Two Frameworksfor Discrete Mathematics
Kirby McMaster
Brian Rague
Steven Hadfield
Two Frameworks for Discrete Mathematics 2
Frameworks and Gestalts
We use the term gestalt to refer to a mental framework that allows
students to organize course topics into a unified whole (greater than
the sum of the parts).
Davis and Hersh: People vary dramatically in what might be called their cognitive style, that is, their primary mode of thinking.
Bain: Students bring paradigms to the class that shape how they construct meaning.... Even if they know nothing about our subjects, they still use an existing mental model of something to build their knowledge of what we tell them.
2
Two Frameworks for Discrete Mathematics 3
Gestalts for Mathematics
Which gestalts are emphasized in Mathematics?
Davis and Hersh: The definition-theorem-proof approach to mathematics has become almost the sole paradigm of mathematical exposition and advanced instruction.
Courant and Robbins: There seems to be a great danger in the prevailing over emphasis on the deductive-postulational character of mathematics. The element of constructive invention, of directing and motivating intuition, remains the core of any mathematical achievement.
Polya: Mathematics has two faces. Mathematics presented in the Euclidean way appears as a systematic, deductive science; but mathematics in the making appears as an experimental, inductive science.
3
Two Frameworks for Discrete Mathematics 4
Measuring Mathematical Gestalt
In previous research, we were able to characterize and measure
two frameworks for mathematics: Logical Math gestalt – based on proving theorems. Computational Math gestalt – based on solving problems.
Our methodology assumed that: The words people use are suggestive of their mental state. In particular, the words used frequently in a book indicate the
gestalt of the author.
4
Two Frameworks for Discrete Mathematics 5
Purpose of this Study
In this study, we apply our Logical Math (LMATH) and Computational
Math (CMATH) scales to a sample of 25 Discrete Math books.
Our primary purpose is to assess the level of emphasis these books give to each gestalt.
Our findings have relevance in the development of approaches for teaching mathematical topics in computing courses, especially Discrete Math courses.
5
Two Frameworks for Discrete Mathematics 6
Logical Math Gestalt
The LMATH scale consists of 10 word/groups and weights.
As expected, theorem, proof, and definition are on the scale.
Scale words used to convey logical order in proofs include let, since, follow, hence, and therefore.
Two general math concepts on the scale are set and function.
The LMATH weights for each word/group can be used to calculate
an overall weighted-average score for a specific book.
6
Two Frameworks for Discrete Mathematics 7
Word/Group Books AvgStdFreq
Weight
theorem/lemma/corollary 25 428.2 18.34
let 25 418.1 17.78
set/subset 25 333.1 13.03
proof/prove 25 329.9 12.85
function/map 25 300.9 11.23
hence/thus/therefore 25 230.3 7.28
definition/define 25 221.4 6.78
show/shown 25 191.3 5.10
follow/following 24 175.4 4.21
since 24 160.8 3.40
TOTAL 100.00
7
LMATH Scale
Two Frameworks for Discrete Mathematics 8
Word/Group Weight StdFreqLMATHScale
theorem/lemma/corollary 18.34 214.8 39.4
let 17.78 *413.1 73.4
set/subset 13.03 *962.3 125.4
proof/prove 12.85 188.7 24.2
function/map 11.23 *527.0 59.2
hence/thus/therefore 7.28 226.8 16.5
definition/define 6.78 230.0 15.6
show/shown 5.10 237.4 12.1
follow/following 4.21 127.7 5.4
since 3.40 187.3 6.4
TOTAL 377.6
8
LMATH CalculationsRoyden – Real Analysis
Two Frameworks for Discrete Mathematics 9
Computational Math Gestalt
The CMATH scale consists of 10 word/groups and weights.
As expected, problem and solution are on the scale.
Scale words model and algorithm describe how problems are to be solved.
The words variable, equation, and constraint represent components of models.
The word function is on both scales.
The CMATH weights for each word/group can be used to calculate
an overall weighted-average score for a specific book.
9
Two Frameworks for Discrete Mathematics 10
Word/Group Books AvgStdFreq
Weight
problem 25 394.3 18.56
method/algorithm 23 343.1 15.33
solution/solve 25 322.4 14.03
value/variable 24 271.7 10.83
equation/inequality 21 265.9 10.46
function/map 24 257.0 9.90
model/modeling 19 222.3 7.71
point/line 24 175.8 4.78
system/subsystem 23 168.7 4.33
condition/constraint 25 164.3 4.06
TOTAL 100.00
10
CMATH Scale
Two Frameworks for Discrete Mathematics 11
Word/Group Weight StdFreqCMATHScale
problem 18.56 *483.8 89.8
method/algorithm 15.33 262.7 40.3
solution/solve 14.03 *496.9 69.7
value/variable 10.83 *474.4 51.4
equation/inequality 10.46 69.3 7.2
function/map 9.90 114.4 11.3
model/modeling 7.71 268.3 20.7
point/line 4.78 -- --
system/subsystem 4.33 137.4 5.9
condition/constraint 4.06 180.7 7.3
TOTAL 303.7
11
CMATH CalculationsHillier & Lieberman – Operations Research
Two Frameworks for Discrete Mathematics 12
Gestalts in Discrete Math Books
Our analysis of mathematical gestalt data from 25 Discrete Math
books seeks to answer the following questions:
1.How often do LMATH and CMATH word/groups appear in the sample books?
2.What is the distribution of LMATH and CMATH scores?
3.What is the contribution of individual word/groups to the overall LMATH and CMATH scores for selected high-scoring books?
4.What is the relationship between LMATH and CMATH scores?
12
Two Frameworks for Discrete Mathematics 13
LMATH and CMATH Word/Groups
The following tables summarize how often LMATH and CMATH
word/groups appear in the sample Discrete Math books. Nine of the 10 LMATH word/groups appear in the
concordance of more than 20 books.
Two of the 10 CMATH word/groups appear in the concordance of more than 20 books (including function/map, which is on both scales).
Only one Discrete Math book listed model as a frequent word.
Discrete Math books contain substantially more Logical Math words than Computational Math words.
13
Two Frameworks for Discrete Mathematics 14
Word/Group Books Avg
StdFreqset/subset 25 427.0
proof/prove 22 217.0
definition/define 17 210.6
hence/thus/therefore 21 199.2
function/map 22 195.0
follow/following 24 194.9
let 22 191.1
theorem/lemma/corollary 23 180.3
since 22 144.6
show/shown 24 132.4
14
LMATH Word/Groups25 Discrete Math Books
Two Frameworks for Discrete Mathematics 15
Word/Group Books Avg
StdFreqfunction/map 22 195.0value/variable 22 155.1
point/line 14 145.4
method/algorithm 17 145.3
solution/solve 14 139.2
problem 18 127.4
system/subsystem 10 87.2
equation/inequality 9 80.2
condition/constraint 3 60.5
model/modeling 1 52.8
15
CMATH Word/Groups25 Discrete Math Books
Two Frameworks for Discrete Mathematics 16
LMATH and CMATH Distributions
LMATH and CMATH scores vary widely across the 25 Discrete
Math books in the sample.
LMATH scores range from 94.3 to 348.9.The mean is 200.3.
CMATH scores range from 36.8 to 175.2.The mean is 83.8.
A graph of the LMATH and CMATH distributions is shown on the
following slide.
16
Two Frameworks for Discrete Mathematics 17
LMATH and CMATH Distributions25 Discrete Math Books
17
0
2
4
6
8
10
12
14
0-49 50-99 100-
149
150-
199
200-
249
250-
299
300-
349
350+
Gestalt Score
Bo
oks
LMATH CMATH
Two Frameworks for Discrete Mathematics 18
LMATH and CMATH Calculations
Eleven of the Discrete Math books have LMATH scores above 200.
The two books with highest LMATH scores are: Nievergelt (2002), Foundations of Logic and Mathematics
(LMATH=348.9). Gries (1993), A Logical Approach to Discrete Math (LMATH=316.5).
No Discrete Math book has a CMATH score above 200. The book with highest CMATH score is: Rosen (1999), Handbook of Discrete and Combinatorial Math
(CMATH=175.2).
The LMATH calculations for Nievergelt and Gries, along with the CMATH calculations for Rosen, are shown in the following tables.
18
Two Frameworks for Discrete Mathematics 19
Word/Group Weight StdFREQLMATHScale
theorem/lemma/corollary 18.34 *689.8 126.5
let 17.78 -- --
set/subset 13.03 *586.9 76.5
proof/prove 12.85 *536.1 68.9
function/map 11.23 146.8 16.5
hence/thus/therefore 7.28 273.7 19.9
definition/define 6.78 *353.2 23.9
show/shown 5.10 96.4 4.9
follow/following 4.21 279.3 11.8
since 3.40 -- --
TOTAL 348.9
19
LMATH CalculationsNievergelt – Foundations of Logic and Mathematics
Two Frameworks for Discrete Mathematics 20
Word/Group Weight StdFREQLMATHScale
theorem/lemma/corollary 18.34 *422.8 77.5
let 17.78 103.0 18.3
set/subset 13.03 *411.0 53.6
proof/prove 12.85 *716.3 92.0
function/map 11.23 192.9 21.7
hence/thus/therefore 7.28 164.8 12.0
definition/define 6.78 *342.7 23.2
show/shown 5.10 101.1 5.2
follow/following 4.21 220.0 9.3
since 3.40 108.6 3.7
TOTAL 316.5
20
LMATH CalculationsGries – A Logical Approach to Discrete Math
Two Frameworks for Discrete Mathematics 21
Word/Group Weight StdFREQCMATHScale
problem 18.56 206.5 38.3
method/algorithm 15.33 *341.3 52.3
solution/solve 14.03 105.4 14.8
value/variable 10.83 182.1 19.7
equation/inequality 10.46 55.3 5.8
function/map 9.90 241.3 23.9
model/modeling 7.71 52.8 4.1
point/line 4.78 249.6 11.9
system/subsystem 4.33 101.3 4.4
condition/constraint 4.06 -- --
TOTAL 175.2
21
CMATH CalculationsRosen – Handbook of Discrete and Combinatorial Math
Two Frameworks for Discrete Mathematics 22
LMATH vs. CMATH Relationship
The relationship between LMATH and CMATH scores for the
sample books is displayed on the next slide as a scatter plot.
The relationship between the two gestalt scales is slightly negative (correlation coefficient = -0.343).
Books with high LMATH scores have relatively low CMATH scores.(The converse is not true.)
For 24 of the 25 Discrete Math books, the LMATH score is higher than the CMATH score.
22
Two Frameworks for Discrete Mathematics 23
LMATH vs. CMATH Scatter Plot25 Discrete Math Books
23
0
50
100
150
200
250
0 50 100 150 200 250 300 350
LMATH Score
CM
AT
H S
core
Two Frameworks for Discrete Mathematics 24
Which Framework for Discrete Math?
Our intent in this study was not to determine which framework is
"best“ for Discrete Math, since each gestalt provides value for the course.
Both Logical Math and Computational Math encourage abstraction: By manipulating symbolic objects in proofs, and Through constructing models to solve problems.
The Discrete Math instructor should provide an appropriate blend of problem-solving and theorem-proving in the course.
The difficult task will be finding a textbook that supports both gestalts, since most Discrete Math books favor Logical Math.
Programming assignments are one way to support Computational Math in a Discrete Math course.
24