Content uploaded by Felix Kamuche
Author content
All content in this area was uploaded by Felix Kamuche on Aug 20, 2015
Content may be subject to copyright.
Journal of College Teaching & Learning – August 2005 Volume 2, Number 8
25
Relationship of Time and Learning Retention
Felix U. Kamuche, (Email: fkamuche@morehouse.edu), Morehouse College
Robert E. Ledman, Morehouse College
ABSTRACT
This study explores the degree to which students’ understanding, or knowledge, may deteriorate over
time. Our specific focus was examining the importance of course sequencing in a curriculum. We
measured student performance in quantitative business courses and found that, over time, the
correlation between students’ performance in those courses declined significantly. We further found
that that decline was not linear.
INTRODUCTION
ny college professor who teaches junior or senior level courses has probably wondered if his or her
students learned anything in prerequisite courses because they seem to not know some fundamental
concepts. The concept of prerequisites is based on the premise that the learning of a subject is
cumulative over several courses and one has to learn certain concepts before learning others. For instance, we do not
expect a student to be able to successfully complete an advanced accounting course if he or she has never been
exposed to basic accounting principles. This study focuses on assessing the extent to which the time between
sequential courses may contribute to the loss of learning. Bahrick (1979) suggests that students do not retain the
information acquired in a class for long after an exam. Higbee (1977) echoes that sentiment. More recent studies,
however, show that although some information is forgotten (Conway, Cohen, & Stanhope, 1991; Semb, Ellis, &
Aranjo, 1993), the retention loss is not as great as expected (Cooper and Greiner, 1971; Semb, Ellis, & Montague,
1990). Wetzel, Konoske, & Montague, (1983), Bahrick, (1984), and Bahrick and Hall (1991) found that retention can
last up to fifty years.
The studies of retention have explored, at least somewhat, whether retention differs for different types of
knowledge. Semb, Ellis, and Aranjo (1993) define four types of knowledge—recall, recognition, comprehension, and
cognition. In their study, retention of recall knowledge was significantly lower than the other three. Semb and Ellis
(1994) argue that there are two dimensions of the retention interval that can affect retention—length, and what
occurred during the interval. With regards to the length of the interval, there are consistent findings that the amount
retained declines in a non-linear manner (Bahrick, 1984; Bahrick and Hall, 1991; Glasnapp, Poggio & Ory, 1978).
Farr (1987) suggests that the degree of original learning is the most important variable to long-term retention.
Bahrick and Hall (1991), in a study of Spanish and mathematics students, found a strong correlation between the level
of original learning and long-term retention. Specifically they studied the relationship between the number and level
of courses completed and retention. The students who completed more and higher level courses retained more. The
study reported here attempts to confirm this relationship by examining the correlation between a student’s grade in
one course and the student’s grade in the subsequent related course.
HYPOTHESES
We hypothesized that the grades of students who take the two courses with a shorter time between them will
correlate more highly and that correlation will decline over time in a non-linear manner as suggested by previous
studies.
H1: The correlation between student grades in statistics and decision science will be significantly lower if the gap
between the two courses includes the summer (3 months) when compared to the correlation between the
grades if students take the courses with four weeks between the courses (fall to spring semesters).
A
Journal of College Teaching & Learning – August 2005 Volume 2, Number 8
26
H2: The correlation between student grades in statistics and decision science will be significantly lower if
students take the courses with a full semester and summer gap between them (fall and fall or spring and
spring semesters) when compared to the correlation between the grades if the gap between the two courses
includes only the summer (3 months).
H3: The decline in student retention, as demonstrated in hypotheses 1 and 2, is nonlinear.
METHODOLOGY
Over three years we tracked the class performance of male students in two quantitative business courses
(statistics and decision sciences). The decision science course required use of concepts learned in statistics. Only
students who took both courses from the same instructor were included in the data. The tests in these two courses
were quantitative problem-solving based examinations. The professor uses the same test problems each term.
To measure the effects of time we compared students who took the two courses in three different
sequences—fall to spring (a 4 week gap), spring to fall (a three month gap) and from spring to spring or fall to fall (an
eight month gap).
RESULTS
The results of the correlation provide very strong support for hypothesis one. The students who took the
courses in the fall to spring sequence had a correlation of 0.86 between their test scores in the two courses. Students
who took the courses in a spring to fall sequence had a correlation of 0.71 between their test scores. That correlation
is significantly lower that the correlation for the first group of students (p<. 001). The support for hypothesis two is
less strong but indicates a further decline in retention. Students who took the two courses in a spring and spring or fall
and fall sequence had a correlation of 0.68 between test scores in the two courses. This is significantly lower than the
previous test group of students who took the two courses in a spring to fall sequence (p<06).
These results clearly suggest a non-linear decline in student retention of knowledge, thus supporting
hypothesis three. Examining the total decline in test performance over time further supports that conclusion. For
instance, 91% of the drop in correlations over eight months (fall to fall or spring to spring) is accounted for in the first
three months as suggested by the decline in correlation for those students who took the course sequence in spring and
fall. More than 50% of the correlation decline occurred after one month as shown by comparing correlation results for
students who took the two courses with only a semester break between compared to those who took the two courses in
the spring and fall with a summer break.
To confirm that the correlations represented a decline in performance we performed an analysis of variance
shown in Table 1. The results of that analysis were significant at the p<. 002 level. Those results make it clear that
the class performance scores were in fact declining over time.
Table 1: Analysis of Variance
ANOVA
df
SS
MS
F
Significance F
Regression
1
388.2994912
388.2995
9.538943
0.002426
Residual
140
5698.946988
40.70676
Total
141
6087.246479
ANOVA
df
SS
MS
F
Significance F
Regression
1
388.2994912
388.2995
9.538943
0.002426
Residual
140
5698.946988
40.70676
Total
141
6087.246479
Journal of College Teaching & Learning – August 2005 Volume 2, Number 8
27
DISCUSSION
This study contributes to the existing literature because of its specific design. Previous studies have relied on
tests to assess the retention of learning. The typical approach was to give students a test on a subject some time after
completing the course. This study is the first we could find that examines actual student class performance in courses
that require the use of knowledge learned in a previous course, specifically a pre-requisite course. Since the sample
was drawn from those students who took both courses from the same professor, the issue of what material was
covered with what degree of emphasis does not come into play.
This study suggests that the sequencing of courses can be important to students’ success. One surprising
aspect of the findings is the extent of decline in retention in the first weeks after the first course was completed. Over
50% of the retention decline seen after three months is evident after only four weeks. This finding seems to be
consistent with the views of Bahrick (1979) and Higbee (1977). However, as Semb, Ellis and Aranjo (1993) found,
retention of recall knowledge seems to decline more rapidly than other types of knowledge. Since the courses
involved in this study required a significant use of recall of concepts and methods from the statistics course to the
decision science course, the substantial early drop in knowledge retention may be a manifestation of what Semb, Ellis,
and Aranjo observed.
Farr (1987) and Bahrick and Hall (1991) found that the degree of learning was correlated to the retention.
Those findings suggest that students who did well in the first course (statistics) should also retain more knowledge,
and therefore do well in the subsequent decision science course. The high correlation (.86) between student grades for
those who took the two courses with only a four-week semester break is consistent with the conclusions of Farr (1987)
and Bahrick and Hall (1991).
This study has implications for course scheduling, advising, and curriculum planning. For example, students
should be encouraged to plan their academic sequences in a manner that takes these results into account. It is fairly
standard practice at our institution, for instance, to strongly encourage students to take all math requirements in a
semester-to-semester pattern and not “take a break” for a semester. The role of pre-requisites and the way courses are
scheduled to facilitate minimizing time between closely dependent courses may need to be given more consideration
in schedule planning by department chairs and deans. For example, it may make sense to offer more sections of a
course like statistics in the fall and subsequently offer more sections of decision science in the spring to accommodate
the maximum number of students. A similar pattern might make sense in other course like languages, accounting, and
mathematics. This kind of course scheduling pattern would facilitate more students taking sequential courses with
minimal time between them.
LIMITATIONS
Clearly further studies are needed to confirm these findings, especially in other disciplines. It should be
noted, though, that the findings in this study agree with previous studies cited and those studies included a variety of
subject matter ranging from social sciences to languages, to laboratory sciences. One limitation in this study relates to
student motivation. Our sample is derived from a student population that has well above average SAT scores who are
attending a highly regarded liberal arts college. The sample size of 141 in this study was large enough to likely
account for variations in GPA. Another area in need of more study is possible gender differences. The sample in this
study is all male since our institution is all male. While none of the previous studies cited noted any differences based
on gender, that is an area that needs further study specifically exploring the issue of possible gender differences.
REFERENCES
1. Bahrick, H. P. (1979). Maintenance of knowledge: Questions about memory we forgot to ask. Journal of
Experiment Psychology: General, 108:296-308.
2. Bahrick, H. P. (1984). Semantic memory content in permastore: Fifty years of memory for Spanish learning
in school. Journal of Experiment Psychology: General, 113:1-29.
Journal of College Teaching & Learning – August 2005 Volume 2, Number 8
28
3. Bahrick, H. P. and Hall, L. K. (1991). Lifetime maintenance of high school mathematics content. Journal of
Experiment Psychology: General, 120:20-33.
4. Conway, M. A., Cohen, G. and Stanhope, N. (1992). Very long-term memory of knowledge acquired at
school and university. Applied Cognitive Psychology, 6:467-482.
5. Cooper, I. L. and Greiner, J. M. (1971). Contingency management in an introductory psychology course
produces better retention. Psychological Record, 21:391-400
6. Farr, M. J. (1987). The Long-term Retention of Knowledge and Skills: A Cognitive and Instructional
Approach. New York: Springer-Verlag.
7. Glasnapp, D. R., Poggio, J. P., and Ory, J. C. (1978). End-of-course and long-term retention outcomes for
mastery and nonmastery learning paradigms. Psychology in the Schools, 15:595-603.
8. Higbee, K. L. (1977). Your Memory: How it Works and How to Improve It. Englewood Cliffs, NJ: Prentice
Hall.
9. Semb, G. B. and Ellis, J. A. (1994). Knowledge taught in school: What is remembered? Review of
Educational Research, 64:253-286.
10. Semb, G. B., Ellis, J. A. and Aranjo, J. (1993). Long-term retention of knowledge learned in school. Journal
of Educational Psychology, 85:305-316.
11. Semb, G. B., Ellis, J. A. and Montague, W. E. (1990). Long-term memory for knowledge taught in school.
(Tech. Rep. No. NPRDC TR-91-1). San Diego: Navy Personnel Research and Development Center.
12. Wetzel, S. K., Konoske, P. I., and Montague, W. E. (1983) Estimating skill degradation for aviation
antisubmarine warfare operators (AWs): Loss of skill and knowledge following training. (Tech. Rep. No.
NPRDC TR-83-31). San Diego: Navy Personnel Research and Development Center.