Ed 187577

DOCUMENT RESUME
ED 187 577 SE 031 006

AUTHOR Suydam, Marilyn N., Ed.; Kasten, Margaret L., Ed.
TITLE Investigations in Mathematics Education, vol. 13, No.
2.
INSTITUTION Ohio State Univ., Columbus. Center for Science and
Mathematics Education.
PUB DATE 80
NOTE 71p.
AVAILABLE PROM Information Reference Center (ERIC/IRC), The Ohio
State Univ., 1200 Chambers Rd., 3rd Floor, Columbus,
OH 43212 (subscription 16.00, $1.75 single copy) .
EDRS PRICE MF01/PC03 Plus Postage.
DESCRIPTORS *Algorithms; Elementary Secondary Education; Games;
Learning Disabilities: *Mathematics Curriculum:
*Mathematics Education; *Mathematics Instruction;
*Problem Solving; Rational Numbers; Ratios
(Mathematics); Teacher Attitudes; Teacher Education;
Teacher Effectiveness.
IDENTIFIERS *Mathematics Education Research
ABSTRACT
Fourteen research reports related to mathematics '
education are abstracted and analyzed. Two of the reports deal with
teacher education, two with problem solving, three with basic
operations, and one each with learning disabled students, rational
numbers, proportional reasoning, counting, teacher effectiveness,
group cooperation on mathematical tasks, and verbal problem solving.
Research related to mathematics education which was reported in RIE
and CIJE between October and December 1979 is also listed. (MK)
INVESTIGATIONS IN MATHEMATICS EDUCATION
Editor Advisory Board
Marilyn N. Suydam Edward M. Carroll

The Ohio State University New York University
Jane D. Gawronaki
San Diego County
Department of Education
Lars
C. Jansson
Associate Editor of Manitoba
University
Margaret L. Kasten Thomas E. Kieren

The Ohio State University University.of Alberta
Published quarterly py
The Center for Science and Mathematics Education

The Ohio State University
1945 North High Street
Columbus, Ohio 43210
With the cooperation of theERIC; Science, Mathematics, and Environmental

Education. Information Analysis Center
Volume 13, Number 2 - Spring 1980
Subscription Price: $6.00 per year.. Single Copy Price: $1.75

Add 50c for Canadian mailings and'$1.00 for foreign mailings.
iii
INVESTIGATIONS IN MATHEMATICS EDUCATION
Spring 1980
Bright, George W. IDENTIFICATION OF EMBEDDED FIGURES BY

PROSPECTIVE ELEMENTARY SCHOOL TEACHERS. School
Science and Mathematics79: 322-327; April 1979.
Abstracted by ARTHUR F. COXFORD 1
Bright, George W.; Harvey, John G.; and Wheeler,
Margariete Montague. USING GAMES TO RETRAIN SKILLS
WITH BASIC MULTIPLICATION FACTS. Journal for
Research in Mathematics Education 10: 103-110;
March1979.
Abstracted by ROBERT ASHLOCK and DOUGLAS EDGE 4
Good, Ron. CHILDREN'S ABILITIES WITH THE FOUR BASIC

ARITHMETIC OPERATIONS IN GRADES K-2. School Science
and Mathematics 79: 93-98; February 1979.
Abstracted by JAMES M. MOSÉR 9
Greenstein, Jane and Stráin, Phillip S. THE UTILITY OF

THE KEY MATH DIAGNOSTIC ARITHMETIC TEST FOR ADOLESCENT
LEARNING DISABLED STUDENTS. Psychology in the Schools
14: 275-282; July 1977.
Abstracted by J. DAN KNIFONG 14
Kieren, Thomas E. and Nelson, Doyal. THE OPERATOR CON-

STRUCT OF RATIONAL NUMBERS IN CHILDHOOD AND ADOLES-
CENCE. Alberta Journal of Educational Research
24: 22-.30; March 1978.
Abstracted by LESLIE P. STEFFE 17
Kurtz, Barry and Karplus, Robert. INTELLECTUAL DEVELOP-

MENT BEYOND ELEMENTARY SCHOOL VII: TEACHING FOR
PROPORTIONAL REASONING. School Science and Mathe-
matics 79: 387-398; May-June 1979.
Abstracted by LARRY SOWDER 22
Mayer, Richard E. DIFFERENT RULE SYSTEMS FOR COUNTING

BEHAVIOR ACQUIRED IN MEANINGFUL AND ROTE CONTEXTS
OF LEARNING. Journal of Educational Psychology.
69: ,537-546; 1977.
Abstracted by THOMAS R. POST 25
Rakow, Ernest A.; Airasian, Peter W.; and Madaus, George
F. ASSESSING SCHOOL AND PROGRAM EFFECTIVENESS:
ESTIMATING TEACHER LEVEL EFFECTS. Journal of Educa-
tional Measurement 15: 15-21; March 1978.
Abstracted by JANE O. SWAFFORD 34
Robinson, Mary L. AN EXPERIMENT IN TEACHING GROUP
COOPERATION ON MATHEMATICS TASKS. School Science
and Mathematics 79: 201-206; March 1979.
Abstracted by SALLY H. THOMAS 37
Schoenfeld, Alan H. EXPLICIT HEURISTIC TRAINING AS A

VARIABLE IN PROBLEM-SOLVING PERFORMANCE. Journal
for Research in Mathematics Education 10: 173-187;
May 1979.
Abstracted by RICHARD E. MAYER 39
Silver, Edward A. STUDENT PERCEPTIONS OF RELATEDNESS

, AMONG MATHEMATICAL VERBAL PROBLEMS. Journal for
Research in Mathematics Education 10: 195-210;
May 1979.
Abstracted by GAIL SPITLER 44
Thornton, Carol A. EMPHASIZING THINKING STRATEGIES IN

BASIC FACT INSTRUCTION. Journal for Research in
Mathematics Education 9: 214-227; May 1978.
Abstracted by JOSEPH N. PAYNE 50
Vance, J. H. ATTITUDES TOWARD MATHEMATICS AND MATHE-

MATICS INSTRUCTION OF PROSPECTIVE ELEMENTARY
TEACHERS. Alberta Journal of Educational Research
24: 164-172; September 1978.
Abstracted by THOMAS COONEY 54
Webb, Norman L. PROCESSES, CONCEPTUAL KNOWLEDGE, AND

MATHEMATICAL PROBLEM-SOLVING ABILITY. Journal
for Research in Mathematics Education 10: 83-93;
March 1979.
Abstracted by LEN PIKAART 58
Mathematics Education Research Studies Reported in

Journals as Indexed by Current Index to Journals
in Education (October .- December 1979) 61
Mathematics Education Research Studies Reported in

Resources in Education (October - December 1979) 63
Bright, George W. IDENTIFICATION OF EMBEDDED FIGURES BY PROSPECTIVE
ELEMENTARY SCHOOL TEACHERS. School Science and Mathematics X79: 322-
'327; April 1979.
Abstract and comments prepared for I.M.E. by ARTHUR F. COXFORD,

University of Michigan.
1. Purpose
The goal was to determine whether prospective elementary school
tt--
teachers coulf identify certain geometry figures when the figure was
embedded in a more inclusive figure. In particular, were overlapping
figures identified as easily as non-overlapping figures; were embedded
figures identified in random order; and were embedáed triangles and
quadrilaterals equally easy to identify?
2. Rationale
In previous work Bright found that the performances of young
children (X 8.1 years) in identifying overlapping and and non-
overlapping figures were not the same. The investigation suggested
that one explanation for the performance differences might have been
that elementary school teachers could not teach the skill because
they could not identify overlapping figures themselves.
3. Research Design and Procedures

A set of four quadrilateral drawings and a set of four triangle
drawings which included overlapping and non-overlapping figures were
prepared. The relevant vertices were lettered. One set of drawings
was randomly distributed to each of 145 prospective elementary school
teachers in university methods and mathematical content classes. The
subjects were instructed to isolate all triangles or quadrilaterals
in each figure and to copy each figure so isolated in numbered spaces
provided on the page with the drawing. Thirty minutes were allowed.
The data were analyzed to determine thè numbers of overlapping and non-
overlapping figures identified and the order in which they were found.
4. Findings
For the two most complex triangle figures, 95 percent and 51 percent
of the non-overlapping and overlapping triangles were identified. Simi-
larly, 92 percent and 49 percent of the non-overlapping and overlapping
quadrilaterals were identified. The differences in each case were signi-
ficant (p <.001).
With regard to the order of identifying the figures, 59 percent
identified the four non-overlapping figures first in a comparable pair
of triangles and quadrilateral figures, and 63 percent identified the
non-overlapping figures first in a second set of two comparable figures.
Finally, the comparisons of identification rates for triangles and quad-
rilaterals showed that more triangles than quadrilaterals were identi-
fied (p <.01).
5. Interpretations
The author concluded that the three specific questions were
answered negatively. Non-overlapping figures were more easily identi-
fied and generally were identified first, and triangles were more
easily identified than quadrilaterals.
The author suggested that the results led to the following
educational implications:
a) Since prospective elementary teachers experience difficulty
identifying overlapping figures, they cannot be expected
successfully to guide students in similar activities.
b) Since quadrilaterals are less well identified than triangles
and both are important simple figures, attention needs to be
given to learning activities which involve general as well
as special quadrilaterals.
c) More attention may need to be directed toward developing
problem-solving performance of prospective elementary school
teachers, since only about 50 percent of the subjects
correctly completed the four most complex figures.
Abstractor's Comments
The rates of identifying overlapping and non-overlapping triangles
and quadrilaterals are interesting in themselves. As geometry teachers
are well aware, picking out simple figures from more complex figures is
not easy for many students. It is not unexpected, then, that the data
reported here confirm this informal observation.
The author would have been on more solid ground in his report if
he had discussed the order in which the students recorded the figures
they found. Data should be generated to decide the issue of the corre-
lation between the record and the actual order in which the figures were
recognized.
A second criticism pertains to the broad implications the author
drew from the data gathered. Certainly it is not clear that the
moderate performance in identifying overlapping triangles and quadri-
laterals is due to inability to identify such figures. An alternate
hypothesis could be that which figures were identified was a function
of the way the students interpreted the instructions which were given.
Before claiming inability or difficulty, further data need to be
gathered to dispose of alternate hypotheses.
The final two implications are, in the reviewer's opinion, reason-
able but seem to follow only indirectly from the data. Again alternate
hypotheses are available which need to be investigated.
The study has interesting data which should be used to generate
other conjectures about identification of overlapping figures. The
data are not sufficient to draw firm conclusions. Implications for
classroom practice and teacher education should be withheld until
additional data are gathered.
Bright, George W.; Harvey, John G.; and Wheeler, Margariete Montague.
USING GAMES TO RETRAIN SKILLS WITH BASIC MULTIPLICATION FACTS. Journal
for Research in Mathematics Education 10: 103-110; March 1979.
Abstract and comments prepared for I.M.E. by ROBERT ASHLOCK and DOUGLAS
EDGE, University of Maryland.
1. Purpose
The purpose of this study was "to investigate the use of games as
an instructional procedure in the retraining skills with basic multipli-
cation facts."
2. Rationale
A summary of research on the cognitive effects of games in mathe-
matics learning shows "that the research in this area is sparse, that the
findings are fragmentary and inconclusive, and that there have been no
systematic attempts to discover the effects Of games." Yet elementary
school teachers assume that games assist learning and use them regularly,
as evidenced by a cited study.

Two similar studies were conducted during the first ten days of the
school year, one in 1976 and one in 1977. Regarding subjects, "The 1976
sample was 14 intact classes (2 fourth-grade, 4 fifth-grade, 2 fifth/sixth-
grade, and 6 sixth-grade) from three elementary schools. The 1977 sample
was 10 intact classes (3 fifth-grade, 4 fifth/sixth-grade, and 3 sixth-
grade) from two elementary schools." The subjects were from white, mid-
dle-socioeconomic families.
Both studies used a teams-games-tournament model as a scheme for
classroom organization for instructional games. Two games from Developing
Mathematical Processes (DMP) were used, MULTIG and DIVTIG. Cooperating
teachers received training ,for their participation in the study, and
"agreed to refrain from teaching basic multiplication facts for the dura-
tion of the study."
In 1976, MULTIG was assigned to the fourth-grade classes, `U LTIG
and DIVTIG were randomly assigned (by grade level) to half of the remain-
ing classes, and DIVTIG only to the other classes. "On the first day, a
20-item, 15-minute power pretest of basic multiplication facts was given
to a random selection of one-half of the students." Of the 20 items,
10 were selected from the 36 multiplication facts used in MULTIG; the
other 10 were selected from the remaining 64 facts. A placebo addition
test was given to the other students. On the second day, teachers demon-
strated the appropriate game and explained the tournament and its struc-
ture. "During days 3-9, students played the game in the tournament struc-
ture and maintained records." On the final day, all students were given
a posttest equivalent to the multiplication pretest.
In 1977, MULTIG and DIVTIG were randomly assigned by grade level to
half of the classes. On the first day, all students were given a 5-minute,
100-item speed test followed by a power test. For the next eight days,
the sequence of events paralleled those of 1976. On the tenth day, all
students were administered a 5-minute, 100-item speed test followed by
a power test. The power tests were equivalent to the 15-minute, 20-item
posttest given in 1976.
In order to test the primary research question for both studies ("Can
games be used to retrain skills with basic facts?"), Wilcoxon matched-
pairs, signed-rank test statistics (Wilcoxon-T) were computed "for the
subscore of game-specific items on the power test, for the total score
on the power test, and, in the 1977 study, for the speed test score."
To test the first two secondary research questions ("Does pretesting on
the skills alter the posttest performance of students?"), "the researchers
ran an analysis of variance on the data of each class, with the two levels
of treatment defined as pretest and no pretest." Also, "F statistics
were computed for each class, one for the game-specific items on the power
posttest, and one for the total power posttest score." To test the other
secondary research question ("Are retraining effects altered by the use
of different drill and practice games?"), an ANOVA was run "on the 1976
means of all but the fourth-grade classes."
4. Findings
For each of the two years studied, the authors present pretest and
posttest means and standard deviations. For 1976, the pretest game-spe-
cific items score (10 items only) and total-items score (all 20 items) are
based on only half of each class, as the other half of the class was
given the placebo addition test. The posttest means and the standard
deviations are based on the full class size. Examination of the 1976
results showed, for the game-specific items, some increase in the means
for 11 of the 14 classes, no change for one class, and some moderate de-
crease for two classes. For the total item scores, 12 of the 14 classes
showed some increase in the mean scores; the other two showed decreases.
These mean differences between the pretest and posttest scores (using
the Wilcoxon-T) were, however, significant at the .025 level in favor
of the posttest scores.
For 1977, the pretests and posttests were given to the full classes.
Examination of these results showed that on both the game-specific items
score and the total items score, there were no decreases from pretest to
posttest mean scores. Again the differences were moderately small, yet
using the Wilcoxon-T they were significant at the .001 level. Pre- and
post-speed test scores were also given. Differences were significant,
again at the .001 level, but in this case the differences appear much
larger. Fifth-grade classes showed an average increase of 129 percent;
combined fifth-sixth-grade classes showed an average increase of 23 per-
cent; and the sixth-grade classes showed an average increase of 27 per-
cent. The overall average percentage increase was 56 percent.
Using the 1976 data, none of the 28 F-statistics calculated to de-
termine the effect of pretesting game-specific items and total items
was significant at the .05 level. The F-statistics obtained to compare
performance between the two-game treatment (MULTIG and DIVTIG) and the
one-game treatment (DIVTIG only) also were not significant at the .05
level.
5. Interpretations
The authors indicated, primarily on the basis of the "substantial
increases in the speed test scores", that it was reasonable to conclude
that the game treatment is an effective way to help students retrain their
skills with the basic multiplication facts.
Abstractors' Comments
The authors have presented a very clear rationale for conducting
these studies. Opinions about the use of games, like many other beliefs
in education, are generally accepted in practice but have not been ade
quately researched. The primary research question and the secondary re-
search questions do follow directly from their rationale.
The authors stated'that their interpretations must be considered
in the context of several limitations. Two.of these limitations, owing
to their importance, require special comment. It was acknowledged that
the study was conducted during the first ten instructional days of the
school year. Different children return from their summer holidays with
different expectations. Some children are eager to return to school;
others anticipate a return to "drudgery". This latter group especially
might be very excited about the games-tournament approach. The very
reason that many teachers choose to use games is for the motivation and
competition they provide. These may be factors that would not be as
significant at other, less atypical, times of the school year.
The second limitation that should be noted is that no attempt was
made to compare the effects of games with any other treatment. The total
absence of a control group is somewhat surprising. Simply returning to
school and being required to do "school-thinking", which presumably makes
some use of numbers, might improve one's facility with all previously
trained basic facts. Even if this increase is not as large as that re-
corded by a games treatment, it might be large enough to mitigate against
finding the games treatment significant at the .05 level.
The readability of the article could have been improved somewhat by
more careful attention to providing brief, explanation-oriented phrases
or sentences. Several examples are illustrated:
a) The authors wrote under the procedures section that "a random
selection of one-half of the students" were given the 15-minute
power pretest. Clarification of whether this is one-half of
total students or per class is only available by reading the
fine print under one of the tables, "the pretest was given ran-
domly to half of each class", or by interpreting a sentence listed
under the results section, "... the researchers ran an analysis
of variance on the data of each class, with two levels of treat-
ment defined as pretest and no pretest".
b) During a description of the treatment, the authors indicated
that "... During days 3-9, students played the games :n the
tournament structure and maintained records". No mention of
the amount of time, per day, was provided. Casually, only to
support an argument, was the "15-minutes a day" mentioned in one
of the final paragraphs of the articles.
c) Very careful reading (and likely rereading) is necessary to un-
derstand:
"In 1976, MULTIG was assigned to both
fourth-grade classes. MULTIG and DIVTIG were
randomly assigned, by grade level, to half of
the remaining classes. DIVTIG was used with
the other classes."
`` It might have been helpful tb indicate "... the two-game treat-
ment,DIVTIG,
MULTINGwas
andassigned by grade level to half the remain-
ing classes; and the one-game treatment, DIVTIG, was used with the other
half of the remaining classes."
d) Although the Wilcoxon-T appears to be appropriately used, as only
half of the 1976 classes were pretested, one must assume that of
the data presented at the posttest level, only half of the data
(those paired with pretest scores) were used to calculate the
Wilcoxon-T scores. As not all readers of the article would be
very familiar with the Wilcoxon matched-pairs, signed-rank pro-
cedure, it would have been helpful to include a sentence ex-
plaining the "half-use" of the 1976 posttest data.
A final area of comment is related to the "large" increase in speed
test scores. The authors appear to be particularly impressed with the
overall 56 percent increase in the 1977 speed scores. Yet this increase
is obtained from three sets of classes. One, the fifth-grade classes,
showed a 129 percent increase; the other two, combined fifth-sixth-grade
classes, had only 23 percent and 27 percent increases. What factor(s)
account for this great disparity? What factor(s) are operating within
these fifth-grade classes that are not operating within the other classes?
This study is a positive contribution to this area of research, but
should be considered with the following in mind: its lack of generali-
zability (to the whole school year) and its emphasis on overall average
percentage score increases attributable only to the treatment effect.
Good, Ran. CHILDREN'S ABILITIES WITH THE FOUR BASIC ARITHMETIC OPERA-
TIONS IN GRADES K-2. School Science and Mathematics 79: 93-98;
February 1979.
Abstract and comments prepared for I.M.E. by JAMES M. MOSER,

University'of Wisconsin-Madison.
1. Purpose
The purpose was to determine the ability of young children to under-
stand the operations of addition, subtraction, multiplication, and divi-
sion from the point of view of sets of concrete objects.
2. Rationale
The author had written a previous position paper suggesting that the
sequence of early mathematics instruction be changed in favor of inter-
relating the four basic operations. Beginning with concrete experiences,
the difficulty level should then increase with number size rather than
by type of operation. Several other researchers, including Piaget, were
cited as having presented evidence suggesting that concrete, manipulative
experiences should come first. Thus, the author sought to determine in
the present study whether the four basic operations were accessible to
youngsters within a concrete materials context.

Fifteen children each from grades K, 1, and 2 from the Florida State
University Developmental Research School were individually interviewed
during a 20-30 minute period, being given a set of seven tasks. The
tasks included one-to-one correspondence, number. conservation, counting,
adding, subtracting, multiplying, and dividing. Subjects were given
poker chips to use. A specially constructed number board depicting setS
of chips circled with string and arranged in order (see Figure 1) was
used with the last four tasks. Those tasks are listed below.
Figure 1. Number board used in the interviews,
depicting sets of poker chips...
Task Four: Adding'.

1. Do you see a group on the board just like my group (of poker
chips)? How can you tell?
2. Do you see a group on the board just like your group of chips
(child has eight chjps)? How do you know?
3. Which group on the board do you need to add to yours to make
mine? How do you know?
4. If you add this group of chips (three) to this group (four),
which group on the board will be just the same? How can you
'tell?
Task Five: Subtracting.
1. How many chips do I have to remove to make this group (of
chips) just like this group (on board)?
2. How many chips are in this group (six)? If I take these.two
away, how many will be left?
3. If I start out with this group of c,Jips (seven) and I end up
with this group (four on board), how many chips did I lose?
Task Six: Multiplying.
1. If I give you two groups like this (set of 3 chips); how many
will you have?
2. If I give you three groups like this (3 chips), how many will
you have?
3. If I give you four groups like this (2 chips), how many will
you have?
Task Seven: Dividing.
1. Can you make two equal groups out of this group (six)?
2. Can you make two equal groups out of this group (seven)?
3. Can you make three equal groups out of this group'(six)?
4. Can you make three equal groups out of this group (eight)?
For. each task, questions were asked using both poker thips and the
number board, until the interviewer was satisfied that each child had
indicated his or her best ability to respond.
4. Findings
All 45 students were able to count to 10 and establish one-to-one
correspondence using the pokér chips. Table 1 shows data for number
conservation and Table 2 presents data for the four tasks dealing with
the basic operations.
Table 1
Conservation of Number Task
Conserver Transitional Nonconserver
Grade K 6 4 5
Levtl 1 8 4 3
2 14 1 0
28 9 8
Table 2
Students' Success Tasks 4-7 by Grade Level
Conserver Transitional Nonconserver
K 1 2 K 1 2 K 1 2
Task 4 (Adding) 6 8 14 4 4 1 5 2 0
Task 5.(Subtracting) 6 8 14 3 4 1 5 2 0
Task 6 (Multiplying) 1 6 14 1 2 1 0 1 0
Task 7 (Dividing) 2 4 14 0 2 1 0 1 0
Most children categorized as transitional or itonconservers tended to rely

exclusively on counting in responding to questions in Tasks 4 and 5.
5. Interpretations
Counting can be used as a successful procedure in dealing with
simple addition and subtraction problems when objects are present.
The sharp drop in success on Tasks 6 and 7 for children in grades K
and 1 can be explained by their reliance on perception and/or count-
ing as means for solving number-related problems.
Making simple classes and series, conserving number, and using
reversibility all begin to appear at about 6 to 7 years of age and it
is not until this point that children can get beyond counting as a
means to solve "operations on numbers" problems. By second grade it
appears that most children have a concrete, intuitive concept of the
operations of addition, subtraction, multiplication, and division. If
multiplication and division are dealt with in the context of manipulative
materials, it appears that children who are able to conserve number
can understand these operations much as they conceptualize addition
and subtraction.
Much of the kindergarten and first-grade mathematics curriculum
should consist of concrete activities related to classification, seria-
tion, conservation, and reversibility. When logical operations are
introduced to children who are developmentally ready to understand
their meaning and usefulness, all four operations should be inter-
related. Since place value is one of the most difficult concepts for
children to understand, multi-digit numbers should be avoided until
the children show that they understand the four basic operations.
I would like to make some comments about the tasks, their adminis-
tration, and scoring. In the one-to-one correspondence task, the author
states that 8 red and 9 white chips were used to determine ability to
establish one-to-one correspondence between the sets. How can that be
when no such correspondence exists between two sets of unequal size?
Next, the counting task (tasks?) was characterized as determining ability
to count up to ten. How was this done? Was simple rote counting the
criterion or wns the attempt to look at "rational counting?" Without
belaboring the point, was the counting task made up of more than one
sub-task as was the case with the four operations tasks, and if so,
how was it scored? In the same vein, we do note that the four opera-
tions tasks were made up of several sub-tasks. How was success rated
--if a child passed one, two, or all three parts? On a much higher
level of concern, I feel that these short tasks are very superficial
measures of whether a child can deal with the basic operations, much
less exhibit any understanding of those operations.
The almost perfect success rate on the addition and subtraction
tasks by the subjects from grades K and 1 deserves mention. It is
practically unbelievable, especially when one looks at the difficulty
of the tasks involved. Sub-tasks 1 and 3 in subtraction are well known
and documented in the literature as being quite hard for children. The
fact that kindergarteners and first graders. did so well is astonishing.
I am led to suspect one or both of two possibilities exist. First, the
so-called "questions" by the interviewer during the task session con-
stituted a large measure of teaching and coaching. Second, the subjects
were extremely atypical. This may well be the case when subjects are
drawn from a university-based laboratory school. In either instance,
the results are open to serious question in terms of generalizability.
Finally, I want to express my concern over yet another example of
'a far-too-often-used practice, namely drawing conclusions and inferences
not based upon the evidence. I have ho real quarrel--in fact, I support
on an intuitive and theoretical basis--the author's suggestion that we
reconsider the sequencing of the elementary mathematics curriculum in
terms of interrelating the four basic operations. But the fact that 15
probably atypical second graders succeeded on some superficial tasks
purportedly measuring understanding of and ability with those four
operations does not imply that the curriculum should interrelate them.
And certainly the results have nothing at all to say about place value
and its potential difficulty.
In short, this article contributes very little to what we know about
how children can and should solve problems in arithmetic.
Greenstein, Jane and Strain, Phillip S. THE UTILITY OF THE KEY MATH
DIAGNOSTIC ARITHMETIC TEST FOR ADOLESCENT LEARNING DISABLED STUDENTS.
Psychology in the Schools 14: 275-282; July 1977.
Abstract and comments prepared for I.M.E. by J. DAN KNIFONG,

University of Maryland.
1. Purpose
The study attempts to judge the suitability of using the Key Math
Diagnostic Arithmetic Test (Key Math test) as a diagnostic tool with
learning-disabled adolescents.
2. Rationale
The identification of learning-disabled (LD) children has often
relied on linguistic ability. One possible reason for this orientation
is the lack of a suitable mathematics test. The Key Math test seems a
likely candidate for both the identification and description of a child's
learning disabilities. Specifically, would the Key Math test identify
LD adolescents through performance means, subtest performance patterns,
or computational error patterns which might then be correlated with the
child's years behind grade level? And, would the test also serve to
describe individual children's difficulties with mathematics?

Eighty-two adolescents (74 boys and 8 girls) between the ages of
12 and 17 who had been identified as learning disabled were given the
Key Math test. Each adolescent was at least two years behind grade
level and had been identified by a school psychologist as being learn-
ing disabled. Although several different tests were used by the psycho-
logists, each adolescent had been tested on the WISC-R or WAIS and
Bender-Gestalt. The Key Math subtest patterns and response errors from
these students were then compared with those of a normal population.
A Factor analysis of the subtest scores was performed and the correla-
tions between the means of each subtest score and year's behind grade
revel were computed.
4. Findings
The LD pattern of scores of the 14 subtest nearly matched those of
a normal fourth-grade population. This was interpreted as a "peaking"
effect in the LD adolescent's ability, probably due to the abstract
nature of mathematics beyond the fourth-grade level.
Statistically significant Pearson Product Moment Correlation Coeffi-
cients ranging from -.18 to -.44 were established for 11 of the 14 sub-
tests vis-a-vis years behind grade level. The three correlations which
were not significant were explained in terms of the content of the sub-
tests.
The fourteen subtests are grouped by the developers as covering
three major content areas: Content, Operations, and Applications. How-
ever, a factor analysis using a verimax rotation indicated only two
major factors: "Operations" and "Applications." The third subtest
grouping, "Content," split between these two factors. This result was
interpreted as suggesting that the original topical categories were
inappropriate for working with LD children.
Five computational error patterns were identified and the frequency
of their occurrence among normal and LD children was compared. Only one
of these patterns (the faulty recall of a number fact) was more frequent
among the normal group.
5. Interpretations
It was cohcluded from the Key Math test that LD adolescents:
a. function at the concrete level;
b. cannot do problems requiring "chained" operations,
e.g., if 3+4 - [J, then + 6 - ; and
c. follow a specific pattern of computational error.
It was further concluded that although the factor analysis does not
support the three-part categorization of the subtests proposed by the,
developers, the test is nevertheless a useful diagnostic instrument for
LD children.
Several of the interpretations and recommendations are not supported
by the scope of the study or the data.
Since specific criteria are not offered for determining "concrete
level functioning," there is no basis for claiming that the test res-
ponses indicated these adolescents were at that level.
It is interesting that the LD adolescents "peaked out" at the fourth-
grade level, but it is not clear whether this is attributable to the
abstract nature of the later mathematics or some other cause.
The fact that a factor analysis does not support the three content
areas of the test as proposed by its developer may indicate that these
categories are inappropriate for LD children, or it may indicate that
this sample tended to do about as well on the third category as they did
on the other two. A careful analysis of the test content would be neces-
sary to answer this question.
The only specific computational error pattern noted for these adoles-
cents was that they made fewer number fact errors than the normal group and
that they made more of the other five types of errors than the normal
group.
In the context of this study there are two separate, but related,
projected uses of a diagnostic mathematics test. Such a test should
identify LD children on the basis of their mathematics performance and
describe individual student weaknesses as an aid to planning instruction.
Clearly, adolescent LD children score very low on this test and their low .1
score correlates with their years behind grade level. However, the study
does not address the question of whether there are other, non-LD children
who score equally poorly on these tests. This would seem to be a crucial
question if the test is to be used to identify LD children. The second
projected use of the test rests on its content validity, an issue not
addressed by the study. Whether the Key Math test, or any other test,
actually includes a balanced range, type, and difficulty level of prob-
lems covering a particular grade level of the mathematics curriculum can
only be determined by a systematic analysis of those items. Such an
analysis may or may not reveal the Key Math test as comprehensive and
thorough.
Kieren, Thomas E. and Nelson, Doyal. THE OPERATOR CONSTRUCT OF RATIONAL
NUMBERS IN CHILDHOOD AND AdOLESCENCE--AN EXPLORATORY STUDY. Alberta
Journal of Educational Research 24: 22-30; March 1978.
Abstract and comments prepared for I.M.E. by LESLIE P. STEFFE,

The University of Georgia.
1. Purpose
Rieren and Nelson explored the view of rational numbers as operators
held by children and adolescents. Five questions were investigated:
a. Is there a change over age of reaction to operator settings?
b. Do there appear to be stages in this development?
c. Will there be difficulties with the concept of inverse in these
settings?
d. Are there differences in reaction to unit as opposed to non-unit'
fractions?
e. What mechanisms do children and adolescents use in handling these
settings?
2. Rationale
While formal instruction with whole numbers has been focused on com-
putation; it is based on a large amount of informal and practical
experience. Instruction with rational numbers has been based almost
solely on a computational construct. There is little relationship to
a body of experience of children or to the broad practical spectrum. The
study is based on a much broader construct of rational numbers which has,
as its capstone, mature functioning in a spectrum of rational number '
settings. Such functioning includes a well-developed concept of equiv-
alence experience with five basic subconstructs, and mechanisms by which
persons deal with the meaning of rational number situations.
3. Research Design and Procedure

Forty-five subjects were randomly selected, 5 in Grade 4 and 10 in
each of Grades 5, 6, 7, and 8. Each 10 consisted of five females and
five males. No more than two subjects came from any one classroom. "
Each subject was interviewed on a one-to-one basis using an apparatus
called a "packing machine," which packaged sheets of paper. Six tasks
were designed and administered as a power test in the following order:
a "1/2" task followed by "2/3," "1/3," "1/3 x 1/2," and "3/4 x 1/2"
tasks.* Because of performance, all subjects did not complete all tasks.
On any task, a subject could be faced with five direct and five inverse
items. For example, on the "1/2" task, the numbers of input sheets were
16, 10, 8, 18, and 6. Each interview followed a set protocol, adapted
to insure clarity. The machine was introduced and tried out to see how
it worked, using up to six trials with feedback. The subject merely
counted the input and output. Five prediction trials followed the
initial trials, where the subject was given the input and was asked to
record the output. Reasons were requested for several examples and a
request was made for the subject to tell how the machine worked. Follow-
ing this, five prediction trials were given, where the subject was given
the output and was asked to record the input. Reasons and a characteri-
zation of the working of the machine were again requested. After the
subject was unable to make predictions or understand how a machine func-
tioned, a partitioning strategy for running the machine was taught. The
subject then "ran" the machine and the experimenter had to answer ques-
tions.
4. Findings
a. The means for four age levels were as follows: 9.4 for subjects
less than 11 years of age; 15.5 for subjects between 11 and 12; 28.9 for
subjects between 12 and 13; and 31.3 for subjects greater than 13.
Differences of the means were significant between the first and second
age groups and the second and third age groups.
b. Six descriptive categories of responses were established: I,
Mastered no tasks; II, Mastered "1/2" tasks (direct and inverse); III,
Mastered unit tasks ("1/2" and "1/3" tasks); IV, Mastered unit and com-
position; V, Mastered unit. non-unit,and composition; VI, Mastered all
items using proportions. Attaining four of five items in a task consti-
tuted mastery.
*A "3/4" task was administered also, but the report does not indi-
cate its place in the sequence.
c. Of the first age group, 8 of 12 subjects were in category II,
3 were in I, and 1 was in III. Of the second age group, 6 of 10 were
in II, 2 were in II, and 2 were in I or IV. Of the third age group,
3 of 11 were in V, 3 were in IV, and 5 were in II or III. Of the
fourth (and highest) age group, 1 of 12 was in VI, 2 were in V, 3were
in IV, and S were in II or III.
d. Most responses were either correct (81) or incorrect (58) on
both items of a pair of direct and inverse items. Only one response
was incorrect on a direct item and correct on its corresponding inverse.
Thirteen responses were the other way.
e. Almost one-half of the response were correct for the unit items
but incorrect for the non-unit items. Only one response was the other
way. Fifteen out of 104 responses were correct for both.
f. Students thought subtractively and not multiplicatively in
analyzing problem settings--for the "2/3" operator, (12 -)8) would be
interpreted as "it's subtracting 4," and (30-420) as "it's subtracting
10," not focusing on the multiples involved.
g. .Over 91 percent of the subjects mastered the "1/2" tasks.
Students would give "1/2" responses when confused on other machines.
h..Twenty-seven subjects learned to "run" the machine, interchang-
ing roles with the interviewer--all subjects Grade 6 or above and 6 of
15 subjects below Grade 6. In all cases, only the direct aspect was
learned with very limited carryover to inverse and composition tasks.
5. Interpretation
The authors note that the data indicate strong age and stage
development'of students' reactions to operator interpretation for
rational numbers. Three levels of development are hypothesized: (a)
a "1/2" oriented level where a child's fractional conceptualization is
dominated by "1/2"; (b) a transitional level where subjects can handle
units and composition of unit operators; and (c) a level where all forms
of operators are handled. A fourth level is conjectured where operators
are incorporated into á system.
The authors note alsó that thinking of operators as ratios is not
necessary for achieving at the third level noted above. Only one
student, a 13-year=old, indicated use of proportions in analyzing the
machines--she named the composition of "3/4" and "1/2" as "3/8" after
one trial. Other students used a partitioning mechanism to get at
non-unit operators.
The authors are to be commended for conceiving and carrying out
this experiment. It represents a very difficult and, correspondingly,
an understudied area of research in mathematics education. There are
(at least) three critical achievements in a child's mathematical life--
A
the idea of ten as a unit, the idea of fraction, and the idea of an
unknown. Because these three ideas are so critical to the mathematical
progress of children, empirical work done in an attempt to elucidate
children's conception (and acquisition) of them must be both encouraged
and, at the same time, widely discussed. It is in this spirit that my
comments are made.
my first comments pertain to the claim that the data indicate stage
development for operator interpretations. Very careful consideration
must be given to the nature of developmental stages. Stages carry with
them the requirements of discontinuity of behavioral organization,
invariant sequence, hierarchical relations between successive stages,
structural organization, and preparation and achievement periods. In
view of these requirements, the authors rightly hypothesize three levels
of development of the operator interpretation. But, what are the cri-
teria for levels? Levels seem to involve measurement-a situation where
one level being of higher rank than another a priori demands measurement.
But the "levels" suggested do not seem to be of the nature of just
measurement sequences. They are much more like stages! Why the dilemma?
The data are cross-sectional and are not concerned with the acqui-
sition of operator interpretation of fractions--only with how children
react to interviews. Piaget's stages were never with reference_to more
advanced mathematics--they always dealt with rudimentary concepts--and
with good reason. To unearth stages with regard to more advanced mathe-
matical concepts than Piaget studied requires intensive study of the
dynamics of acquisition rof particular concepts--much more detailed and
intensive study than interviews allow. Interviews can be highly
suggestive and are not to be discouraged. But one cannot observe
possible processes of change without a chance to understand what those
processes might be. For example, how is "level 1" functionally inte-
grated into "level 2"? Does "level 2" develop and then reorganize
"level 1" or is "level 2" an abstraction of "level 1"? While other
important questionsineed to be asked and considered, it is more
important to note that research methodology used in mathematics
education needs to be expanded and changed to be consonant with the
intentions of the investigators.
My second set of comments pertain to the claim that thinking of
ratios as operators is not necessary for achieving a level where a child
can handle all forms of operators. I am not disputing that the 13-year-
old child used proportional reasoning when she named the composition of
"3/4" and "1/2" as "3/8." Even though she could have simply found the
product of 3/4 and 1/2, these authors know proportional reasoning when
they see it! However, because other students used a partitioning
mechanism to get at non-unit operators does not mean they were not
capable of proportional reasoning. The question is whether structures
which allow a task tb be assimilated are the.same ones that an observer
thinks he "sees" the subject using to execute a task. 'It is entirely
possible that whole-number knowledge was brought to Bear to solve a
task which was conceptualized by virtue of proportionality. As no case
was made by the authors that the children could hot engage in propor-
tional reasoning, their claim should be recast as a conjecture..
Again. my comments are meant only to extend discussioñ on what are
to me very important issues raised by the experiment. They are not to
be thought of as being derogatory of the study, as it was well done. .
Kurtz, Barry and Karplus, Robert. INTELLECTUAL DEVELOPMENT BEYOND ELE-
MENTARY SCHOOL VII: TEACHING FOR PROPORTIONAL REASONING. School Science
and Mathematics 79: 387-398; May-June 1979.
Abstract and comments prepared for I.M.E. by LARRY SOWDER,

Northern Illinois University.
1. Purposes
a) "A brief teaching program, suitably designed, can enable ninth
and tenth grade students in prealgebra classes to become profi-
cient in proportional reasoning.
b) "A'form of this teaching program making use of manipulative ma-
terials will be more effective than a form using pencil and paper
activities only.
c) "Student attitudes toward this teaching program will favor the
form with manipulatives over the pencil and paper form." (p. 387)
2. Rationale
Several status studies, built on Piagetian work with proportional
reasoning, have indicated that many adolescents have difficulty with prob-
lems involving proportions. Thus, whether instruction can improve stu-
dents' performance on such tasks deserves investigation. Laboratory ap-
proaches "invite the active participation of all students, form a concrete
basis for abstractions, and facilitate peer group interactions" (pp. 387-
388). Hence, studying the effect of a laboratory approach seemed desirable.

There were two series of investigator-designed and piloted lessons,
one involving manipulatives and the other involving only paper and pencil.
In these lessons the students worked with constant ratio, constant sum,
and constant difference relationships from tables of data, with laboratory
settings (it is not clear that the manipulative version and the paper-pencil
version involved the same settings), and with solving proportion problems.
The lessons required 14 class periods.
Eight intact prealgebra classes, four for the manipulative version
and four for the paper-pencil version,. served as subjects. Four teachers
were inbolved, each teaching each version to one class; these teachers
underwent a 16-hour training course on the lessons. One class from each
of six other teachers provided a control group. All the classes came
from four schools.
Equivalent forms of investigator-written, eight-item tests were given
as the pretest, the posttest, and the (two-month) delayed posttest.
Scoring of the items was based on "level" of response--a seven-category
system describing the process of solution and documented in earlier studies.
The upper levels reflect increasingly sophisticated forms of proportional
reasoning. Attitude data were based on an attitude survey and observations.
The analyses reported were based largely on. single items from the
first four items on the test. Two of the other items were on secondary
objectives (e.g., drawing graphs), but no'reason was given for ignoring
the remaining two items, except that they were similar to two of the items
reported.
4. Findings
Roughly two-thirds of the students in the experimental classes ex-
hibited proportional reasoning on the posttest, whereas only about one-
sixth did on the pretest. The corresponding change in the control classes
was from about one-tenth to one-third. Similar results held for the de-
layed posttest with the experimental groups.
5. Interpretations
a)"This study has shown that it is possible to advance the use of
proportional reasoning of many secondary school students by means
of a well-designed teaching program requiring approximately three
weeks of school time" (p. 396).
b)".,.,the cognitive gains were equal for the two experimental groups"
(p. 397).
c)"The attitude surveys and classroom observations made during the
study ... showed that the manipulative version was considerably
more popular than the pencil and paper version. ... We conclude
that the manipulative laboratory materials had a motivating value
for many students and should therefore be considered insecondary
school mathematics classes when student interest and attitude to-
' ward the subject are important issues" (p. 397).
The status studies cited demand that we pee whether instruction can
improve students' use of proportional reasoning. I particularly liked
the authors' choice of subjects and the use of a two-month delayed post-
test. If long-term improvement can be accomplished with relatively un-
scholarly students, we teachers can take heart. On the other hand, some
experienced teachers might sag, "Why didn't they ask me whether it could
be done? I've been doing it for years!" Now there is documentation.
It is not clear why the analyses centered on only four of the six.
relevant test questions. An authors' comment on that point, and on their
feelings about the possible effects of the lack of random assignment,
might have been comforting.
As the authors note, this study does not shed any light on which
facets or combinations of facets of the lessons--e.g., laboratory activi-
ties, or the focus on the relationships possible among variables--might
'be responsible for the learning. Max Bell has described many educational
studies as being at the "earth, air, fire, and water" stage of knowledge.
I would so classify this study. (The investigators might disagree, since
they used a Piagetian-based approach in developing the lessons.) In any
case, such studies can be valuable, particularly when they suggest "mole-
cules" that might be basic ingredients in learning.
Mayer, Richard E. DIFFERENT RULE SYSTEMS FOR COUNTING BEHAVIOR ACQUIRED
IN MEANINGFUL AND ROTE CONTEXTS OF LEARNING. Journal of Educational
Psychology 69: 537-546; 1977.
Abstract and comments prepared for I.M.E. by THOMAS R. POST,

University of Minnesota.
1. Purpose
This series of three experiments was designed to test the hypothesis
that different underlying rule systems (cognitive objectives) can be ac-
quired even though the same levels of mastery performance on a given be-
havioral objective are achieved. These studies attempt "to specify more
clearly the characteristics of different learning outcomes capable of
supporting the same criterion behavior and to test more clearly the idea
that different instructional conditions can lead to the acquisition of
different outcomes supporting the same behavior" ,(p. 538). Examined is
"counting behavior" in base three using both numbers and letters to rep-
resent the patterns.
2. Rationale
In this study behavioral objectives of instruction which focus pri-
marily on the behavior produced by the learner are distinguished from
"cognitive objectives" which specify "the psychological processes and
structures that are sufficient to produce the needed behaviors" (Green,
1976,-p. 123). Wertheimer (1959) has argued that concepts learnéd through
meaningful apprehension of relations has a potentially broader learning
outcome capable of superior transfer and long-term retention despite the
fact that a less meaningful approach might produce the same level of per-
formance on an immediate mastery test. Similarly Scandura (1970) and
Ehrenpreis and Scandura (1974) have suggested that the same mathematical
skill can be supported by a set of simple associations (low level rules)
or by a higher order rule system capable of generating those associations.
An early evaluation of Sesame Street (Bogatz and Ball, 1971) indica-
ted that the program has been demonstrated to be highly successful in
imparting a number of basic cognitive skills such as labeling and rate
counting, but "has been much less effective in teaching more complex rule-
governed behavior" (Henderson et al., 1975, p. 480),
These and other results encourage the search for instructional pro-
cedures which might be responsible for producing broad and narrow rule
systems in learning a mathematical skill such as counting.

In this series of three experiments, 68 subjects in introductory
psychology at the University of California, Santa Barbara, learned to
recite the first 12 or 18 responses in one of two formats in base 3.
A meaningful context of learning was defined as a way to relate the
counting to past experience with base 10.
Experiment 1: All 24 students in this study learned the same count-
ing series, but for half the format was numerals (0, 1, 2, 10, 11, ...,
200) and for half it was letters (w, d, r, dw, dd, ..., rww). The num-
erals or letters were presented one at a time using a shielded typewriter.
One transfer test consisted of four addition problems involving no carry-
ing and four problems involving carrying (for each format); a second
transfer test (on counting) required students to supply the next 10 items
aftér a given numeral or letter.
Subjects were randomly assigned to treatment; Group Number received
only number symbols during learning and transfer; Group Letter received
only letter symbols. Each subject was instructed by the experimenter to
type on the tape what he or she thought came next after the given symbol.
All subjects continued until correctly anticipating all symbols in the
list twice in a row. Transfer lists were then administered.
Experiment 2: ,The procedure was identical to that for Group Letter
in Experilent 1 except that half of the 24 subjects (drawn from the same
population as in Experiment 1) were shown the letter-to-digit conversion
card prior to learning (Group Before) and half were shown the same card
after learning but before the transfer test (Group After).
Experiment 3: This experiment refined the procedures of Experiment
2 by using a more homogeneous subject population (SAT scores averaged 600),
by using a shorter counting sequence (13 items),-and by controlling the
time spent on each response,(10 seconds). Items were presented via a
slide projector (onto a screen) to two or three subjects, who wrote re-
sponses on paper. Only letters were used; the Before and After instruc-
tions indicated to subjects that they could "think of the letters as
numbers, w - 0, d 1, r - 2, if that helps you."
4. Findings
Experiment 1: The 18 responses for 1 through 200 (or d through rww)
were divided into six blocks with three positions within each block. (The
positions were items ending in l or d, 2 or r, and 0 or w.) Subjects in
Group Letter produced an average of 79.5 errors in learning to criterion,
compared to 9.8 errors for Group Number [F(1,22) - 44.29, p < .001].
This result was expected due to the fact that Ss had more experience in
counting with an ordered set of numbers than random letters.
The Group x Block Interaction was examined to determine if the
groups had in fact used different strategies in learning the sequences.
If so, Group Letter might produce a serial position curve similar to
that obtained with nonsense syllables -- a strong primary effect and a
leveling off -- while Group Number would not. The significant Group x
Block Interaction [F(5,110) = 8.14, p < .001] resulted from the fact that
Group Letter, as hypothesized, did perform better on the early blocks,
while Group Number showed the reverse trend. The average per-pupil er-
rors for Group Letter were 7.2, 12.6, 13.8, 15.0, 15.9, and 15.0, for
Blocks 1 - 6 respectively. The corresponding errors for Group Number
were 2.0, 1.6, 1.8, 1.6, 1.6, and 1.2.
A second prediction is that Group Letter should have approximately
equal difficulty for base change and nonbase change positions, since all
are equally meaningless, but Group Number should have relatively more dif-
ficulty with base changes than nonchanges. This prediction is partially
consistent with the reliable Group x Position Interaction [F(2,44) = 11.43
p - .001], in which Group Number had relatively more difficulty with base`
change than nonchange positions, but'Group Letter tended to have a more
balanced pattern of errors. The average errors for Group Letter were
25.2, 19.2, and 35.4 on the two nonbase change positions and one base
change position, respectively, and the corresponding errors for Group
Number were 1.8, 0.8, and 7.2. It was noted that since Group Number out-
performed Group Letter in all positions, the above interaction is not dis-
ordinal; however, an analysis of the percentage of errors to total errors
shown in the left portion of Table 1 revealed that Group Number committed
'a higher proportion of errors at the base change position than Group Letter
(t(22) - 6.12,p < .001). The difference in the proportion of errors at
the base change is consistent with the hypothesis that different learning
strategies were used.
Table 1
Proportion of Total Errors Occurring at Base
Change Positions for Two Groups
Group Experiment
1 2 3
Letter or After .48 .45 .43

Number or Before .78 .55 .61
t test, p < .001 - .05 < .001
Note. Base change position ends with w or 0. Experi-

ment 1 compares Letter and Number Groups; Experiments
2 and 3 compare After and Before Groups.
Since both groups learned the same sequence to the same criterion
of mastery (two errorless trials), no difference in post-test performance
was predicted. Because of differential error patterns, it was concluded
that the groups may have used different learning strategies; i.e., if
Group Letter Ss learned by a rote strategy, the predicted outcome of
learning would be a memorized chain of 18 sëquential responses. If sub-
jects in Group Number learned by telating the new sequence to their past
experience with number systems, the predicted outcome of learning would
be a system of'rules for generating the first 18 items in base 3. Thus,
it was predicted that Group Number would 'do better on transfer tasks, and
this was.borne out. Group Number performed better than Group Letter on
arithmetic (t(22) 10.77, p < .001) and on counting (t(22) di 3.65, p < .01 )
related post-tests.
Experiment 2: Group After (identical to Group Letter in Experiment
1) averaged 82.2 errors in learning to criterion, compared to 37.6 errors
for Group Before (F(1,22) - 4.71,'p < .05). The pattern for Group After
was similar to that of Group Letter in Experiment 1, with better perfor-
mance in early blocks than middle or later blocks. The two interactions
tested, Group x Block and Group x Position, occurred in the predicted di-
rection but' were not significant (p - .05). Group Before performed sig-
nificantly better on counting transfer test (p = .05), suggesting that Group
Before acquired a broader learning outcome. Apparently the conversion
table served as a meaningful context (see Table 2) of learning which al-
lowed broader encoding of the to-be-learned sequence.
Experiment 3: As in Experiment 2, Group After committed significantly
more errors than Group Before (50.0 vs. 17.5, F(1,18) - 24.27, p < .01).
This time,Group x Block Interaction (F(3,54) - 6.38, p < .011 reached sta-
tistical significance. Once again the Group x Position Interaction failed
to reach statistical significance. Group Before's higher proportion of er-.
rors at base change portions was, however, significant [t(18) - 3.37, p < .01].
Group Before performed significantly better than Group After in the arithme-
tic test (t(18) - 2.72, p < .02] (see Table 2).
Table 2
Proportion Correct Response on Transfer
Tests for Two Groups
Group Exp. 1 Exp. 2 Exp. 3

Add Cou Add Cou Add Cou
Letter or After .01 .18 .49 .61. .17 .21
Number or Before .56 .72 .49 .86 .50 .61
t test, p < .001 < .01 ns .05 < .05 < .02
Note. Experiment 1 compares Letter and Number Groups; Experiments 2 and 3

compare After and Before Groups.
The studies suggest that qualitatively different learning processes

were used by the various treatment. groups. In general, Ss in Group After
(or Group Letter in Experiment 1) seemed to use what could be called a
rote learning set. Ss in Group Before (or Group Number in Experiment 1)
seemed to use a different strategy that could be called a meaningful learning
set (i.e., relating presented material to an integrated set of existing know-
ledge and in particular relating to part-counting experience in base 10).
5. Interpretations
These results indicaté thatthe same behavior (e.g., counting to 12 or
18 in base 3) can be supported by entirely different learning processes
and learning outcomes. Further, these results caution against teaching
a particular problem-solving behavior by using behavioral objectives for
,mastery that ignore the underlying cognitive objectives of instruction
(Greeno, 1976). To avoid narrow learning outcomes such as exemplified
by Group After in the present study, instructional objectives should be
sensitive not only to what behavior is learned, but also to how it is
learned and structured in memory. There are situations in which narrow
learning outcomes might be preferred, but when new learning must build
on the outcomes of present learning (such as in number systems), narrow
learning would have particularly serious consequences.
Like some previous studies on the sequencing of an advance organizer,
the present results'argue against the straightforward idea that Group
After and Group Before must have learned the same thing, since they were
presented with the same stimuli. Broader learning of new technical skills
may be enhanced by providing a meaningful context prior to learning rather
than after, presumably because the context serves to establish a meaning-
ful learning set (Ausubel, 1968).
The concern of rote vs. meaningful learning is not new. Bruner,
Dienes, Piaget, Gagne, and others have all addressed this issue in one
way or another. Each has advocated'the development of coherent schemas
or structures which relate past to pretest learnings and lays the foun-
dation for the meaningful assimilation and accommodation of new experi-
ences into one's existing cognitive map.
Mayer's three studies, like earlier investigations, conclude that
"broader learning of new technical skills may be enhanced by providing
a meaningful context prior to learning rather than after, presumably be-
cause the context serves to establish a meaningful learning set." Hardly
an idea foreign to mathematics educators, it is nonetheless of crucial
importance. This paper offers additional evidence to support the exist-
ence of these differential cognitive structures.
Mayer, in this series of three related experiments, has inferred
certain aspects about the nature of the underlying cognitive structures
which support overt behavior -- an admittedly difficult-area. These
inferences, perhaps by necessity, are based on what might be thought of
as circumstantial evidence. At some points in this paper the author ap-
pears to be arguing the converse of an admittedly true statement. That
is, given that the statement "if X Then Y" is true, the converse "if Y
then X" is believed to be true. This may or may not be the case. Mayer
has established in several cases by past research that X does, in fact,
imply Y. A potential difficulty arises when one concludes from this that
"if Y then indeed X." It must be pointed out that Mayer is careful to
qualify his conclusions with the use of "if," "might," "could," etc.
The danger of overly ambitious conclusions is present, however, unless
the reader is careful.
Mayer's use of college-age students with "no prior experience with
counting in nonbase 10 number systems" may have changed the perceived
nature of the problem task from that of a counting-related problem to
that of a problem more concerned with pattern recognition. This research,
therefore, may not really be dealing with counting-behavior as suggested.
The author's distinction between rote and meaningful context of learn-
ing is in need of further consideration. Mayer defines meaningful context
as "providing a way to relate the counting series to past experience with
base 10, e.g., presenting the letter to digit conversion list prior to
learning." A rote context is presumably the absence of this.
The fact that subjécts might, in fact, exhibit a variety of strate-
gies in attempting to deal with the "counting tasks" seems to have been
overlooked. Mayer suggests that students will relate the task to past
experience with base 10 and that thè crucial issue (in Experiment 2 and 3)
is whether or not Ss are presented with conversion (letter to number) in-
formation before or after instruction. An alternate possibility exists.
It is indeed possible to address successfully the counting issue without
recourse to the structural properties of base 10 or numeration systems
in general. When faced with the sequence w, d, r, dw, dd, dr, rw,. ...,
it is possible to predict the next element in the sequence by simply ob-
serving the patterns which emerge. It would also be feasible to use the
same procedure if the numbers 0, 1, and 2 replaced the letters w, d, and
r, respectively. Several individuals, when asked by this reviewer to
predict the next number (letter) in the sequence, responded in this man-
ner; i.e., not on the basis of previous knowledge of base 10 but by search-
ing for recurring patterns; be they number or letter. This point is made
only to suggest that alternate solution strategies exist for the problems
solution in much the same way that real problems usually elicit various,
solution strategies from a variety of individuals. This then casts some
doubt on the appropriateness of Mayer's definition of meaningful context.
The assumption that the presentation of a numerical sequence is somehow
more meaningful than the presentation of a letter sequence because it alone
can be related to previously established cognitive structures (base 10) is
questionable. An alternate hypothesis would suggest that meaningful con-
text in this case implies a good deal more than familiarity with the base
10 numeration system; i.e., the ability to generalize a pattern given a
small number of examples. It seems possible that, for a given individual,
either context (letter or number) could constitute a meaningful context
given appropriate previous experiences.
It would have been useful and relatively easy for the author to have
interviewed Ss in these experiments to ascertain the nature of the strat-
egies employed. It is unfortunate that this was not done. Some informa-
tion was gathered regarding solution methods used, but this was not re-
ported in the article.
This raises the question as to why individual students given the
number sequence outscored those given the letter sequence in Experiment
1 and those given the conversion Suggestion (i.e., let w - 0, d - 1, and
r gi 2 in Experiments 2 and 3), if it were not because of direct related-
ness of the sequence to previously developed cognitive structures? The
answer may lie in the domain of the perceived relatedness of the context
within which the problem is embedded, not in the meaningfulness of the
problem itself. Subjects are generally more comfortable with what seems
to them to be a more concrete embodiment of the problem situation. 'Having
had more experience with number related patterns, it would seem likely
that subjects would be more comfortable in this area. This is analogous
to the uneasiness which first-year algebra students face when first con-
fronted with the concept of variable.
It seems necessary to distinguish meaningful context from meaningful
learning.
The difficulties resulting from the fact that both groups in Exper-
inent 1 received different posttests was rightfully acknowledged by Mayer.
He suggests (and I agree for the reasons just discussed) that letters
may be inherently more difficult regardless of thé experience. It would
have been interesting to have each group take the others' posttest as a
measure of transfer. If the Number Group really did develop a higher
order rule system, as suggested, it should be able to deal effectively
with the letter sequence with or without the letter to number conversion
suggestion. That is, it should be able to deal comfortably with struc-
tural similarities (isomorphisms) between the number and letter patterns.
It is useful to re-affirm continually the importance of meaningful
learning experiences in mathematics education. Overly conservative trends
such as tie Back-to-the-Basics movement have a tendency to ignore or sub-
limate this aspect of instruction to the more visible and easily measured
overt performance on clearly specified behavioral objectives. Unfortu-
nately, those objectives most easily measured are precisely those which
reflect algorithmic procedures. These, in turn, can be taught in either
rote or meaningful ways. All too often the rote is employed out of pro-
portion to its importance. Mayer is correct in urging us to consider
dimensions of student learning beyond that implied by mastery of behav-
ioral objectives. "Instructional objectives should be sensitive not only
to what is learned but also to how it is learned and structured in memory.."
This is a much more difficult assignment.
References
Bogatz, C. A. and Ball, S. The second year of Sesame Street: A contin-
uing evaluation. Princeton, N.J.: Education Testing Service, 1971.
Ehrenpreis, W. and Scandura, J. M. The algorithmic approach to curriculum

construction: A field test in mathematics. Journal of Educational
Psychology, 1974, 66, 491-498.
Green, J. G. Cognitive objectives of instruction: Theory of knowledge

for solving problems-and answering questions. In D. Klahr (Ed.),
Cognition and Instruction. Hillsdale, N.J.: Erlbaum, 1976.
Henderson, R. W.; Swanson, R.; and Zimmerman, B. J. Training seriation

responses in young children through televised modeling of hierarchy
sequenced rule components. American Educational Research Journal,
1975, 12, 479-490.
Scandura, J. M. Role of rules in behavior: Toward an operational def-

inition of what (rule) is learned. Psychological Review, 1970, 77,
516-533.
Wertheimer, M. Productive thinking. New York: Harper & Row, 1959.

Rakow, Ernest A.; Airasian, Peter W.; and Madaus, George F. ASSESSING
SCHOOL AND PROGRAM EFFECTIVENESS: ESTIMATING TEACHER LEVEL EFFECTS.
Journal of Educational Measurement 15: 15-21; March 1978.
Abstract and comments prepared for I.M.E. by JANE SWAFFORD,

Northern Michigan University.
1. Purpose
The purpose of the study was to explore the possibility that the
absence of significant differences in comparisons of school or program
effectiveness might be due to varying teacher effectiveness within the
schools or programs. Specifically, the study sought to determine what
proportion. of the variance not accounted for by between-schools differ-
ences could be attributed to differences between-teachers-within-schools
rather than to individual student differences on error variation.
2. Rationale
Large-scale studies of program effectiveness have most often led to
the conclusions that different programs have little differential impact
since within-school (or within-program) variation is largely relative_ta.
between-school (or between-program) variation. However, it is notéd
that in many of these studies, the data have been aggregated across
teachers within each school or program. Hence, variance among teachers
within each school is assigned to individual. pupil or error variation.
This study was designed to investigate the wisdom of ignoring teacher
effectiveness in assessing program or school effectiveness.

Data from the First International Study of Achievement in Mathe-
matics (IEA) were utilized in this investigation. From the American
IEA sample, a subset of 108 schools at the 8th- and 10th-grade and 33
schools at the 12th-grade level were selected. Each school had two or
more teachers who each had at least five students participating in the
IEA study. Using the IEA data, variance estimates and percentages for
school, for teacher-within-school, and for pupil-within-school-and-
teacher were calculated on measures of (1) mathematics achievement,
(2) interest in mathematics, and (3) socioeconomic status.
4. Findings
On the measure of mathematics achievement, the within-school
variance was somewhat higher than the between-school variance at the
three grade levels, ranging from 60 percent to 75 percent of the total
variance. However, the variance between teachers within school was
found to account for approximately 25 percent of the total variance at
the three grade levels and from 30 percent to 40 percent of the within-
school variance. Hence, teachers-within-school did explain much of the
variance in mathematics achievement.
On the measure of interest in mathematics, the' within-school vari-
ance was considerably higher than the between-school variance, ranging
from 83 percent to 92 percent of the total. However, the teacher-
within-school variance was small and rather constant at each grade
level at 6 percent to 8 percent of the total variation. Hence,
teachers-within-school did not explain much of the variation in
students' attitudes
On the measure of socioeconomic index, moderate between-school
variations (22 percent to 25 percent of total) and large within-school
variations (72 percent to, 78 percent of.,._total) were found. However,
teacher-within-school variation accounted only for between 6 percent
and 8 percent of the total variance. Hence, teachers-within-school
did not explain much of the variation in socioeconomic status.
5. Interpretations
The study concludes that a sizable component of the variation
traditionally defined as error variance in studies of school or pro-
gram effectiveness can be associated with the differential effects of
teachers-within-schools.
The use of existing data-pools to investigate questions concern-
ing research design and interpretations, unfortunately, has its perils.
This study both acknowledges and illustrates the problems. A study
designed to answer one set of questions is sometimes poorly designed
to address another. The IEA study was designed to assess international
differences in student achievement, not teacher effectiveness. The
sample was chosen to be representative of the student population in
each country. Information concerning the identity of the student's
teachers was incidentally collected. Information concerning the level
of the student's mathematics course was not. Hence, rather than dif-
ferences in teacher effectiveness, the observed differences between
students with different teachers could reflect differences in the
level of the mathematics course (e.g., General Mathematics versus
Algebra II) or differences in ability grouping for the same course.
While one might suspect that within a school or program there can
be wide variation in the effectiveness of individual teachers, this
study does not substantiate it. It does, however, reiterate the cau-
tion against ignoring the teacher variable in program evaluation and
suggest a research design for estimating the impact of that variable,
guidance was given only if there was complete frustration in the group.
The control groups did the worksheets with the teacher directing the
reading, answering the questions, and giving "direct assistance" to
those with wrong answers.
The groups were compared on an observation checklist and comment
sheet on.
1) whether or not they could complete the task, and
2) how the pupils worked effectively with each other.
The analysis consisted of comparing the number of groups that completed
the slide-rule task without teacher assistance and the pupils' subjec-
tive comments.
4. Findings
1) All the experimental groups completed the slide-rule task; none
of the control groups did.
2) Experimental groups helped and encouraged each other; control
groups did not. Examples of the interactions are quoted in the
article.
5. Interpretations
Students can be helped to work effectively in groups on mathematics
problems. The author states: "Furthermore, while doing this, they are
gaining self-confidence, interest in math, and a more healthy attitude
toward their peers. ... Children's personalities bloom." This doesn't
happen often "while they're learning mathematics."
0. 1) No comparison was made on what content was learned in training
tasks. Did those with group-oriented instruction learn more
than those with teacher direction?
2) Teachers who are deciding on how to use classroom time need to
compare the learning by students in well-trained groups with
good, teacher-oriented, full-class instruction.
3) One rural class of 28 taught by six preservice teachers does not
generalize to most classrooms.
4)" The affective results quoted above in "Interpretations" need to
be verified.
Schoenfeld, Alan H. EXPLICIT HEURISTIC TRAINING AS A VARIABLE IN
PROBLEM-SOLVING PERFORMANCE. Journal for Research in Mathematics
Education 10: 173-187; May 1979.
Abstract and comments prepared for I.M.E. by RICHARD E. MAYER,

University of California, Santa Barbara.
1. Purpose
The author lists three main goals: (1) "tó study the impact of
instructión in five heuristics on some students' performance on a series
of problems"; (2) "to see if other students working the same problems...
but not receiving the heuristics, instruction would use or intuit the
strategies"; and (3) "to see, by,.comparing the two groups, if explicit
instruction in heuristics 'makes a difference'."
2. Rationale
There has been much written about how to, teach students to become
better problem solvers (such as Polya's classic How to Solve It and
Mathematical Discovery), but such work is not closely connected to
research and theory in the psychology of problem solving. Similarly,
there has been much research and theory on problem-solving heuristics
in cognitive psychology (such as Newell and Simon's Human Problem
Solving), but little connection with practical issues in mathematics
education. The present study attempted to bridge this gap by taking
heuristics derived from cognitive psychology and studying their use in
real-world mathematical situations. Thus, the tools of cognitive psycho-
logy--descriptions of the mechanisms underlying problem-solving
performance--are applied to the problems of mathematics education.
The result is an attempt to.determine the "mechanisms" used by mathe-
matical problem solvers and see if these "mechanisms" can be success-
fully taught.

Four subjects served in .the experimental group and three subjects
served in the control group. Subjects were experienced in mathematics
and science, since they were all upper-division mathematics' and science
majors at the University of California,' Berkeley.
All subjects took a five-problem pretest and a five-problem post-
test, with a maximum of 20 minutes allowed for solving each problem.
The problems included proofs, complex algebra story problems, and
series sum problems. All subjects received written and tape-recorded
instruction on how to solve 20 problems (including the five pretest
problems) over a two-week'period.
The independent variable was type of instruction: the experimental
group was given a list and description of five useful strategies, and
explicitly told that a given instructional problem should be solved by
a specific strategy. For experimental subjects, all problems in an
instructional session were solved by the same strategy. The control
group'was given the same set of 20 problems and the same written and
recorded instruction, except that no list and description of the
strategies was given, no explicit mention of the used strategy was
given, and the problems solved in a session were not all of.the Same
strategy type. The dependent variable was the change in score from
pretest to posttest; in addition, thinking-aloud protocols were taken
so that more detailed differences in performance could be determined.
4. Findings
The experimental group increased from an average of 20 percent
correct on the pretest to an average of 65 percent correct on the post-
test, while the control group averaged 25 percent on both tests. In
spite of the extremely low sample size, the differences in change scores
are significant. Analysis of protocols indicated that the control group
did not tend to make efficient use of heuristics, while the experimental
did for several types of heuristics.
5. Interpretations
The author concluded that
a detailed look at the students' solutions to the test
problems supports the results suggested by the statis-
tics: Conscious application of a problem-solving
strategy does make a difference.(pp. 182-3)
Moreover, students can transfer training, at least on some strategies,
from practice problems to posttest problems. The inconclusive results
on other strategies suggests that more work is needed in selecting
and teaching the strategies. However, "the fact that a student knows
how to use a strategy is no guarantee whatsoever that the student will
indeed use it" (p. 183).
Students intuit heuristics minimally. Therefore, students need to
be taught strategies explicitly. Further research is discussed: (a)
"loosening" the format to make it approximate real-world conditions, or
(b) varying the heuristics being studied.
On the importance of being cognitive. This paper shows a great
improvement over previous work on "teaching problem solving" because
the author deals with the mechanisms underlying performance rather
than raw empiricism. This line of research demonstrates that it is
not necessary to return to earlier research techniques in which "method
A" is compared to "method B," a choice for 'which is better' is based
on posttest scores, and no understanding of underlying cognitive mech-
anisms is provided. It is a pleasure to see that the "raw empiricism"
of the past is finally being laid to rest.
This paper is representative of the growing symbiotic relationship
between cognitive psychology and mathematics education. The domain of
mathematics learning and problem solving provides an excellent arena
for development and testing of cognitive theory; and the analytic tools
of cognitive psychology provide an excellent starting point for improve-
ments in mathematics instruction. For example, the recent advances in
our understanding of algebraic problem solving by people like Hayes and
Greeno and of physics problem solving by Larkin and Simon are highly
encouraging. Only mutual benefit can come from the natural give and
take between cognitive psychologists and mathematics/science educators.
All you need is more. The introduction, method section, and
results section of the paper--while clear and well writte n--all could
be improved by providing "more of the same." The introduction seems
to list related research work but makes no effort to describe the
existing work in enough useful detail. The procedure allows for a
nicely controlled study, and the study shows an overall level of tight
experimental control that is sadly lacking in many similar studies.
However, the results are seriously limited by the fact that only seven
subjects were used, and by the fact that all of them were already
highly experienced in mathematics and science. Admittedly, the
results are statistically significant--even when subjected to a
t-test, which the author• did not attempt. However, the importance of
the finding demands that an additional replication be carried out
before the author can make too many claims. Furthermore, less exper-
ienced subjects should be used, since it is they who are likely to be
most in need of strategy training.
The same request for "more of the same" could be made with respect
to the analysis of protocols. The author makes a nice start at'describ-
ing the differences in protocols of subjects in the two groups. How-
ever, much more detail is needed. In particular, there is a need to
tie the protocol analysis to some specific theory of problem solving.
Analyzing protocols can become a trap in which the researcher has so
much specific information that he or she cannot see any general patterns
in the results. The present author has not fallen into this trap, but
there is a need to become more concrete about the general patterns
located in the protocols. A final problem with the protocol procedure
is that "thinking aloud" is not a normal or natural way to solve prob-
lems. It is not possible to determine how this procedure influenced
the "number correct" data for the posttest nor is it possible to
disentangle the effects of "thinking aloud" from the effects of the
experimental treatment.
Can we teach problem solving? This paper deals with an old question
--Can we teach problem solving?--and deals with it in a creative way.
However, there are many pitfalls in "training of problem solving"
research. First, there is the problem of defining problem solving.
This is rarely done. The goal of instruction for problem solving should
be clearly specified. There are many instructional procedures which go
under the label of "teaching for problem solving," but they vary widely
in content. We need an objective definition of what it is that we are
trying to teach.
A second problem concerns the traditional usefulness of instruc-
tion'in general problem solving. A typical finding concerning programs
developed to teach creative problem solving to school children is that
such programs have either little effect or that they improve problem-
solving performance mainly for specifically similar problem-solving
situations (see Mansfield et al., 1978). If this is also the case
for cognitive techniques--the present experiment deals only with near
transfer to very similar problems--this raises the question of whether
it is useful to assume there are any "general" problem-solving tech-
niques in humans independent of specific problem-solving subject areas.
References
Mansfield, R. S.; Busse, T. V.; and Krepelka, E. J. The effectiveness
of creativity training. Review of Educational Research, 1978, 48,
517-536.
Silver, Edward A. STUDENT PERCEPTIONS OF RELATEDNESS AMONG MATHEMATICAL
VERBAL PROBLEMS. Journal for Research in Mathematics Education 10:
195-210; May 1979.
Abstract and comments prepared for I.M.E. by GAIL SPITLER,

University of British Columbia.
1. Purpose
The principal objective of the study was to examine (a) dimensions
of verbal problem similarity perceived by students and (b) the relation-
ship between a student's perception of problem similarity and performance
on tests of verbal and mathematical ability.
2. Rationale
Polya has suggested that when one is solving a mathematical problem,
it is useful to "think of a related problem." Silver argues that one's
ability to use Polya's related problem heuristic is at least partially
dependent upon the criteria which one uses in forming categories of prob-
lems.' The study may be seen as an extension of the work of Chartoff
(1977) and of Hinsley, Hayes, and Simon (1976). The article under
review ís a report of two studies, with the second being a refinement
and replication of the first. The two studies will be discussed separ-
ately.
3. Research Design and Procedures (Study 1)

The sample consisted of 95 students in four intact eighth-grade
mathematics classes taught by the investigator. All of the testing
occurred during the normal mathematics class period. Measures of verbal
and mathematical ability used were The Educational Records Bureau Mathe-
matics Achievement Test (Mathematics Computation, Mathematics Concepts),
Differential'Aptitude Test (numerical ability, verbal ability, abstract
reasoning ability), and Lorge-Thorndike (verbal and non-verbal IQ). The
testing/treatment occurred over nine days. The first day involved the
completion of the Card Sorting Task (CST) developed by the investigator.
The CST consists of 24 verbal problems, on individual cards, which the
student is asked to sort according to mathematical relatedness. The
problems varied systematically along two dimensions: mathematical
structure and contextual detail. The students also supplied written
explanations of their categories. On the second day the students
attempted to solve problems I-8 of the CST. On the third day the solu-
tions to problems 1-8 were discussed. The fourth/fifth days'and the
sixth/seventh days paralleled days two/three considering problems 9-16
and 17-24, respectively. The number of completely correct solutions
obtained by each individual student was recorded as the Problem-Solving
Task (PST) score. On the eighth day the students were asked to complete
the CST again. On the last day the students completed the Problem Rela-
tions Task (REL) developed by the investigator. The REL consists of ten
verbal problems similar to those included on the CST. The students were
asked to read the first REL problem, think about how they would solve
it, and identify the CST problem which was mathematically related. The
procedure was repeated with each of the ten REL problems.
The CST yielded two scores: a dimensional association score (DAS)
and a pure category score. For each dimension the DAS was obtained by
counting the number of pairs of problems, related along the given dimen-
sion, that a student put into the same category. The pure category
score was obtained by counting the number of instances of student-formed
categories containing at least two problems related along the given
dimension and no problems unrelated to the given dimension. The REL
score was obtained by counting the number of instances in which the
student correctly identified that the given REL problem was related to
a CST problem along a given dimension.
4. Findings (Study 1)
Students used four criteria to categorize the CST problems: mathe-
matical structure, contextual detail, question form, and pseudo-structure.
The latter dimension referred to associations based upon the presence of
a measurable quantity such as age, weight, or time. A major difficulty
with the CST problem set was identified in that many of the problems
were related along more than one of the four dimensions. Only the struc-
ture and contextual detail scores were used in further analysis. There
was a marked increase in the structure DAS score and the pure category
score between sorts 1 and 2. For each of the two dimensions, Pearson
product-moment correlations.coefficients were used tó compare perfor-
mance on the CST and REL tasks. The results indicate significant
(p <.005) positive relationships between the REL score for the structure
dimension and the pure category score for each sort, as well as for the
structure DAS score for each sort. For the context dimension signifi-.
cant positive correlations were found between the REL score and the con-
text DAS score, but the correlations between the REL score and each of
the pure context scores were not significant.
The results suggest generally positive correlations between high
performance on the ability measures and sorting based upon structure,- and
negative correlations between ability measures and sorting based upon
contextual detail. Pre-solution sorting on the basis of structure was
significantly (p <.005) correlated with PST performance, even when the
effects of verbal and nonverbal IQ were simultaneously controlled; how-
ever, the association was not significant when the effects of mathematics
concepts knowledge and mathematical computational ability were controlled.
Post-solution sorting on the basis of structure was significantly (p <.05)
correlated with PST performance even when the effects of IQ, concepts
knowledge, and computational ability were simultaneously controlled.
3. Research Design and Procedures (Study 2)

The sample for this study consisted of 58 eighth-grade students
enrolled in one of three intact mathematics classes taught by the
investigator. None of the subjects had participáted in Study 1. The
CST was revised in light of the findings of Study 1 so that few problem
pairs would be related along more than one dimension. All other proce-
dures were the same as Study 1 except that the REL task was not included.
4. Findings (Study 2)
The results of Study 2 support the findings of Study 1. There was
an indication that all four dimensions were used for classifying; there
was an increase in the number of structure associations between sort 1
and sort 2; there was a positive correlation between measures of ability
and sorting according to mathematical structure, and correlations
between the PST and structural sorting were significant even when the
effects of ability were controlled with the same limitations as noted
in Study 1. In addition, Study 2 indicated a generally negative rela-
tionship between the tendency to sort problems on the basis of question
form and performance on tests of nonverbal IQ, concepts knowledge, com-
putational ability, and problem-solving ability. However, the pre-
solution pure pseudo-structure score correlated positively with PST
performance, but not with general mathematical ability variables.
S. Interpretations
The four identified dimensions used by the students to categorize
the problems are similar to the dimensions identified in similar studies.
The findings suggest that students view problem relatedness in different
ways and that there exists a strong relationship between mathematical
ability and the tendency to perceive the,mathematical structure of
verbal problems and to relate problems on the basis of perceived struc-
tural similarities. The findings concerning the use of pseudo-structure
categories suggest that this category was used by good problem solvers
during the initial sort and that some less able problem solvers increased
their use of this category on the second sort. The need for more
research in this area was noted.
The one contribution made by this study is the identification of
four categorizing strategies; however, as the investigator acknowledges,
one must be careful not to generalize these results, as the problem sets
were carefully constructed to elicit the recognition of certain pre-
determined dimensions. Having acknowledged this contribution, I must
also state that I have many reservations about the study, both in the
conceptualization of the probleti and in specific detail. These concerns
are listed below.
1. Aside from the identification of the four categorizatión styles,
I find the other results of the study self-evident. One of the major
findings of the study is that the sorting of problems according to mathe-
matical structure is positively correlated with various measures of
ability. In fact, would one not define ability and/or intelligence so
as to include the ability to perceive structure in events or situations?
Another major finding of the study is that the PST score positively
correlated with structural sorting even when various measures of ability
and knowledge were controlled. While it is always satisfying to obtain
statistically significant results with such controls, this result also
seems self-evident. Students who perceived the structure of a problem
were able to solve it.
2. It is important to recognize that an underlying assumption of
this study is that the recollection of a related problem is somehow
related to the way in which said problem is initially categorized by
the student. This may or may not be true.
3. The study could have been strengthened. in one or all of the
following ways:
a) by the inclusion of some measure of general information
sorting. As the investigator points out, evidence is available
from a wide variety of fields concerning the perception of common-
ality of stimuli. Many cognitive style constructs are at least
partially defined in terms of the type of categorizing style
employed by the subject. If this study had included some measure
of cognitive style it might have shed some light on the question
of whether the sorting behavior was unique to mathematics or
whether it is a generalized behavior that appears in a variety
of situations.
b) by the inclusion of some measure of the retention of the
problems and the system by which it was categorized. This would
have helped to address the problem of whether initial perceptions
influence later recollection of problems.
c) by the inclusion of problems for which the students did
have algorithms appropriate for the solution of the problem. The
CST set included only problems for which the students did not have
sufficient algorithms for the solution thereof. Do students sort
such problems in the same way that they would sort solvable prob-
lems?
4. The report itself lacks detail in several critical areas. This
may not be the fault of the author. It may be that the brevity and
succinctness required for a journal article is such that the necessary
detail cannot be included. If such is the case, one must raise, ques-
tions about the purpose of such journals. Specifically the report lacks
detail concerning the CST and REL.
a) With respect to the CST, there is no indication as to which
specific mathematical structures were used, why certain structures
were Selected, how the students performed on the CST (frequency of
use of various categories and whether students tended to use one
strategy over the others), and how many problems in the revised
CST actually overlapped along one or more of the dimensions. The
investigator does give some clues concerning the frequency of use
of the various dimensions, but more detail is needed.
b) There is no indication given as to why the REL task was
not included in Study 2.
In summary, I must acknowledge the contribution made by this study
but I have many reservations about the overall report. In fact, one
wonders why the results of Study 1 were included at all, as Study 1
was more a pilot study for Study 2 than an independent experiment. I
heartily concur with one of the investigator's closing comments that
this study raises a myriad of questions for further research, but I
must add that t doubt that it has answered very many questions.
Thornton, Carol A. EMPHASIZING THINKING STRATEGIES IN BASIC FACT
INSTRUCTION. Journal for Research in Mathematics Education 9: 214-
227; May 1978.
Abstract and comments prepared for I.M.E. by JOSEPH N. PAYNE,

University of Michigan.
1. Purpose
The purpose was to explore the effect of teaching thinking strate-
gies for basic facts in grades two and four.
2. Rationale
The results from studies done earlier by Thiele (1938) and Swenson
(1949) were given as reasons to conjecture improvement in basic fact
learning through the explicit teaching of strategies. Recent studies
and recommendations by Rathmell (1978), Ashlock (1971), Hall and
Trafton (1974), Jerman (1970), and Myers and Thornton (1977) were cited
also. The author noted the current practice of not using strategies in
mathematics classrooms.

Intact groups from multi-level classes in grades 2 and 4 were
pooled to form two groups, experimental (E) and traditional (T). For
E, n - 25 in Grade 2 and n - 23 in Grade 4; for T, n - 22 at Grade 2
and n - 20 at Grade 4. Random assignment of treatments to groups was
made. Mean I.Q. for the two groups was slightly above average, about
110.
There were eight weeks of instruction, beginning in September, for
20 minutes per day, three days per week. Regular classroom teachers
were trained and taught the classes. In grade 2, E subjects were taught
these strategies for addition: doubles, sharing (compensation) to
relate facts to doubles, adding to 9, counting on and using 10. For
subtraction E subjects were prompted to think of the addition fact.
In grade 4, E subjects were taught these multiplication strategies:
patterns, relationships, and other techniques (e.g., twice as much,
adding on, subtracting from, finger multiplication for 9s,
commutativity). For division, stùdents were encouraged to think of
the multiplication fact.
The instruction on facts for T-groups was based on the sequence
in the adopted textbook with supplementary teacher-directed drills.
Mathematics time for both T and E outside the fact-instruction time
was on content topics such as geometry, time, or metric measurement. .
The dependent measure was the number of facts correct on identi-
cal pretest, posttest, and retention test (two weeks) with 3-minute
written timing. Data were analyzed using ANOVA with a general 2x 3
design. Scheffe tests for pairwise comparisons and t tests were
carried out subsequently. Results for "harder facts" on addition in
grade two and on multiplication in grade 2 were analyzed separately
but in the same manner. The upper and lower thirds of each group were
interviewed after the retention test in mid-November and again in May.
4. Findings
There was a marked effect in favor of E:
(a) In grade 2, p <.001 on the posttest and. the retention test.
on both addition and subtraction.
(b) In grade 4, p <.001 on the posttest in multiplication, p <.05
on division posttest and p <.05 on retention tests for both
multiplication and division.
In grade 2, actual scores for T declined on subtraction from the pretest
and gains in addition for T were small, about 7. In grade 4, E gained
about 50 on multiplication while T gained about 30. On division, E
gained 49 and T gained27. Gains on harder facts for addition and
multiplication were similar. E gained it on the posttest in grade 2
while T declined by 1. On multiplication, E gained 25 while T gained 9.
From the interview data it was discovered that about three-fourths
of students in E used the strategies that were taught explicitly. About
one-third of the students in T used strategies; the most evident stra-
tegy was the usual one of counting using a ruler or fingers. For facts
mastered during the eight-week period, retention of strategies used
earlier was more consistent for high and for low achievers.
5. Interpretations
The data support the use of thinking strategies to facilitate
learning of basic facts. A discussion of the poor performance of T
in grade 2 suggested that confidence of pupils in their ability to
do the facts was a dodlinant influence. The aithor recommended that
curriculum and classroom efforts focus more carefully on the develop-
ment of thinking strategies prior to drill on basic facts.
The study holds substantial significance in mathematics education
for several reasons:
(1)It reminds the mathematics education profession that older
substantial studies need to be reexamined for replication
and useful implications for problems evident in today's
schools.
(2)It addresses one component of the popular return to the basics
with the need to "know the facts."
(3)It provides evidence on the effectiveness of thinking and
reasoning even on skill-oriented topics.
With a shift in the way facts are organized and taught, there appears
to be a relatively easy way to improve achievement. Furthermore, while
the evidence in this study is not as direct as one would like, the con-
fidence of pupils who are taught the strategies seems to be strengthened.
The results found by the investigator on addition and multiplica-
tion seem quite dependable. The results for subtraction and division
bear some skepticism because of the lack of specificity about the time
the strategies were used and the duration of practice on these strate-
gies. Furthermore, it was not clear just how the strategies for sub-
traction and multiplication were related to addition and multiplication,
respectively.
The interview data on facts that were mastered suggest that such
an experiment should extend over a longer period of time, perhaps a full
school year, under a mastery-learning scheme.
Children seemed to follow the advice of skipping facts on which
they were unsure. It needs to be determined if children do this as
easily on a pretest as on a posttest because there is some possibility
that the substantial gains could have some spurious component. If
children are more confident, they may be more willing to skip.
There versa few minor irritations in reading the research report.
It took several;readings to discover that the major dependent measure
was the "number" of facts completed correctly in three minutes. In
paragraph two, page 219, there is a statement on subtraction contrary
to the data. The T1, T2, and T3 used in Table 4 were not defined. It
takes some study to realize that they probably refer to pretest, post-
test, and retention test. In lb, page 225, "counting on from the
greater addend" was used as a general description of strategies for
E even though the reference is to both addition and multiplication.
An initial figure on the sequence of events lists a Transfer Period
with Interim Tests yet there were no data and no discussion of this
part of the study.
These minor inconsistencies should not detract from a very useful
piece of work. It is the kind of study that is easy to replicate.
With each replication there is likely to be good in-service for
teachers as well. This is a happy combination indeed.
Vance, J. H. ATTITUDES TOWARD MATHEMATICS AND MATHEMATICS INSTRUCTION
OF PROSPECTIVE ELEMENTARY TEACHERS. Alberta Journal of Educational
Research 24: 164-172; September 1978.
Abstract and comments prepared for I.M.E. by THOMAS COONEY,

University of Georgia.
1. Purpose
The purpose of the present study was to compare the attitudes
toward mathematics and the beliefs about mathematics and mathematics
instruction of prospective elementary teachers at three stages of
preparation at thê University of Victoria. The attitudes and beliefs
of high and low achieving students, students enrolled in primary and
intermediate sections of a methods course, and students enrolled in a
methods course with and without a particular content eburse as a pre-
requisite were also investigated.
2. Rationale
A basic assumption of the study is that the attitudes and beliefs
of elementary school teachers toward mathematics affects the way in
which mathematics is taught. Previous research indicates that pre-
service elementary teachers tend to have more of an informal view of
mathematics after their training in mathematics education than before
the training.

Three Likert-type scales were used. The Attitude Toward Mathe-
matics Scale (ATMS) measured interns' positive and negative feelings
toward mathematics. The Belief About Mathematics Scale (BANS) and The
Belief About Mathematics Instruction Scale (BAMIS) were designed to
measure a formal-informal dimension of belief.. "A high score on the
BANS indicates an appreciation of the creative and investigative
nature of mathematics, as opposed to a view that mathematics is a
closed, rigid subject. A high score on the BAMIS suggests a strong
belief in the importance of student exploration and discovery, as
opposed to teacher explanation, in learning mathematics."
The scales were administered to various groups of prospective
elementary teachers enrolled at the University of Victoria in the
1973-74 and 1974-75 academic years. Subjects were classified in two
ways: by preparation and by achievement. The following three levels
of preparation were defined:
i Beginning of math course for prospective elementary teachers
ii Completion of same course
iii Completion of mathematics methods course
Two levels of achievement were defined: high and low. The levels
were based on high school or university mathematics courses completed
and on grades received.
A 3 x 2 (preparation by achievement) design was used to compare
scale scores of independent groups at the three stages and two achieve-
ment levels. Subjects per cell were 21, 28, 34, 42, 25, and 28. "At
each stage, only those subjects who had responded to the scales for the
first time at that stage were included in the sample. Some students
were tested at two different stages. Data were obtained for 46 subjects
who completed the scales prior to and upon completionsbf the mathematics
course in 1974-1975 and for 27 subjects who completed the scales upon
completion 1973-1974 and upon completion of the methods course in 1974-
1975. Pre- and posttest scores were compared using the correlated
t-test."
4. Findings
Students began the mathematics course with fairly positive atti-
tudes toward mathematics. The attitudes of these students did not
differ significantly from those who had completed the course or from
those who had completed both the content and the methods course.
Students who completed both the content and methods courses had a more
informal view of mathematics than entering students or students who had
completed just the content course. Students began the content course
with neutral beliefs about mathematics instruction. Students developed
a more informal view of instruction upon completion of the mathematics
course and still more so after completion of the methods course. High-
achieving students bad more favorable attitudes and a more informal view
of mathematics instruction than low achievers.
Analysis of data from students who completed the scales at two
successive stages of preparation also resulted in several findings. •
Students who successfully completed the mathematics course tended to
develop a more favorable attitude and a more informal view toward
mathematics. Successful completion of the methods course tended to
promote. positive attitudes toward mathematics and informal beliefs
about mathematics and mathematics instruction.
5. Interpretations
It appeared to the investigator that the mathematics education
program was reasonably effective in developing desirable attitudes
toward mathematics and mathematics instruction. High achievers tended
to be more positive that low achievers. This is probably because
teachers with a poor knowledge of mathematics ate more likely to empha-
size drill activities and feel more constrained in promoting students'
exploration in mathematics. Transfer students and other students who
had another mathematics course besides the one of interest here weré
found to have similar attitudes and beliefs upon completion of the
methods course to those who had taken the standard course.
Some results seemed contradictory. One analysis indicated that
attitudes of students at the three preparatory stages did not differ
significantly. Yet it was reported that students who took the instru-
ments on a "pre-post" basis developed a more favorable attitude. No
explanation was given for the apparent conflict.
Still it is "comforting" to think that mathematics teacher educa-
tion programs can have a positive effect on interns. However, the
findings should be viewed in the context of other questions. To what
extent do changes in attitudes and beliefs affect preactive or inter-
active phases of teaching? Do'the changes in any way transcend beyond
paper and pencil changes? If so, how? What specific aspects of a
methods course are more likely to influence changes in attitudes and
beliefs?
In fairness, these questions were not the concern of the present
investigation. However, unless we know more about how attitudes and
beliefs affect a teacher's performance and come to understand how
changes in a teacher's effect occur, we hive not truly gained much
knowledge. Also, there is the additional'concern that the instruments
used in this and other similar studies provide only a rather narrow
perspective on teacher effect and affect.
In closing, I ,found the article rather interesting and informative.
But unless the basic problems of teacher education are dealt with in a
more comprehensive manner, progress in upgrading our teacher education
programs is not likely to be very fast coming.
Webb, Norman L. PROCESSES, CONCEPTUAL KNOWLEDGE, AND MATHEMATICAL PROB-
LEM-SOLVING ABILITY. Journal for Research in Mathematics Education 10:
83-93; March 1979.
Abstract and comments prepared for I.M.E. by LEN PIKAART,

Ohio Unibersity.
1. Purpose
The study was designed to investigate the relationship and relative
importance of (a) conceptual knowledge variables and (b) process variables
in mathematical problem solving.
2. Rationale
Conceptual knowledge may be envisioned as the facts, concepts, prin-
ciples, and algorithms that are needed to solve problems, whereas process
refers to the techniques, strategies, or heuristics needed to recall and
construct information while solving problems. Referring to an analysis
by Mayer, Webb identifies the following three stages of a learning process:
reception of material, existence Of prerequisite knowledge (conceptual
knowledge)'and activation of assimilative set (process). Hi notes the
difficulty of attempting to quantify the relative contribution to problem
solving of conceptual knowledge and process because each construct is de-
pendent on the other. That is, problem solving which is dependent on pro-
cess in one instance may be part of learning and become conceptual know-
ledge in a later instance.

A sample of forty high school students taking second-year algebra was
selected from a group of volunteers in four schools that represented dif-
ferent social and economic areas. Later, three subjects had to be elimi-
. ,nated.
A pretest, administered at each school, consisted of 16 scales of
mathematics achievement, attitudes, mental abilities, and problem solving.
All but one scale was selected from the National Longitudinal Study of
Mathematical Abilities. Later, students were interviewed individually
and tape recorded as they solved, thinking aloud, eight problems in a
test called the "Problem-Solving Inventory." The following three sets of
data were used to code the problem-solving protocols:
a)Heuristic strategies. Eighteen variables, such as "draws a rep-
resentative diagram," were identified to represent problem-solving
processes. (Originally "counting, algebraic, and other errors"
were classified as checklist variables, but later these three var-
iables were reclassified as process-sequence variables.) These
variables were checked only once if the process was observed on
each problem.
b)Process-sequence variables. Fifteen variables, such as "reads
the problem" or "uses trial and error" were noted each time the
student employed the process.
c)Score variables. Success in problem-solving was measured by a
total score sranging from 0 to 5 on each problem with three sub-
parts--approach (0-1), plan (0-2), and result (0-2).
Principle component analyses were performed on the 16 pretest scales,
the 18 heuristic strategies scales, and the 15 process-sequence variables,
reducing the number of variables to 4, 5, and 4, respectively. These con-
structs accounted for from 61% to 70% of the total variance in each set
and were identified as follows:
Pretest Heuristic Strategies Process-Sequence

Components Components Components
Mathematical Achievement Pictorial Representation Deductive Production
Verbal Reasoning Product Checking Random Production
Field-dependence Concrete Representation Non-production
Negative Anxiety Récall Recall Production
Sudden Insight
Finally, regression analyses were performed using the total score of

the Problem-Solving Inventory as the'dependent variable and changing tfie
order of entering the independent variables above.
4. Findings and Interpretations

Tables of the regression analyses are presented depicting the results
when Pretest Components were entered first and when Heuristic Strategies
Components were entered first. However, when all components were intro-
duced stepwise, the sequence of selection drew from both sets of components.
Multiple R2 reached 0.72 with 6 components, 3 Pretest and 3 Heuristic
Strategies. "The large overlap of the components in accounting for vari-
mince in problem-solving scores provides evidençe that conceptual know-
ledge and processes are interrelated."
The study is an important link in understanding the problem-solving
process and should be of interest to all mathematics educators. It is
well conceptualized, designed, and reported. Webb is well aware that it
has limitations--subjects "...were not randomly selected to represent a
particular population, ... the small number of problems prohibited making
generalizations across a large number of problem-solving situations,"
the talking-aloud method may not have indicated all strategies being em-
ployed, the pretests were general measures of abilities rather than spe-
cific measures of conceptual knowledge needed to solve the problems, and
the scoring technique for the dependent variable may have influenced the
results.
It is important to note that the regression analyses lead to opti-
mistic results. The multiple correlation is a statistic [hit mathemati-
cally optimizes the common variance between two sets of variables. That
is, at best about 74% of the variance in the Problem-Solving Inventory
can be accounted for by the other variables in the study.
Still, is an exploratory study, it presents evidence that both con-
ceptual knowledge and process variables are important components of the
ability to solve problems. It is to be hoped that other investigations
will be conducted to explore the relationship in other samples of a pop-
ulation and to explore the effects of instructional treatments on the
interrelation of these components.
MATHEMATICS EDUCATION RESEARCH STUDIES REPORTED IN JOURNALS AS INDEXED
BY CURRENT INDEX TO JOURNALS IN EDUCATION`.
October - December 1979
EJ, 202 556 Halford, G. S. An Approach to the Definition of Cognitive

Developmental Stages in School Mathematics. British Journal of
Educational Psychology, v48 n3, 298-314, November 1978.
EJ 202 875 Nielsen, Linda. Feminism and Factoral Analyses: Alleviating

Students' Statistics Anxieties. College Student Journal, v13 nl,
51-56, Spring 1979.
EJ 203 066 Brandt, Ron; And Others. What It All Means. Educational
Leadership, v36 n8, 581-85, May 1979.
EJ 203 523 Shannon, A. G.; Clark, B. E. Mathematical Attitudes of Some

Polytechnic Students. British Journal of Educational Technology,
v10 nl, 59-68, January 1979.
EJ 203 712 Kurtz, Barry; Karplus, Robert. Intellectual Development

beyond Elementary School VII: Teaching for Proportional Reasoning.
School Science and Mathematics, v79 n5, 387-98, May-June 1979.
EJ 203 714 Cohen, Martin P. Scientific Interest and Verbal Problem

Solving: Are They Related? School Science and Mathematics,
v75 n5, 404-08, May-June 1979.
EJ 203 752 Burnett, James. Self-Paced Fortran. Educational Research

and Methods, vll n2, 53-56, 1979.
EJ 203 986 Sherman, Julia. Predicting Mathematics Performance in High

School Girls and Boys. Journal of Educational Psychology, v71 n2,
242-49, April 1979.
EJ 205 403 Wang, C. Y. Mathematics in Biomedicine. American Mathemat-

ical Monthly, v86 n6, 498-502, June-July 1979.
EJ 205 439 Padilla, Michael J.; Smith, Edward L. Experimental Results

of Teaching First Grade Children Strategies for Nonvisual Seriation:
Journal of Research in Science Teaching, v16 n4, 339-45, July 1979.
EJ 205 442 Treagust, David F. Comments on "The Acquisition of Proposi-

tional Logic and Formal Operational Schemata during the Secondary
School Years." Journal of Research in Science Teaching, v16 n4,
363-67, July. 1979.
EJ 205 498 Kolata, Gina Bári. Institute Idea Divides Mathematicians.'

Science, v205 n4405, 470-72. August 1979.
EJ 206 963 Suydam, Marilyn N.; Weaver, J. F. Research on Mathematics

Education Reported in 19/8. Journal for Research in Mathematics
Education, v10 n4, 241-320. July 1979.
EJ 206964 Bishop, Alan J. Visualising and Mathematics in a Pre-
Technological Culture. Educational Studies in Mathematics,
v10 n2, 135-46, May 1979.
EJ 206 965 Austin, J. L.; Howson, A. G. Language and Mathematical

Education. Educational Studies in Mathematics, v10 n2, 161-97,
May 1979.
EJ 206 997• Sherzer, Bill. Where Are the Computers? Computing Teacher,
v6 n4, 6-9, May 1979.
EJ 207 063 Sherrill, James M. Subtraction: Decomposition versus Equal

Addends. Arithmetic Teacher, v27 nl, 16-17, September 1979.
EJ 207 064 Wheatley, Grayson H.; And Others. Calculators in Elementary

Schools. Arithmetic Teacher, v27 nl, 18-21, September 1979.
EJ 207 078 Swadener, Marc. What Mathematics Skills Hiring Officials

Want. Mathematics Teacher, v72 n6, 444-47, September 1979.
EJ 207 086 Muller, David. Perceptual Reasoning and Proportion. Math-

ematics Teaching, n87, 20-22, June 1979.
EJ 207 088 Preece, Muriel. Mathematics: 'The Unpredictability óf Girls?

Mathematics Teaching, n87, 27-29, June 1979.
EJ 207 107 Dickson, Linda. A Case Stúdy Based on the Mathematical

Achievements and Experiences_of Ten London Transport Craft Appren-
tices. International Journal of Mathematical Education in Science
and Technology, v10 n2, 251-78, April-June 1979.
EJ 207 154 Tebbutt, Maurice J. The Growth and Eventual Impact of Cur-
riculum Development Projects in Science and Mathematics. Journal
of Curriculum Studies, v10 nl, 61-73, January-March 1978.
EJ 207 155' Good, Thomas; And Others. Curriculum Pacing: Some Empiri-
cal Data in Mathematics. Journal of Curriculum Studies, v10 nl,
75-81. January-March 1978.
MATHEMATICS EDUCATION RESEARCH STUDIES REPORTED IN RESOURCES IÑ EDUCATION
October - December 1979
ED 170 994 Rittenhouse, Robert K.; Spiro, Rand J. Conservation in $Deaf

and Normal-Hearing Children. 12p. M7O1/PC01 available from
EDRS.
ED 171 324 Park, Ok-Choon; Tennyson, Robert D. Adaptive Design Strate-

gies for Selecting Number and Presentation Order of Examples in
Coordinate Concept Acquisition. 29p. MF01/PCO2 available from
EDRS.
ED 171 421 Ohe, Pia. How Do Young Children Learn Geometric Concepts.
19p. MF01/PCO1•available from EDRS.
ED 171 514 Lindvall, C. Mauritz; Ibarra, Cheryl Gibbons. The Relation-

ship of Mode of Presentation and of Schodl/Community Differences to
the Ability of Kindergarten Children to Comprehend Simple Story
Problems. 13p. MF01/PC01 available from EDRS.
,ED 171 515 Begle, E. G. Critical Variables in Mathematics Education:

Findings from a Survey of the Empirical Literature. 165p. Docu-
ment not available from EDRS.
ED 171 517 - Johnson, Carl S.; Byars, Jackson A. Influences on Mathemat-

ical Preparation of Secondary School Teachers of Mathematics. 13p.
MF01/PCO1 available from EDRS.
ED 171 518 Mullis, Ina V. S. Effects of Home and School on Learning

Mathematics, Political Knowledge and Pblitical Attitudes. 72p.
MF01/PCO2 available from EDRS.
ED 171 520 Hector, Judith H. Using a Calculator to Teach Fraction Com-

putation in Basic Arithmetic: Research and Observations. 9p.
MFO1/PC01 available from EDRS.
ED 171 523 Evaluation of Teaching and Learning Models for Mathematics

and Reading. Final Report. 24p. MF01/PC01 available from
EDRS.
ED 171 524 Pullman, Howard W. Cognitive Structure and Performance in

Mathematics. 27p. MFO1/PCO2 available from EDRS.
ED 171 525 Pullman, Howard W. The Relation of the Structure of Language

to Performance in Mathematics. 23p. MFO1/PCO2 available from
EDRS .
ED 171 529 Williams, S. Irene; Jones, Chancey-O. A Survey of the Use
of Hand-Held Calculators in Advanced Placement Calculus Courses.
College Board Program Report P78-1. 26p. MFOIÎPCO2 available
from EDRS.
ED 171 532 Ben-Baruch, Nehemyah. Measurement of the Readability of
Hebrew Language Mathematics Textbooks. 96p. MF01/PC04
available from EDRS.
ED 171 536 Schoenfeld, Alan H. Presenting a Model of Mathematical Prob

lem Solving. 20p.'"MFO1/PC01 available from EDRS.
ED 171 537 Schultz, Karen A. Visual-Spatial Operations and the Develop-

ment of Problem Solving Abilities. llp. MFO1/PCO1 available from
EDRS.
ED 171 539 Berliner, David C. Allocated Timel Engaged Time andbAcademic
Learning Time ih Elementary School Mathematics Instruction. 25p.
MFO1 available from EDRS. PC not available from EDRS.
ED 171 540 Easley, J. A., Jr. Mathematical Foundations of Forty Years

of Research on Conservation in Geneva. Revised Edition. 28p.
ED 171 542 . Sherman, Alan. The Effect of a Community College Education

on the Cognitive Development of Students in Liberal Arts and Labor-
atory Technology Curricula: An Intervention Study. 22p. MFO1
available from EDRS. PC not available from EDRS.
ED 171 544 Davis, Robert B. Conceptualizing the Structures Underlying

Cognitive Behavior - The Usefulness of "Frames". 19p. MFO1/PCO1
available from EDRS.-
ED 171 548 Sachar, Jane; And Others. Error Analysis on Literals and
Numerals in Solving Equations. 10p. MFO1/PC01 available from
EDRS.
ED 171 549 Ebmeier, Howard; Good, Thomas L. An Investigation of the
Interactive Effects Among Student Types, Teacher Types, and Instruc-
tional Types on the Mathematics Achievement of Fourth Grade Students.
63p. MF01 available from EDRS. PC not available from EDRS.
ED 171 550 House, Peggy A. Mathematics Anxiety and the Minnesota Tal-
ented Youth Mathematics Project. 24p. MFO1 available from EDRS.
PC not available from EDRS.
ED 171 551 Davis, Robert B. Error Analysis in Nigh School Mathematics.

Conceived as Information-Processing Pathology. 24p. MFO1/PC01
ED 171 552 Fox, Lynn H.; Brody, Linda E. Sex Differences in Attitudes
and Course-Takingin Mathematics Among the Gifted: Implications
for Counseling and Carëer Education. 27p. KFO1/PCO2 vailable
from EDRS.
ED 171 553 Davis, Robert B.; McKnight, Curtis C. Towards Eliminating
"Black Boxes"; A New Look at Good vs. Poor Mathematics Students.
2Op. MF01/PC01 available from EDRS.
ED 171 560 Mathison, Marjorie A. Interventions in Math Anxiety for
Adults. llp. MF01/PC01 available from EDRS.
ED 171 561 Suydam, Marilyn N., Ed. Investigations in Mathematics

Education, Vol. 12;.No. 2, Spring 1979. MF01/PC03 available.
from EDRS.
ED 171 565' Morris, Charles J.;, Bowling, J. Michael. Math Confidence'
and Performance as a Function of Individual Differences in Math
Aptitude. 13p. MF01/PC01 available from EDRS.
ED 171 568 Stover, Georgia. An Analysis of the Changes Made by Sixth

and Eighth Grade Students When. Reviewing Arithmetic Word Problems.
41p. MFO1/PCO2 available from EDRS.
ED 171 569 Mitchell, Charles E. Children's Performance on Selected
Addition and Subtraction Situations Involving the Existence and
Non-existence of Whole Number _ Solutions: A Report of Initial
Piloting. '14p. MF01/PC01 available from EDRS.
ED 171 572 Suydam, Marilyn N. Calculators: A Categorized Compilation

of References. 188p. MF01/PC08 available from EDRS.
ED 171 573 Suydam, Marilyn N. The Use of Calculators in Pre-College
Education: A State-of-the-Art Review, 21p. MF01/PC01 available
from EDRS.
ED 171 576 Padilla, Michael J.; 011ila, Lloyd. The Effect of Small
Group Teaching on Acquisition and _Transfer of Nonvisual Seriation
Abilities. 20p. MF01/PC01 available from EDRS.
ED 171 584 Friend, Jamesine E. Column Addition Skills. 52p. MFO1/ .

PC03 available from EDRS.
ED 171 585 Suydam, Marilyn N. Investigations with Calculators: Abstracts

and Critical Analyses of Research. Supplement. 38p. MF01/PCO2
ED 171 587 Brandau, Linda I.; Dossey, John A. Processes Involved in

Mathematical Divergent Problem Solving. 32p. MF01 available
from EDRS. PC not available from EDRS.
ED 171 588 Brandau, Linda. Processes and Heuristics Involved in Math-

ematical Divergent Problem Solving: A Further Analysis. 31p.
MFOI/PCO2 available from EDRS.
ED 171 589 Weide, Bruce. W. Statistical Methods in Algorithm Design

and Analysis. 186p. MF01/PC08 available from EDRS.
ED 171 590 Lester, Frank K., Jr. A Procedure for Studying the Cognitive
Processes Used During Problem Solving: An Exploratory Study. 18p.
ED 171 781 Oxford, Rebecca L.; And Others. Making Room for Foster
Children: A Criterion-Referenced Approach to ESEA Title I
Migrant Program Evaluation. 28p. MFOl/PCO2 available from
EDRS.
ED 171 783 Schmidt, William H. Measuring the Content of Instruction.

Research Series No. 35. 19p. MFO1/PC01 available from
EDRS.
ED 171 798 Roderick, Stephen A.; And Others. Evaluation of a Success-

ful Remedial Summer School Program. 36p. MFO1/PCO2 available
from EDRS.
ED 172 776 Piele, Donald T. Microcomputers "Coto" School. 13p. MFO1/
PCO1 available from EDRS.
ED 172 893 Salmond, Lowell; And Others. Annual Report of the Develop-
mental Mathematics Curriculum Committee of the American Mathematical
Association of Two Year Colleges: Report of Subcommittee Activity
for 1977-1978. 26p. MFO1/PCO2 available from EDRS.
ED 173 069 Schultz, James E.; And Others. A Comparison of, Alternatives
and an Implementation of a Program in the Mathematics Preparation
of Elementary School Teachers. Final Report. 92p. MFO1/PC04
ED 173 080 Chalupsky, Albert B.; And Others. The Canadian Experience:
Implications for Metric Conversion in Education. 128p. MFO1/PC06
ED 173 081 Sandman, Richard S. Factors Related to Mathematics Anxiety

in the Secondary School. 15p. MFO1 available from EDRS. PC
not available from EDRS.
ED 173 113 Fuson, Karen C., Ed.; And Others.. Explorations in the Model-
ing of the Learning of Mathematics. 249p. MFO1/PC10 available
from EDRS.
ED 173 122 Butler, Martha L. The Effects of the Summer Institute: "New
Career Options for Women Through Mathematics" on Mathematical, Social,
Academic and Career Anxieties and on Sex Role Attitudes. 32p. MFO1/
PCO2 available from EARS.
ED 173 124 Williams, Barbara I.; And Others. Math Tutoring Study, Study
II: Paraprofessionals as Tutors. 295p. MFO1/PC12 available
from EDRS.
ED 173 143 Whitted, Betty Lou. A Comparative Study of Needs and Attitudes
of Parents and Educators in Regard to the Metric System. 180. MFO1/
PC08 available from EDRS.
ED 173 153 Wolcott, Diane Marie. Selected Concrete Logical Piagetian

Tasks as They Relate to Mean and Item Scores on the Math Conceit
Section of the Stanford Math Achievement Test, Form A. 73p. MFO1
ED 173 154 Annis, Linda. The Effect of Cognitive Style and Learning
Passage Organization on Study Technique Effectiveness. 8p.
ED 173 173 Wagner, William J. Application of Stochastic Learning Theory

to ElementaryArithmetic Exercises. Technical Report No. 302.
Psychology and Education Series. 131p. MFO1 available from EDRS.
PC not available from EDRS.
ED 173 174 Kuhs, Therese; And Others. A Taxonomy for Classifying Ele-
mentary School Mathematics Content. Research Series No. 4. 71p.
ED 173 175 Elmore, Patricia B.; Vasu, Ellen S. A Spectrum Analysis of

Attitudes Toward Mathematics: Multifaceted Research Findings. 16p.
MFO1/PCO1 available from EDRS.
ED 173 177 Wong, Shirley Mabel. The Secondary Consumer Mathematics Cur-
riculum in H.C.-Present and Proposed. 58p. MFO1/PC03 available
from EDRS.
ED 173 360 Williams, John D.; And Others. Stages of Piagetian Tasks
and Spatial Relations in University Students. llp. MFO1/PCO1
ED 173 657 Motivating Girls to Prepare for Math-Related Occupations.

Final Report. 82p. MFO1/PC04 available from EDRS..
ED 173 739 Jordan, Valerie Barnes; Jensen, C. Mark. Summary and Review:
Correlational Studies of Measures of Piagetian Cognitive Development
and Mathematics Achievement. 16p. MFO1/PCO1 available from
EDRS.
ED 174 395 Stroup, Kala; Jasnoski, M. L. Do Talented Women Fear Math?
10p. MFO1 available from EDRS. PC not available from EDRS.
ED 174 400 Wilson, James W., Ed.; And Others. Journal for Research in
Mathematics Education, Vol. 9, No. 4, July 1978. 85p. MFO1
ED 174 408 Alternatives for Slow Learners in Secondary School Math. 66p.
Ed 174 455 Powell, William R.; Bolduc, Elroy J. Indicators for Learning
and Teacher Competencies in the Basic Skills: Mathematics. Florida
Educational Research and Development Council, Inc., Research Bulletin,
Volume 13, Number 1, Summer 1979. 42p. MFO1 available from EDRS.
PC not available fróm EDRS.
ED 174 456 East, Phillip; Moursund, David. Calculators b Computers in
the Classroom. 67p. MFO1/PC03 available from EDRS.
ED 174 477 Boswell, Sally L. Nice Girls Don't Study Mathematics: The
Perspective from Elementary School. MF01/PC01 available from
ERRS.
ED 174 491 Dorf, Richard C. Final Report on National Science Foundation
"Women in Engineering Program," July 1, 1976-July 8, 1978. 110p.
MF01/PC05 available from EDRS.
ED 174 492 McKnight, Curtis C. The Use of "Frames" in Conceptualizing

Learning and Problem Solving in Mathematics. 34p. MFO1/PCO2
ED 174 663 Sternberg, Robert J. The Development of Linear Syllogistic

Reasoning. Technical Report No. 5. Cognitive Development Series.
31p. MF01/PCO2 available from EDRS.

Ed 187577

Uploaded by

Copyright:

Available Formats

Ed 187577

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ed 187577

Uploaded by

Copyright:

Available Formats

DOCUMENT RESUME

ED 187 577 SE 031 006

Editor Advisory Board

Marilyn N. Suydam Edward M. Carroll

Margaret L. Kasten Thomas E. Kieren

The Center for Science and Mathematics Education

With the cooperation of theERIC; Science, Mathematics, and Environmental

Volume 13, Number 2 - Spring 1980

Subscription Price: $6.00 per year.. Single Copy Price: $1.75

INVESTIGATIONS IN MATHEMATICS EDUCATION

Bright, George W. IDENTIFICATION OF EMBEDDED FIGURES BY

Good, Ron. CHILDREN'S ABILITIES WITH THE FOUR BASIC

Greenstein, Jane and Stráin, Phillip S. THE UTILITY OF

Kieren, Thomas E. and Nelson, Doyal. THE OPERATOR CON-

Kurtz, Barry and Karplus, Robert. INTELLECTUAL DEVELOP-

Mayer, Richard E. DIFFERENT RULE SYSTEMS FOR COUNTING

Schoenfeld, Alan H. EXPLICIT HEURISTIC TRAINING AS A

Silver, Edward A. STUDENT PERCEPTIONS OF RELATEDNESS

Thornton, Carol A. EMPHASIZING THINKING STRATEGIES IN

Vance, J. H. ATTITUDES TOWARD MATHEMATICS AND MATHE-

Webb, Norman L. PROCESSES, CONCEPTUAL KNOWLEDGE, AND

Mathematics Education Research Studies Reported in

Mathematics Education Research Studies Reported in

Abstract and comments prepared for I.M.E. by ARTHUR F. COXFORD,

3. Research Design and Procedures

3. Research Design and Procedures

Abstract and comments prepared for I.M.E. by JAMES M. MOSER,

3. Research Design and Procedures

Task Four: Adding'.

Most children categorized as transitional or itonconservers tended to rely

Abstract and comments prepared for I.M.E. by J. DAN KNIFONG,

3. Research Design and Procedures

Abstract and comments prepared for I.M.E. by LESLIE P. STEFFE,

3. Research Design and Procedure

Abstract and comments prepared for I.M.E. by LARRY SOWDER,

3. Research Design and Procedures

Abstract and comments prepared for I.M.E. by THOMAS R. POST,

3. Research Design and Procedures

Letter or After .48 .45 .43

Note. Base change position ends with w or 0. Experi-

Group Exp. 1 Exp. 2 Exp. 3

Note. Experiment 1 compares Letter and Number Groups; Experiments 2 and 3

The studies suggest that qualitatively different learning processes

Ehrenpreis, W. and Scandura, J. M. The algorithmic approach to curriculum

Green, J. G. Cognitive objectives of instruction: Theory of knowledge

Henderson, R. W.; Swanson, R.; and Zimmerman, B. J. Training seriation

Scandura, J. M. Role of rules in behavior: Toward an operational def-

Wertheimer, M. Productive thinking. New York: Harper & Row, 1959.

Abstract and comments prepared for I.M.E. by JANE SWAFFORD,

3. Research Design and Procedures

Abstract and comments prepared for I.M.E. by RICHARD E. MAYER,

3. Research Design and Procedures

Abstract and comments prepared for I.M.E. by GAIL SPITLER,

3. Research Design and Procedures (Study 1)

3. Research Design and Procedures (Study 2)

Abstract and comments prepared for I.M.E. by JOSEPH N. PAYNE,

3. Research Design and Procedures

Abstract and comments prepared for I.M.E. by THOMAS COONEY,

3. Research Design and Procedures

Abstract and comments prepared for I.M.E. by LEN PIKAART,