American Economic Association

The Effects of Attendance on Student Learning in Principles of Economics

Author(s): Garey C. Durden and Larry V. Ellis
Source: The American Economic Review, Vol. 85, No. 2, Papers and Proceedings of the
Hundredth and Seventh Annual Meeting of the American Economic Association Washington, DC,
January 6-8, 1995 (May, 1995), pp. 343-346
Published by: American Economic Association
Stable URL: http://www.jstor.org/stable/2117945
Accessed: 14-11-2015 05:24 UTC

REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/2117945?seq=1&cid=pdf-reference#references_tab_contents

You may need to log in to JSTOR to access the linked references.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact [email protected].

American Economic Association is collaborating with JSTOR to digitize, preserve and extend access to The American Economic
Review.

http://www.jstor.org

This content downloaded from 129.78.139.28 on Sat, 14 Nov 2015 05:24:59 UTC
All use subject to JSTOR Terms and Conditions
 The Effects of Attendance on Student Learning
in Principlesof Economics

By GAREY C. DURDEN AND LARRY V. ELLIS*

Does attendance affect performance in difference in the performanceof students

college economics courses? David Romer with no classroomattendancevis-a-visthose
(1993) found that attendancedid contribute attending class. Finally, Stephen Buckles
significantlyto the academicperformanceof and M. E. McMahon (1971) found atten-
students in a large intermediatemacroeco- dance at lectures that simplyexplainedma-
nomics course that he taught in the fall terial covered in reading assignments did
of 1990. (See the Summer 1994, Journal not enhancestudents'understandingof eco-
of Economic Perspectives [vol. 8, no. 3, nomics. In this paper we present new evi-
pp. 205-15] for numerous comments on dence on the effects of class attendanceon
Romer.) This conclusion held even after student performance.Our resultspertainto
controllingfor student motivationwhich, it the Principlesof Economicscourse as it is
may be argued,is the true factor determin- taught in a two-semester sequence at a
ing performanceand is only approximated medium-size, comprehensive state univer-
by attendance. An earlier study by Kang sity.
Park and Peter Kerr (1990) found that at-
tendancewas a determinantof studentper- I. The Data
formance in a money and banking course,
but not as important as a student's GPA The data for this studywere collected by
and the percentile rank on a college en- surveyingstudentsat the end of the semester
trance exam. A study by Robert Schmidt in several sections of the Principlesof Eco-
(1983) reported that time spent attending nomics course (both micro and macro). A
lectures contributed positively to perfor- questionnairewas administeredover three
mance in a macroeconomic Principles semesters,Springand Fall 1993 and Spring
course. 1994. The data on absences are the esti-
On the other side of the ledger is evi- mated numberof classes missed as reported
dence from Neil Browneet al. (1991) show- by the students themselves.' The observa.
ing that studentswho did not attend a typi- tions on studentgrades(on a ten-pointscale
cally structuredclass with lectures did just as collected,or as a percentageof the possi-
as well on the Test of UnderstandingCol- ble course points for those using a point
lege Economics (TUCE) as those students system)were normalizedto a single instruc-
who attended a standard microeconomic tor's gradingscale. This was done to mini-
Principlescourse. They also reported,how- mize any grade effects in the data across
ever, that those students who attended instructors.Table 1 providesdescriptionsof
the lectures performed better on essay the variablesemployed in the study, along
questions than those who did not. A with their means and standarddeviations.
similar study by Campbell McConnell and
C. Lamphear (1969) found no significant

1Sincesome classeswere large,rollwas not calledin

all sections.However,one researcherwas able to cor-
* Department of Economics, Appalachian State relate attendanceon eight unannouncedquizzeswith
University,Boone, NC 28608.We thankJohnSiegfried, student reported absences. The correlationwas 0.79
Tim Perri,BarryElledge, and Peter Kennedyfor help- (p < 0.01), even thoughsome studentsreportedmore
ful comments. than eight absences.
343

This content downloaded from 129.78.139.28 on Sat, 14 Nov 2015 05:24:59 UTC
All use subject to JSTOR Terms and Conditions
344 AEA PAPERS AND PROCEEDINGS MY 1995

TABLE 1 -DESCRIPTION OF VARIABLES (n = 346) II. Empirical Results

Variable Description Table 2 reportstwo ordinaryleast-squares

AVE Student course average (OLS) regressionswith course grade aver-
ABS Number of absences
1 if absences = 1 or 2, 0 otherwise
age (AVE) as the measure of student per-
ABS1 Dummy:
ABS2 Dummy: 1 if absences = 3 or 4, 0 otherwise
formance and the dependent variable. In
ABS3 Dummy: I if absences = 5 or 6, 0 otherwise regression(1), absences (ABS) enters as a
ABS4 Dummy: 1 if absences = 7 or 8, 0 otherwise continuousindependentvariablealong with
ABS5 Dummy: 1 if absences > 8, 0 otherwise
MSAT Math SAT score
other variableswhichwere includedto con-
VSAT Verbal SAT score trol for differences in background,ability,
GPA Grade-point average (x 100) on a four-point scale and motivation across students. The esti-
HSCHECON Dummy: 1 if had high-school
otherwise
economics, 0
mated coefficient on the number of ab-
CALC Dummy: 1 if taken college calculus, 0 otherwise sences per semester(ABS) has the expected
RACE Dummy: 1 if white, 0 otherwise sign and is statistically significant at the
Dummy: 1 if sorority or fraternity member,
FRATSOR
0 otherwise
1-percentlevel. Using the mean values for
ECON Dummy: 1 if previously had a college economics those independentvariablesthat are signif-
STUDEC
course, 0 otherwise
Time spent studying economics (hours per week)
icant at the 5-percent level, regression(1)
MF Dummy: 1 if male, 0 if female impliesthat an averagestudentwho has not
EDUC Dummy: 0 if either of the student's parents had missed any classes duringthe semester ob-
education or less, 1 if either
a high-school
parent had some college, 2 if either parent had
tains a course grade average of 74.8 per-
a college degree, and 3 if either parent studied cent. Alternatively,an averagestudentwho
at the graduate level
COLPREP Dummy: 1 if the high school program was college
has the averagenumberof absencesduring
preparatory, 0 otherwise the semester (3.5) achieves a course grade
EXCURR Dummy: 1 if one or more
activities, 0 otherwise
extracurricular
averageof 73.7 percent. It appearsthat the
HRSCAR Number of credit hours carried during current opportunitycost to the student in terms of
semester
in job
grade averageof those 3.5 absences is low,
HRSWK Number of hours worked per week
STATE Dummy: 1 if from North Carolina, 0 otherwise
but it could result in at least a half-letter-
grade reduction for those who are on the
Variable Mean SD numerical margin.
In order to explore further the relation-
AVE 72.234 10.000
ship between absences and academic per-
ABS 3.512 3.660
0.277 0.448
formance, we included absences as a di-
ABS1
ABS2 0.159 0.366
chotomousindependentvariable.As shown
ABS3 0.162 0.369
in Table 1, ABS1 equals 1 if the numberof
ABS4 0.061 0.239 absencesis 1 or 2, and is 0 otherwise;ABS2
ABS5 0.104 0.306 is 1 if the numberof absences is 3 or 4, 0
MSAT 515.647 89.372 otherwise;and so on. The results with ab-
VSAT 466.960 76.235 sences entered this way are reportedin re-
GPA 269.538 56.020 gression(2) in Table 2. Regression(2) indi-
HSCHECON 0.454 0.499 cates that absences do not affect student
CALC 0.532 0.500 performance until a typical student has
RACE 0.951 0.216 missedfive or more classes(i.e., with ABS3).
FRATSOR 0.214 0.411 Zero throughfour absences(i.e., ABS1 and
ECON 0.419 0.494 ABS2) have no statisticallysignificanteffect
STUDEC 2.400 1.883 on course grade average.
MF 0.621 0.486 Regression (2) also reveals that as the
EDUC 1.621 1.131 number of absences increases above the
COLPREP 0.538 0.499 thresholdlevel of 4, the negativeimpacton
EXCURR 0.720 0.450 grades increases.The coefficienton the bi-
HRSCAR 13.081 5.481
naryvariabledistinguishingmore than eight
HRSWK 7.818 11.097
absences exceeds (in terms of absolute
STATE 0.899 0.302
value) the coefficient on the variable for

This content downloaded from 129.78.139.28 on Sat, 14 Nov 2015 05:24:59 UTC
All use subject to JSTOR Terms and Conditions
VOL. 85 NO. 2 BETTER LEARNING FROM BETTER MANAGEMENT 345

TABLE 2-DETERMINANTS OF STUDENT PERFORMANCE the results of David Brasfieldet al. (1992)
IN PRINCIPLES OF ECONOMICS
(DEPENDENT VARIABLE: AVE)
who found that students who have had a
first course in calculus performbetter than
studentswho have not.
Independent (1) (2)
variable Coefficient t value Coefficient t value
At one time, the preponderanceof evi-
Constant 32.941 3.869** 33.919 3.941*
dence suggestedthat havingtaken a course
ABS -0.328 2.691** in high-schooleconomics had either no ef-
ABS1 0.465 0.382 fect or, possibly,a negativeeffect on student
ABS2 -1.538 1.105
ABS3 -3.228 2.286* performance in college economics (John
ABS4 -3.475 1.828t Siegfried and Rendigs Fels, 1979). Recent
ABS5 -3.521 2.117* studies, however, have reported a contrary
GPA 0.083 9.796** 0.078 9.125**
CALC 3.304 3.704* * 3.352 3.752** result, that a high-schooleconomics course
HSCHECON 2.738 3.304* * 2.766 3.346** improves performanceat the college level
VSAT 0.014 2.353* 0.014 2.350*
RACE 4.173 2.159* 4.524 2.327*
(A. Myatt and C. Waddell, 1990; Brasfield
MSAT 0.011 2.009* 0.011 1.972* et al., 1993). Our results are consistentwith
FRATSOR -1.987 1.903t -1.886 1.804t the recent studies and show that having
EDUC 0.718 1.899t 0.695 1.829t
ECON -1.091 1.271 -1.027 1.190 taken a high-school economics course
HRSWK -0.041 1.087 -0.049 1.283 (HSCHECON) contributes positively and
MF 0.663 0.735 0.736 0.797 significantlyto studentperformancein Prin-
EXCURR - 0.615 0.627 - 0.595 0.604
STATE - 0.699 0.509 - 0.737 0.536 ciples of Economics. Not surprisingly,we
STUDEC 0.103 0.457 0.093 0.411 also find that parents' educational attain-
HRSCAR -0.015 0.193 -0.021 0.266
COLPREP -0.156 0.178 -0.097 0.110
ment is positivelyassociatedwith students'
0.435
performance. Minority students (RACE)
Adjusted R2: 30.431
and studentswho are membersof a frater-
tStatistically significant at the 10-percent level. nity or sorority (FRATSOR) do not per-
'Statistically significant at the 5-percent level. form as well, other things constant.
**Statistically significant at the 1-percent level.
Finally,we find no gender-relateddiffer-
ences in student performance.This is con-
traryto much of the reported evidence on
gender effects in the literature.Many stud-
seven or eight absences, which in turn is ies have found that males score higherthan
larger than the coefficient on the variable females on multiple-choiceexamsin college
for five or six absences. The data seem to economics courses (see Siegfried, 1979;
suggest what many professorshave thought Keith Lumsdenand Alex Scott, 1987). Our
all along:the typicalstudentis not adversely result is consistent, however,with a recent
affected by a few absences, but excessive study(MaryWilliamset al., 1992)that found
absenteeism (in this case, five or more no significantdifferencein the performance
misses) is associatedstronglywith poor aca- of males and females. Their result held
demic performance. acrossdifferentcourses and also for perfor-
The estimates reported in Table 2 also mance measuresother than multiple-choice
reveal interesting results with respect to exams.
other independent variables that were sig-
nifrcant at the 10-percent level. Previous III. Conclusions
studies, including a recent one by Park
and Kerr (1990), confirmour findings that The results of this study indicate that
GPA and college-entrance-exam scores attendance does matter for academic
(i.e., MSAT and VSAT) are amongthe most achievementin a Principles of Economics
important determinantsof student perfor- course. The evidence suggests that the ef-
mance in college economicscourses. In ad- fect is nonlinear, becoming importantonly
dition, we find that having had a course in after a student has missed four classes dur-
calculus has a significantpositive effect on ing the semester.What reallyseems to mat-
studentperformance.This is consistentwith ter is excessiveabsenteeism.

This content downloaded from 129.78.139.28 on Sat, 14 Nov 2015 05:24:59 UTC
All use subject to JSTOR Terms and Conditions
346 AEA PAPERS AND PROCEEDINGS MAY1995

Why absencesbeyond a certain threshold Browne,Neil M.; Hoag,John H.; Wheeler,Mark

level affect performanceis not clear. Is it V. and Boudreau, Nancy. "The Impact of
because there are only a certain numberof Teachers in Economic Classrooms." Jour-
lectures that the student can affordto miss nal of Economics, 1991, 17, pp. 25-30.
before it begins to affect comprehensionof Buckles, Stephen G. and McMahon, M. E.
the material?Or does this "thresholdeffect" "Further Evidence on the Value of Lec-
simply reflect the fact that the better stu- ture in Elementary Economics." Journal
dents make good class attendance a habit of Economic Education, Spring 1971, 2(2),
and typicallydo not miss more than three or pp. 138-41.
four classes per semester?Furtherworkwill Lumsden, Keith G. and Scott, Alex. "The
be needed to answerthese questionsdefini- Economics Student Reexamined: Male-
tively. Female Differences in Comprehension."
Finally, our results largely confirm the Journal of Economic Education, Fall 1987,
findingsof previous studies with respect to 18(4), pp. 365-75.
most other factors which have been found McConnell, Campbell R. and Lamphear, C.
to affect studentperformancein economics. "Teaching Principles of Economics with-
As exceptions, we find no difference be- out Lectures." Journal of Economic Edu-
tween the estimated average scores of fe- cation, Fall 1969, 1(4), pp. 20-32.
males and males and that having had a Myatt, A. and Waddell, C. "An Approach to
high-schooleconomics course increases the Testing the Effectiveness of the Teaching
averagestudent'sgrade (all ceteris paribus, and Learning of Economics at High
of course). These latter two findings are School." Journal of Economic Education,
consistent with the more recent literature, Summer 1990, 21(3), pp. 355-63.
indicating that the high-school economics Park, Kang H. and Kerr, Peter M. "Determi-
course may be more effective than in past nants of Academic Performance: A
years,2 and that women may be more in- Multinomial Logit Approach." Journal of
clined to take those analyticalhigh-school Economic Education, Spring 1990, 21(2),
courses which foster the type of reasoning pp. 101-11.
that enhances performancein college eco- Romer, David. "Do Students Go to Class?
nomics classes. Should They?" Journal of Economic Per-
spectives, Summer 1993, 7(2), pp. 167-74.
REFERENCES Schmidt,RobertM. "Who Maximizes What?
A Study in Student Time Allocation."
Brasfield, David; Harrison, Dannie and McCoy, American Economic Review, May 1983
James. "The Impact of High School Eco- (Papers and Proceedings), 73(2), pp.
nomics on the College Principles of Eco- 23-28.
nomics Course." Journal of Economic Ed- Siegfried,John J. "Male-Female Differences
ucation, Spring 1993, 24(2), pp. 99-111. in Economic Education: A Survey." Jour-
Brasfield, David; McCoy, James and Milkman, nal of Economic, Education, Spring 1979,
Martin. "The Effect of University Math 10(2), pp. 1-11.
on Student Performance in Principles of Siegfried,John J. and Fels, Rendigs."Research
Economics." Journal of Research and De- on Teaching College Economics: A Sur-
velopment in Education, Summer 1992, vey." Journal of Economic Literature,
25(4), pp. 240-47. September 1979, 17(3), pp. 923-69.
Williams, Mary L.; Waldauer, Charles and
Duggal, Vjaya G. "Gender Differences in
Economic Knowledge: An Extension of
2This suggests improved teacher training in college the Analysis." Journal of Economic Edu-
and in economics education programs. cation, Summer 1992, 23(3), pp. 219-31.

This content downloaded from 129.78.139.28 on Sat, 14 Nov 2015 05:24:59 UTC
All use subject to JSTOR Terms and Conditions