. The results
of the investigation revealed
that 83% of the colleges and universities surveyed offered first aid as a separate course, or as part
of another course. In addition, Winkelman (1977), stated that
"for some time community colleges and four-year colleges and universities have offered first aid
and emergency care courses." Consequently, "first aid and
emergency care subject matter has also been found to be included in more comprehensive kinds of
health and safety education courses such as a general safety and
accident prevention course or a course concerning the organization and administration of health
programs."
The first aid and emergency care instruction provided
at the college and/or university level must be more comprehensive and utilize a more in depth
method of
evaluation. The majority of the classes provided by the respective training agencies are developed
so that minimal effort is required to successfully complete the
instructional program. The amount of time required to complete the instructional program varies,
but can be obtained in a limited time period. At the collegiate
level, basic first aid and emergency care instruction is provided during the course of a semester
and/or quarter. The amount of time that is devoted to this
instruction is dependent on the respective institution. Despite the variation in the time allotted,
instruction at the college level generally provides a more in depth
introduction to the topic of first aid and emergency care than is provided in community based
training programs.
As with other subject content areas, current
instruction is heavily scrutinized with an emphasis placed on accountability. Measurement and
evaluation, of which
testing is an example, are two tools utilized to address accountability issues and the planning of
course content and delivery.
...instruction at the college level
generally provides a more in depth introduction to the topic of first aid and emergency care than is
provided in community based
training programs.
|
The measuring instruments used with members of the
community taking a specific course should not be the same as those utilized at the collegiate level.
Testing for
the purpose of student certification is a major concern, but is not the only concern present among
college instructors. The college instructor must issue a final grade
for the course.
According to the literature, first aid and emergency
care measurement and evaluation was initially introduced at the senior high school level in 1940
with the
development of the General First Aid Test for Senior High School Students
(McCloy & Young, 1954). Measurement and evaluation at the collegiate level
had
its inception at Brooklyn College in the early 1940s (Doscher, 1943). As a result of this impetus,
cognitive standardized testing in the area of first aid and
emergency care at the collegiate level developed sporadically in the ensuing decades (Serdula,
1957; Casperson, 1959; Gilbert & Windsor, 1977; Burckes, 1982).
Despite these early efforts, there is a noticeable paucity within the literature pertaining to
continued contributions or research since the early 1980s focusing on
measurement and evaluation.
Purpose
The purposes of the project were: 1) develop a valid,
reliable,
and objective first aid and emergency care knowledge test for college and university students; 2)
develop initial
national norms for two and four year college and university students.
In order to provide more comprehensive first aid and
emergency care instruction at the college/university level, an instrument was needed for consistent
evaluation. the outcome of this project consisted of a valid, reliable test with accompanying
national norms which provide an instructor with the opportunity to compare the test performance
of his/her students to the performance of others across the nation. Such as test could also be used
to identify specific areas in the first aid and emergency care program in need of revision or
modification.
The initial step in the construction of the instrument
was the development of the table of specifications. According to Lien (1980), "the table of
specifications is a
prescriptive guide in regards to the specific content of a test that is being developed." This
involved an in depth examination of universally used first aid and
emergency care textbooks to develop a list of possible subject content areas that could be
included in basic first aid and emergency care instruction at the college
and university level. This was accomplished by analyzing the most widely adopted textbooks
namely: First Aid: Responding to Emergencies, (American Red
Cross, 1993), and First Aid and CPR (National Safety Council, 1991). A
thorough analysis of the two books suggested twelve subject content areas that could be
considered for inclusion in college level instruction. These topics included: emergency
identification and recognition; basic life support; control of blood loss; shock;
cold related injuries; musculoskeletal injuries; poisoning; burns and heat injuries; splinting,
dressing, bandaging; specific medical emergencies; first aid kits and
supplies; and victim transport.
Table 1. Table of Specifications: Subject Content Area Weightings of
National Jury Panel of Experts |
Subject Content
Area |
Percentage
Subject Content
Should Cover on
Test |
Number of Test
Items per
Content Area (75
total) |
Emergency Identification
& Recognition |
11.9 |
9 |
Basic Life
Support |
18.3 |
13 |
Control of Blood
Loss |
9.9 |
8 |
Shock |
10.0 |
8 |
Cold Related
Injuries |
5.4 |
4 |
Musculoskeletal
Injuries |
9.2 |
7 |
Poisoning |
5.7 |
5 |
Burns and Heat
Injuries |
6.7 |
5 |
Splinting, Dressing,
Bandaging |
6.6 |
5 |
Specific Medical
Emergencies |
8.6 |
7 |
First Aid Kits and
Supplies |
3.4 |
2 |
Victim
Transport |
4.4 |
3 |
Total |
100.0 |
75 |
A national jury panel consisting of forty-four members
was requested to assign weights (percent value) to each of the subject content areas provided in
relationship
to the number of corresponding items on two seventy-five item parallel form instruments. Each
member of the panel had prior experience either in the development
of a basic first aid and emergency care textbook, program, or in the provision of first aid and
emergency care instruction at the college level. Each juror had
extensive knowledge and experience from which to make judgements concerning which subject
content areas should be included on a basic first aid and emergency
care knowledge test. The weightings provided by the panel of jurors were tabulated to ascertain
the number of test items that should be devoted to each subject
content area. The table of specifications is provided in Table 1.
Due to the considerable variability in the amount of
time scheduled for a class period at the various colleges and universities, it was determined that
two equivalent
forms containing seventy-five multiple-choice items each would be appropriate. Most college and
university courses range from 45-75 minutes in length. It was
concluded that an instrument containing seventy-five items could be completed in those college or
university classes that were limited to forty-five minutes in length.
This was based on the assumption that the typical lower division college or university student
could be reasonably expected to respond to an average of two (2)
multiple choice items per minute. Gronlund (1985), stated that "as a rough guide, the average
high school student should be able to answer...one multiple choice
item...per minute." Using this as a guide, it was assumed by the researchers that due to the fact
that the average college student is at an advanced level of
intellectual maturity than high school students, that two multiple choice items per minute would
be acceptable.
Next, members of the panel of jurors evaluated each
test item that was constructed in regards to whether the item should be included, revised, or
discarded. Items
were then randomly selected from each content area to be placed in the parallel form of the
preliminary drafts with respect to the percentages enumerated in the
specifications table.
The two preliminary drafts of the instrument were
administered as a pilot at twelve institutions to a population of 313 students (n=313). Form A of
the preliminary
instrument was administered to a student population of 156. Form B was piloted with a sample of
157 students. The 12 participating institutions represented 6
two-year community and/or junior colleges and 6 four-year colleges and/or universities randomly
selected from each of the six geographical districts of the
American Alliance for Health Physical Education, Recreation, and Dance
(AAHPERD).
Concerning validity it must be remembered that "tests
themselves are never valid...rather the concept of validity is linked to the inferences we draw
based on the use
of tests" (Popham, 1990). Rather the professional must make judgements regarding the validity
of such score-based inferences. These judgements are based on
evidence of which there is three types, of which one was specific to this investigation, content
evidence. "In general, content-related evidence demonstrates the
degree to which the sample of items, tasks, or questions on a test are representative of some
defined universe or domain of content" (Popham, 1990). The protocol
to ensure this type of validity espoused by Popham (1990) which was followed included: First,
attempts to incorporate suitable content on the test can be carried
out and (documented) during the test development itself, calling on experts where needed to
ensure that the test represents a desired domain of content. Second, is
the completion of post facto judgements concerning the representativeness of the content of the
test.
The content validity of the instrument and test items
were established by the use of two well documented basic first aid and emergency care textbooks
in the
identification of initial subject content areas. The subject content areas included in the
development of the draft instrument were determined through preliminary and
final weightings provided by a 44 member panel of experts who possessed either extensive
experience or specialization in planning, developing, or teaching basic
first aid and emergency care.
The subject content areas included in the
development of the draft instrument were determined through preliminary and final weightings
provided by a 44 member panel
of experts who possessed either extensive experience or specialization in planning, developing, or
teaching basic first aid and
emergency care.
|
The data analysis for pilot instrument Form A revealed
the following: the mean was 41.6, the median was 44.1, and the mode was 44; the range was 54
with a
variance of 155.05 and a standard deviation of 12.4. The item analysis for pilot instrument Form
A revealed that of the seventy-five items, 30 (40%) were at the .41
or higher discrimination level which is considered to be very good. Twenty-seven of the items
(36%) were within the .20 to .40 boundaries which were classified as
satisfactory. Fifteen items (20%) were within the .00 and .19 level of discrimination. A total of
three items were determined to be negative discriminators. The
mean level of item discrimination for the total pilot instrument Form A was .34, which was within
the satisfactory range of acceptance. The discrimination range
utilized was reported by Lien (1980), which classified .41 to 1.00 as very good; .20 to .40 as
satisfactory; .00 to .19 as low; and -.01 to -1.00 as unacceptable.
The item analysis revealed that From A contained ten
items (13%) that were deemed too easy, with difficulty levels above 75%. The majority of the
items, 59
(79%), were within the acceptable level of difficulty which was .25 - 75%. Six items (8%) were
too difficult, with a difficulty level below 25%. The difficulty for
the total pilot instrument Form A was 55.2% The mean for pilot instrument Form B was 44.1, the
median, 46.2; the distribution was bimodal, with the scores of 49
and 47 occurring most frequently. The range was 47, with a standard deviation of 10.4, and a
variance of 109.15.
The item discrimination levels for pilot Form B
revealed that thirteen items, or 17%, where at the .41 or higher level of discrimination.
Thirty-eight (51%) of the
test items were between the satisfactory ranges of .20 and .40. Twenty-one (28%) items were in
the low category concerning discrimination. Negative
discrimination levels were present for three items constituting 4% of the test items. The mean
level of item discrimination for the entire Form B instrument was
determined to be .26.
Form B consisted of seventeen items (22%) that were
too easy. Fifty-one items, or 68%, possessed acceptable levels of difficulty, whereas seven items,
or 9%,
were classified as too difficult. The cumulative difficulty for pilot instrument Form B was also
calculated to be 60.6%.
Kuder-Richardson Formula 21 was utilized in the
calculation of a reliability coefficient. According to Popham (1990), "most well constructed norm
referenced tests
usually hover between .80 and .90." The reliability of Form A was determined to be .89, and
Form B was determined to be .84. Therefore, both instruments
possessed a suitable level of reliability to be considered well constructed according to Popham's
(1990) stipulation.
The parallel forms of the final instrument, Form A and
Form B containing seventy-five items each, related directly to the statistical data provided by the
item analysis
of the preliminary draft. Only those test items that met the following established criteria were
selected to be a part of the final parallel instruments.
- The items discriminated between good and
poor student performance.
The difficulty index for each item was between
25% and 75%.
Each alternative response was selected by at least
three percent of the respondents.
Those pilot instrument items that were deficient in
distractor utilization or level of difficulty, or discrimination were either revised or deleted.
The final forms of the instrument were administered in
order to establish initial national test norms. The sample consisted of two-year community
and/or junior
colleges, and four-year colleges and/or universities from the six geographical districts of
AAHPERD. The final sample of two year institutions represented 17
states and the sample of four year institutions represented twenty-five states. Thirty-two of the
fifty states had at least one intact group of students tested as part of
the final administration. The final sample consisted of n = 927 college and university students.
As in the administration of the pilot instrument, cluster sampling
was employed, whereby the parallel forms of the instrument were administered to an intact class
of students enrolled in a basic first aid and emergency care course at
the respective institution.
The analysis of data followed the same procedure as
for the pilot forms of the instrument. Percentile ranks. Analysis of the data showed that the final
forms of the
instrument met acceptable criteria concerning test construction protocols. The measures of
central tendency for final Form A were: mean 43.1, median 43.5, and
mode 41. The measures of variability for Form A included a range of 52, standard deviation of
9.7, and a variance of 94.09. The Kuder Richardson reliability
coefficient was determined to be .82, with an index of discrimination of .33, and an index of
difficulty of 57%.
The measures of central tendency for final Form B
included a mean of 41.6, median of 41.7, and a mode of 38. The measures of variability included
a range of 49,
standard deviation of 9.8, and a variance of 95.05. The Kuder Richardson Reliability coefficient
was calculated to be .82, the index of discrimination was .34, and
the index of difficulty was 52%. The standard error or measurement for Form A was 4.1, and 4.3
for Form B.
The reliability coefficient of the final instruments was
calculated using the Kuder-Richardson Formula 21. The reliability for both Final Form A and
Form B were
calculated to be the same at .82, thus meeting the criteria between .80 and .90 for a well
developed, norm-referenced test. The standard error of measurement was
4.1 in Form A, and 4.3 in Form B. In addition, t-scores were calculated for the parallel forms of
the final instrument.
The First Aid and Emergency Care Knowledge Test
for College and University Students (Ballard, 1994), lends itself to utilization as a pretest -
posttest instrument to
be used in the
planning of course content delivery. Such utilization would assist in the identification of areas in
which the student(s) is strong/weak to guide instruction. It is
believed that if instruction addressed the subject content areas in which the student(s) had
misconceptions and/or the greatest need, the student would potentially
exhibit a higher level of interest and motivation in learning and feel comfortable in applying the
knowledge in a real world emergency or situation. This is important
in response to a 1998 Annals of Emergency Medicine article which revealed that almost 50% of
adults surveyed would not provide assistance to someone involved
in a roadside accident (Braslow & Brennan, 1998).
This instrument provides a viable alternative to the
existing testing mechanisms and affords the classroom instructor the opportunity to compare the
performance
of students in his/her respective class to a national norm. The instrument is applicable at both the
two-year and four-year college and university setting and
encompasses the guiding principles of the leading first aid and safety organizations. The
instrument is also one which can be administered in a variety of university
class settings. As a result of the instrument having 75 multiple choice items it is comprehensive in
scope, yet can be administered in classes that range from 45 to 75
minutes. We as a profession must continuously strive to develop measurement and evaluation
tools applicable in the college and community setting. Only through
such diligent efforts can we assure that the instruments utilized are valid and reliable for the
populations in question. In addition, numerous changes in first aid and
emergency care theory and practice occur, and must be incorporated in such educational testing
devices.
American National Red Cross. (1993).
First aid: Responding to emergencies. St. Louis: Mosby-Year Book,
Inc.
Ballard, M. D. (1994). Development of a valid and reliable first aid and
emergency care test for college and university students. Unpublished doctoral dissertation,
University of Tennessee, Knoxville.
Braslow, A. & Brennan, R. T. (1998).
Layperson cpr. Annals of Emergency Medicine, 31(5), 654.
Burckes, M. E. (1982). The Burckes
Emergency Care Knowledge Test. Journal of the American College Health Association,
30(4), 185-186.
Casperson, D. G. (1959). A first aid and
emergency care knowledge test for college students (Doctoral dissertation, Indiana University,
1959). Dissertation
Abstracts International, 31, 7379B.
Doscher, N. (1943). Two first-aid examinations
for college students and adult groups. Research Quarterly, 14(2),
228-248.
Gilbert, G. G., & Windsor, R. (1977). The
Ohio State University standard first aid and personal safety achievement test. Journal of
School Health, 47(1), 48-50.
Gronlund, N. E. (1985). Measurement
and evaluation in teaching (5th ed.). New York: Macmillan Publishing
Company.
Lien, A. J. (1980). Measurement and
evaluation of learning (4th ed.). Dubuque, Iowa: W.C. Brown Co.
McCloy, C. H., & Young, N. D. (1954).
Tests and measurement in health and physical education (3rd ed.). New York:
Appleton-Century-Crofts, Inc.
National Safety Council. (1991). First aid
and CPR.. Boston: Jones and Bartlett Publishers.
Poole, R. D., & Ludwig, D. J. (1960). First
aid instruction in colleges and universities. Journal of School Health, 30,
237-239.
Popham, W. J. (1990). Modern
educational measurement: A practitioner's perspective. Englewood Cliffs, New Jersey:
Prentice Hall.
Serdula, G. (1957). Standardization of a
first aid knowledge test for college students. Unpublished doctoral dissertation, Indiana
University, Bloomington.
Ventura, S. J., Peters, K. D., Martin, J. A.,
& Maurer, J. D. (1997). Births and deaths: United States, 1996. Monthly vital
statistics report, vol 46 no 1, supp 2.
Hyattsville, MD: National Center for Health Statistics. 1997
Winkelman, J. L. (1977). An-up-to-date view of
cardiopulmonary resuscitation instruction in colleges and universities. Journal of the
American College Health
Association, 26(2), 108-110.
[Reprint
(PDF) Version of
this article]
Copyright ©
1999 by IEJHE.