|Abstract or Summary
- The purpose of this study was to investigate one of the
effects of the SCIS program in the Corvallis Public Schools.
The investigation was based on a comparison of students
ability to construct a logical theoretical model from a
The total sample consisting of 289 students, was
obtained with no known systematic bias through a table of
random numbers and district teacher lists. The SCIS sample
consisted of 136 students stratified by grade level with
one first grade, one first-second split grade, two second
grades, and two third grades represented. The non SCIS
sample consisted of 153 students stratified along similar
grade level lines. The experimental sample had one
semester of SCIS prior to the present study.
The data gathering session consisted of the researcher presenting the students with a discrepant event in the form
of a "theory box." The theory box was a modified shoe box
with s funnel, on top and a hose visible on the bottom to the
students. The students observed the researcher pour a clear
liquid into the funnel and the liquid came out blue. After
some unobstrusive manipulation by the researcher with the
theory box, the students observed the clear liquid (from the
same original container) being poured into the funnel. The
liquid came out of the bottom hose yellow.
The students were asked to draw what they thought could
be inside of the theory box to explain what they had
observed. The data collected was classified as to logical,
probably-logical, or non-logical from criteria developed by
the researcher in conjunction with members of the Department
of Science Education at Oregon State University.
Using chi-square analysis, after Siegel, it was
1. There was no statistical evidence approaching the
.05 level to support that SCIS was a factor in the
ability of primary children to construct a logical
theoretical model from a discrepant event when
compared on the first and first-second split grade,
second grade, and total first through third grade
2. It was not possible to analyze the third grade
data alone due to lack of high enough expected frequencies in the non-logical, cells for valid use
3. The combined second and third grade SCIS to non
SCTS sample was significant beyond the .10 level,
one tailed. This would suggest a weak possibility
that SCIS could be a factor in the ability of the
combined second and third grade samples to
logically explain a discrepant event.
4. Comparison of grade level as a factor obtained
significance beyond the .10 level, two-tailed, for
the first and first-second split grade sample to
the second grade sample; beyond the .05 level for
the second to third grade sample; and way beyond
the .001 level of significance for the first and
first-second split grade sample to the third grade
sample. Significant results from chi-square
analysis indicated a strong possibility that the
grade level of a child was a factor in the ability
of young children to construct a logical
theoretical model from a discrepant event.
Based upon the data gathered in this study, the
researcher recommends that a similar study be conducted
using a larger sample and covering more grade levels. It
also appears logical, to this researcher, to compare the.
SCIS to non SCIS students after more than just one semester
of experience with SCIS materials. To replicate such a study, further refinement of the researcher. designed
Instruments is indicated.
Along the line of further research, the researcher
recommends that the factors of Sex and I.Q. be explored
with the ability of SCIS and non SCIS students to construct
a logical theoretical model from a discrepant event. It
also seems to this researcher that a significant follow-up
study might be a comparison of student ability to construct
a logical theoretical model from a discrepant event between
students who were classified as pre-operational and concrete
operational in terms of the Piaget conservation tasks.