DOCUMENT RESUME
ED 353 334
AUTHOR
TITLE
PUB DATE
NOTE
PUB TYPE
EDRS PRICE
DESCRIPTORS
IDENTIFIERS
TM 019 400
Sutarso, Toto
Some Variables in Relation to Students' Anxiety in
Learning Statistics.
Nov 92
17p.; Paper presented at the Annual Meeting of the
Mid-South Educational Research Association
(Knoxville, TN, November 11-13, 1992). For related
document, see TM 019 380.
Reports
Research/Technical (143) -Speeches /Conference Papers (150)
MFO1 /PCO1 Plus Postage.
Academic Achievement; Attitude Measures; *College
Students; Correlation; *Educational Attitudes;
Educational Background; Higher Education;
*Mathematics Anxiety; Mathematics Education;
*Predictor Variables; Prior Learning; Rating Scales;
*Statistics; *Student Attitudes; Student
Characteristics
*Students Attitudes Toward Statistics Inventory
ABSTRACT
The purpose of this study was to investigate some
variables that relate to students' anxiety in learning statistics.
The variables included sex, class level, students' achievement,
school, mathematical background, previous statistics courses, and
race. The instrument used was the 24-item Students' Attitudes Toward
Statistics (STATS), which was administered to the statistics classes
at the College of Education and at the College of Commerce and
Business Administration at the University of Alabama (Tuscaloosa).
The STATS required students to describe themselves based on a 0 to 9
scale, with 0 being "does not describe me" and 9 being "describes
me." The sample included 79 male and 97 female students in
undergraduate and graduate statistics classes. The data were analyzed
in contingency tables using chi square statistics to compute
significance of relationships. All data analyses were performed on an
IBM miniframe computer. The association analysis showed that
there
was a significant relationship between students' anxiety in learning
statistics and the variables of students' achievement, statistical
preknowledge, school, and current class level. However, the results
do not provide enough evidence to suggest that there
was a
relationship between students' anxiety in learning statistics and the
other variables (such as college mathematics background, gender,
and
ethnicity). (RLC)
***********************************************************************
Reproductions supplied by EDRS are the best that can be made
from the original document.
***********************************************************************
U.S. DEPARTMENT OF EDUCATION
Office of Educational Research and Improvement
EDUCATIONAL RESOURCES INFORMATION
CENTER (ERIC)
(li'his document hes been reproduced as
received from the person or organization
originating it
O Minor changes have been mad to improve
"PERMISSION TO REPRODUCE THIS
MATERIAL HAS BEEN GRANTED BY
I'd Tel 5.0 Tg
reproduction duality
Points of view or opinions stated in this docu-
ment do not necessarily represent official
OERI position or policy
TO THE EDUCATIONAL RESOURCES
INFORMATION CENTER (ERIC)."
CYZ
ceD
SOME VARIABLES IN RELATION TO STUDENTS'
ANXIETY IN LEARNING STATISTICS
r.z4
Toto Sutarso
The University of Alabama
Paper presented at the annual meeting of
the Mid-South Educational Research Association
Knoxville, Tennessee, November 1992
BEST' COPY AVAILABLE
Discussion
"SOME VARIABLES IN RELATION TO STUDENTS'ANXIETY
IN LEARNING STATISTICS"
Toto Sutarso, University of Alabama
The purpose of the study was to investigate some
variables that relate to students' anxiety in learning statistics.
The variables included sex, class level, students' achievement,
school, mathematical background, previous statistics courses, and
race.
The instrument used was the Students' Attitude Toward
Statistics (STATS) instrument which administered to the statistics
classes at the College of Education, and at the College of Commerce
and Business Administration, the University of Alabama. The sample
consisted of 79 male students, and 97 female students.
The data were analyzed in contingency tables using the
Chi Square statistics to compute significance of relationships. All
data analyses were performed on the IBM 3081-400E (version 4.1).
The association analysis showed that there was a
significant relationship between students' anxiety in learning
statistics and the variables students' achievement, statistical
school,
and current class level. However,
an
interesting phenomena, the data showed that there was not enough
evidence to say that there was a relationship between students
anxiety in learning statistics and the other variables.
preknowledge,
1
SOME VARIABLES IN RELATION TO STUDENTS' ANXIETY
IN LEARNING STATISTICS
Introduction
Statistics
tools,
is considered one of the scientific
not only for describing data but also as an
inductive
method
research
in
methodology.
This
distinguishes it as a very important subject, especially
in higher education.
Statistics is offered at every
college. It is common, in some colleges, for statistics
to be given in series of courses, from°the very basic up
to the advanced level.
Statistics is also considered a complicated field.
Statistics
is
related
to
mathematics,
probability,
calculators, and computers. Previous studies showed that
there was a statistics anxiety among college students
(Roberts & Bilderback. 1980; Roberts & Saxe. 1982; Frank
& Rickard. 1988; Katz & Tomazik. 1988; and Benson. 1989).
Moreover,
Dillon
(1982)
reported
that
there
is
a
statisticophobia in college level statistics classes.
Considering the two aspects above,
investigating
variables which might relate to Students'
Toward Statistics (STATS)
Attitudes
is very important. The main
2
purpose of this study is to investigate some variables
that relate to students' anxiety in learning statistics.
Literature Review
Dillon
(1982)
reported
that
there
was
a
statisticophobia in college level statistics classes.
Statistics anxiety, even less than statisticophobia, is
'dangerous'
not only for the student but also for
statistics itself, and for other sciences in general.
Students who are anxious about a class will feel the
course
is
more
difficult
than
it
should
The
be.
instructional goals will be difficult to achieve. For
statistics
itself,
this
situation will
lead to
an
attitvtde of not liking statistics. Many students may try
to avoid this class. On the contrary, statistics courses
are needed as
a research tool.
Consequently,
since
statistics is a scientific tool in the research method,
the bad attitudes toward statistics may lead to a bad
impact on sciences in general, especially the sciences
that are closely related to it such as psychometry,
sociometry, biometry, and econometrics.
Research tends to show that there is a positive
relationship between STATS and student achievement.
3
Research done by Robert and Saxe (1982), by using the
Student Attitude Survey (SAS), indicated that the more
positive the STATS the higher the statistics achievement.
This finding was supported by Roberts and
who
Reese (1987)
found that regardless of gender grouping, more
positive STATS tended to show a higher course grade.
Moreover,
even
though
the
instruments
used
were
different, the data analysis still revealed the same
finding. This statement was verified by Water, Martelli,
Zakrajsek & Popovich (1988) when they used both SAS and
the Attitudes Toward Statistics (ATS) instrument of Wise
(1985) .
Benson
(1989)
used the Statistical Test Anxiety
(STA) which he developed to measure a student anxiety in
learning statistics. In the issue of the relationship
between the STA score and the course grade, he came up
with the conclusion that the higher the STA score the
lower the course grade. In other words the more anxious
the student was in learning statistics the lower the
course grade. In addition, a similar finding was also
mentioned by Ware and Chastain (1989), the more positive
the STATS, the lower their anxiety, and the higher the
score achieved. In conclusion, the research showed that
4
there was a positive relationship between STATS and
course grade; and that there was a negative relationship
between anxiety and both STATS and course grade.
Some variables were found related to the STATS.
However, others variables were found insignificant or
there were inconsistent findings among the previous
studies.
Research showed that there was
between
STATS
statistics
end
a
relationship
preknowledge.
The
correlation analysis tended to indicate that between the
two
variables
was
there
a
significantly
positive
correlation (Roberts & Saxe, 1982; Roberts & Reese, 1987;
and
Collis,
Oberg,
&
Shera,
1989).
Their research
findings consistently showed the positive relationship
between STATS and statistics preknowledge, or negative
relationship between students' anxiety and statistics
preknowledge.
However, regarding the relationship between STATS
and a number of previous college mathematics completed,
there was an inconsistency among research findings. Some
researchers found that there was a positive relationship
between the number
of
previous
courses completed anc4 the STATS.
college mathematics
The more previous
5
college mathematics courses completed the more positive
the STATS (Robert & Saxe, 1982; Roberts & Reese, 1987).
However, the research done by Benson (1989) showed a
tendency but an insignificant correlation between number
of college mathematics courses completed and statistics
anxiety. His findings were supported by other researchers
who mentioned that there was not enough evidence to say
that there was a relationship between number of college
mathematics completed and STATS (Collis, Oberg, & Shera,
1989; Ware & Chastain, 1989).
Sex differences also revealed an inconsistency among
the research findings. Roberts and Saxe (1982) concluded
that
there
was
a
significant
association
between
students' sex and their attitude toward statistics. Male
students tended to have more a positive attitude then
their
female
counterparts.
Consequently,
the
male
students tended to have a better statistics achievement
than
their
female
counterparts.
This
finding
was
supported by other researchers. It was found that female
students were more anxious than male students
(Benson,
1989; and Zeidner & Safir, 1989). In some studies, male
students tended to have higher scores in statistics then
their female counterparts (Frank & Rickard, 1988; Waters,
8
6
Martelli, Zakrajsek, & Popovich, 1988; Benson, 1989; and
Ware & Chastain, 1989). However, Elmore & Vasu (1986)
found the contrary. They showed that feminist issues were
the significant predictor of statistics achievement.
Female
students
showed
better
than
attitudes
male
students did. Consequently, female students tended to
have
achievement
higher
compared
with
male
their
counterparts. This finding was supported by Raiszadeh &
Ahmadi (1987). Their data analysis showed that female
students had higher statistical achievement than male
students had.
Other variables which related to the STATS were also
found
by
some
researchers.
They were material
and
instructor qualification (Johnson, 1980; Robert & Saxe,
1982; and Reisner,
1985;
Katz
&
1985), teaching methods
Tomazik,
1988),
(Reisner,
computer use
(Ware
&
Chastain, 1989; and Collis, Oberg, & Shera, 1989), the
status of a course whether required or elective, and
calculator attitudes (Roberts & Saxe,
1982). However,
research showed that ethnicity was not an indicator of
significant difference of anxiety
1989) .
(Zeidner
&
Safir,
7
Mpthod of the Study
The instrument used in this study was the STATS
instrument (Sutarso, 1991). The instrument was a set of
statements (24 items) which allowed students to describe
themselves based on 0 to 9 scale with 0 = does not
describe me and 9 = describe me. Each statement related
to attitudes toward statistics.
it had
In addition,
variables which may relate to their attitude in learning
statistics such as current class level, major, number of
previous
college
mathematics
and
statistics
had,
race/ethnic background, and gender. It was piloted with
20 students who took a statistics class. The purpose of
this
pilot
study,
especially,
was
to
try
out
the
instrument made. Based on the pilot study, the instrument
was
revised.
Finally,
it
was
administered
to
the
statistics classes at the College of Education, and the
College
of
University
Commerce
of
&
Alabama.
Business Administration,
The
classes
consisted
the
of
undergraduate statistics classes (BER 345, ST 250, and ST
251), and the graduate statistics classes were (BER 540,
BER 546, and ST 553).
The sample consisted of 79 male and 97 female
students. There were 8.5% Afro-American, 78.4% White, and
8
23%
other ethnic groups
variables
included
background,
in the sample.
were:
score,
class
level,
college/school,
Demographic
sex,
ethnic
statistics
preknowledge, and mathematics background.
The data were analyzed in contingency tables using
the Chi Square statistics to compute significance of
relationships. All data analyses were performed on the
IBM
3081-400E
mainframe
computer
using
the
SPSS-X
(version 4.1).
Result and Discussion
The first statistical comparison was between the
Students' Anxiety in Learning Statistics (SALS) score and
variable course score. The data tended to say that there
was a significant relationship between the SALS and the
course score (p < .002). The relationship between these
two variables was negative which means the lower the SALS
score the higher the students' achievement.
This finding was consistent with the literature
review
which
relationship
stated
between
that
there
Students'
was
a
Attitudes
positive
Toward
Statistics (STATS) and their achievement (Robert, & Saxe,
1982; Robert, & Reese, 1987; Water, Martelli, Zakrajsek,
11
9
& Popovich, 1988; Benson, 1989; and Were, & Chastain,
1989) .
The second statistical analysis was between SALS
score
and
statistics
preknowledge.
Statistical
preknowledge was measured by the number of statistics
courses taken before the student entering the class
surveyed. The Chi Square test showed that there was a
negative relationship between the SALS score and the
statistics preknowledge
(p
<
.01).
relationship shoWed that the more
This significant
a
student had
a
statistics preknowledge, the less their anxiety; or the
more confident a student is in learning statistics, the
more possibility that the
student will
take
other
statistics courses.
The third investigation was testing the association
between SALS score and college mathematical background.
The data tended to say that more mathematics background
succesfully reduced students' anxiety from the worst
level to the medium level. However,
the mathematics
backgound was not a significant predictor for the SALS
(p < .45). This association indicated that mathematics
background
was
not
important
as
preknowledge in relation to the STATS.
ji 2
as
statistics
10
This finding was supported by the conclusion of the
some previous studies (Benson, 1:89; Collis, Oberg,
&
Shera, 1989; and Ware, & Chastain, 1989). However, it was
inconsistent with the result of two other studies which
said there was a significant relationship between the
STATS and previous mathematics completed (Robert & Saxe,
1982; and Robert & Reese, 1987).
The fourth statistical comparison was about the
gender differences in SALS score. The Chi Square test
revealed insufficient evidence to say that there was a
relationship between gender and the SALS score (p < .64).
Also, there was insufficient evidence to say that male
students had more
confident
compared with their
female
in
learning
counterparts
statistics
nor
female
students had better attitude toward statistics. This
finding contradicted the inconsistent conclusions of both
sides in the literature review.
The
fifth
statistical
comparison
was
the
relationship between the SALS score and ethnicity. The
data revealed insufficient evidence to say that there was
a relationship between an anxiety in learning statistics
and ethnic background
(p <
.42).
This statement was
consistent with the conclusion of Zeidner & Safir (1989)
11
who reported that ethnicity was not an indicator of
significant difference of anxiety.
The oixth investigation was testing the association
between the SALS score and the students' school. Because
of the data distribution consideration, the variable
school just divided into three categories: Commerce and
Business Administration (C&BA), Education, and others.
The Chi Square test indicated there was a significant
relationship between the SALS score and the students'
school (p < .004). The data tended to show that students
from School of Education had more confidence in learning
statistics compared with both their C&BA and other
counterparts.
The final comparison was the relationship between
the SALS score and the students' class level. The data
tended to show that senior and graduate students had more
confidence in learning statistics compared with their
junior, sophomore and freshman counterparts. This finding
was
logical
since
the
probability
of
statistics
preknowledge is higher for seniors and graduate students
than for juniors, sophomores, and freshmen.
12
Conclusions
The study showed that there was a significant
relationship
statistics
between
and
some
students'
variables,
anxiety
such
in
as
learning
students'
achievement, statistics preknowledge, school, and current
class level. However, interesting phenomena, the data
:showed that there was not enough evidence to say that
there was a relationship between students' anxiety in
learning statistics and the other varia :es,
such as
college mathematics background, gender, and ethnicity.
The study might be useful for statistics instructors
to know their students' attitude toward statistics, and
realize some variables that relate significantly to their
anxiety in learning the subject. Knowing this, they could
provide better teaching
strategies
to overcome the
problems in teaching statistics course. It also would be
useful for educational researchers to know variables as
an indicator in investigating their relationship.
However, this study still needs improvement. There
is no research without error. Also, the finding here
might be restricted to the sample statistics classes in
the College of Commerce & Business Administration, and
the College of Education, the University of Alabama.
REFERENCES
Jeri.
(1989).
Structural
components
of
statistical test anxiety in adults: An exploratory
model. Journal of Experimental Education, 57, (3),
247-261
Benson,
Collis, B., Oberg, A., & Shera, W. (1989). An evaluation
of
computer-based
instruction
in
statistical
techniques for education and social work students.
Journal of Educational Technology System, 17, (1),
59-71.
Elmore, P. B., & Vasu, E.
S.
(1986). A model of
statistics achievement using spatial ability,
feminist attitude and mathematics-related variables
as
predictors.
Educational
&
Psychological
Measurement, 46, (1), 215-221.
Frank, M. L., & Rickard, K. (1988). Psychology of the
scientist: LVIII: Anxiety about research: An initial
examination
of
a
multidimensional
concept.
Psychological Reports, 62, 455-463.
Katz, B. M., & Tomazic, T. Z. (1988). Changing students'
attitudes
toward
statistics
through
a
nonquantitative approach. Psychological Reports,
(62), 658.
Kim, J. 0., & Mueller, C. W. (1991). Factor Analysis.
Sage University Papers on Quantitative Application
in the Social Sciences. Newsbury Park and London:
Sage Pub.
Raiszadeh, F. M. D., & Ahmadi, M. (1987). Students' race
and gender in introductory business statistics.
Journal of Education for Business Statistics, 63,
(1), 21-23.
Roberts, D. M., & Bilderback, E. W. (1980). Reliability
and validity of a statistics attitude survey.
Educational and Psychological Measurement, 40, (1),
235-238.
Roberts, D. M., & Reese, C. M. (1987). A comparison of
two scales measuring attitudes toward statistics.
Educational and Psychological Measurement, 47, (3),
759-764.
Roberts, D. M., & Saxe, J. E. (1982). Validity of a
statistics attitudes survey: A follow-up study.
Educational and Psychological Measurement, 42, (3),
907-912.
Ware, M. E., & Chastain, J. D. (1989). Computer-assisted
statistical
analysis:
A teaching innovation?
Teaching of Psychology, 16, (4), 222-227.
Waters, L. K., Martelli, T., Zakrajsek, T., Popovich, P.
M.
(1988).
Attitudes
toward
statistics:
An
evaluation of multiple measure. Educational and
Psychological Measurement, 48, 513-516
Wise, Steven L. (1985). The development and validation of
a scale measuring attitudes toward statistics.
Educational and Psychological Measujement, 45, (2),
401-405.
& Safir, M. P. (1989). Sex, ethnic, and
social differences in test anxiety among Israeli
adolescents. Journal of Genetic Psychology, 150,
Zeidner, M.,
(2), 175-185.