- The purpose of this investigation was to develop a
valid and reliable instrument to assess Oregon high school
students' nutrition information.
The following steps were taken in order to achieve this
First, a review of the literature pertaining to high
school nutrition curriculum was conducted.
Second, numerous published tests designed to assess
high school students' nutrition information were examined.
Third, principles of test construction were identified.
Fourth, objectives were selected to serve as a guide
for item development.
Fifth, a pool of 75 items was constructed and used for
the preliminary test.
Sixth, the test directions and several items were revised to serve as the first trial instrument. Five low-scoring
items were omitted.
Seventh, the 70-item, five-option test was administered
to 300 students in three Oregon counties. The data were
analyzed and several items were revised.
Eighth, a panel of 13 experts in the fields of test
construction, health education and nutrition evaluated the
items for scientific accuracy, clarity, importance and
appropriateness for high school students. From their
recommendations the test was revised down to 50 items.
Ninth, the second trial instrument was administered to
4,518 students from 60 randomly selected Oregon high schools.
The resulting data were converted into norms.
Before any meaningful generalizations can be made,
the instrument providing the data must demonstrate a high
degree of validity and reliability. An item analysis of
the second trial administration revealed the 50 items to be
within the accepted range of difficulty. The items discriminated
positively, and all but two distractors were
plausible enough to be selected by at least two percent of
the student respondents. Statistical analysis revealed the
instrument to be valid. The reliability was calculated by
the odd-even number split-halves method and adjusted by the
Spearman-Brown Prophecy Formula. The reliability coefficient
was .899, which, according to standard testing
criteria, is considered high. The second trial test was administered during the 1979
spring school term. One point was awarded for a correct
answer and no points were deducted for an incorrect answer.
Out of a possible 50 points, the mean score for the entire
population was 27.667, with a standard deviation of 9.33.
An analysis of variance test was performed on the variables
of age, sex, grade and size of school. There was a
significant difference in mean scores between age groups,
in that scores consistently increased with age. Females
scored 16 percent better than males. Grades 9 through 12
were represented in the study. There was a significant
difference between each grade level, always in favor of the
upper grade. To insure a representative sample, the Oregon
School Activities Association standard for size of school
was used. A significant difference in mean scores was
demonstrated, with the AAA schools scoring higher than the
AA and A schools. There was also a significant difference
between the mean scores of urban and rural students, with
the urban students scoring higher.
It was concluded that:
1. There was a need for an instrument to assess student
achievement of selected nutrition objectives.
2. The instructional objectives of the School Health
Education Study were useful as a source for the construction
of the instrument.
3. A valid and reliable instrument able to evaluate student achievement of selected nutrition objectives was