Mathematics

for the 21st Century:

What Should Students Learn?

Paper 2: Methods & Tools

July 2014
 


Prepared by:
Sanjoy Mahajan, Ph.D.

With sincere thanks and appreciation for the generous

support of the:

Center for Curriculum Redesign

Boston, Massachusetts
[Link]

July 2014
Copyright © 2014 Center for Curriculum Redesign. All Rights Reserved.
 About the Center For Curriculum Redesign

In   the   21st   century,   humanity   is   facing   severe   dif6iculties   at   the   societal,   economic,   and   personal   levels.  
Societally,  we  are  struggling  with  greed  manifested  in  6inancial  instability,  climate  change,  and  personal  
privacy  invasions,  and  with  intolerance  manifested  in  religious  fundamentalism,  racial  crises,  and  political  
absolutism.   Economically,   globalization   and   innovation   are   rapidly   changing   our   paradigms   of   business.  
On   a   personal   level   we   are   struggling   with   6inding   ful6illing   employment   opportunities   and   achieving  
happiness.   Technology’s   exponential   growth   is   rapidly   compounding   the   problems   via   automation   and  
offshoring,   which   are   producing   social   disruptions.   Educational   progress   is   falling   behind   the   curve   of  
technological  progress,  as  it  did  during  the  Industrial  Revolution,  resulting  in  social  pain.    

The   Center   for   Curriculum   Redesign   addresses   the   fundamental   question   of   "WHAT   should   students  
learn   for   the   21st   century?"   and   openly   propagates   its   recommendations   and   frameworks   on   a  
worldwide   basis.   The   CCR   brings   together   non-­‐governmental   organizations,   jurisdictions,   academic  
institutions,  corporations,  and  non-­‐pro6it  organizations  including  foundations.    

Knowledge, Skills, Character, and Metacognition

CCR   seeks   a   holistic   approach   to   deeply   redesigning   the   curriculum,   by   offering   a   complete   framework  
across  the  four  dimensions  of  an  education:  knowledge,  skills,  character,  and  metacognition.  Knowledge  
must  strike  a  better  balance  between  traditional  and  modern  subjects,  as  well  as  interdisciplinarity.  Skills  
relate   to   the   use   of   knowledge,   and   engage   in   a   feedback   loop   with   knowledge.   Character   qualities  
describe   how   one   engages   with,   and   behaves   in,   the   world.   Metacognition   fosters   the   process   of   self-­‐
re6lection  and  learning  how  to  learn,  as  well  as  the  building  of  the  other  three  dimensions.  

To learn more about the work and focus of the Center for Curriculum Redesign, please visit our website at
[Link]/about/background


ii
 Summary
 

Understanding  mathematics  is  essential  for  full  participation  in  society.  Yet  mathematics  is  learned  mostly  by  
rote,   with   little   understanding   or   possibility   for   transfer.   Thus,   PISA   can   provoke   a   revolution   in   mathematics  
education   worldwide—by   focusing   the   mathematics   assessment   on   distinguishing   rote   learning   from   true  
understanding.  Assessment  drives  teaching  and  learning:  What  PISA  measures,  the  mathematics  curriculum  
worldwide  will  deliver.  

Introduction
Nearly   30   years   ago,   one   of   the   authors   (SM)   last   helped   a   fellow   student   cheat.   The   class   was   civics.   Our  
teacher,  former  classmate  of  Apollo-­‐11  pilot  Buzz  Aldrin,  wondering  whether  we  could  find  Europe  on  a  map,  
sprang   upon   us   with   a   geography   test.   Its   hardest   question   asked   how   far   several   countries   were   from   the  
borders   of   civilization.   My   neighbor   in   the   first   row   of   desks   whispered:   "How   far   is   Cuba   from   the   United  
States?  A  thousand  miles?"  I  whispered  back,  "No,  ninety!"  The  next  day,  Buzz  Aldrin's  schoolmate  returned  
our  tests.  We  knew,  he  said,  more  geography  than  he  had  expected.  However,  one  student  placed  Cuba  90,000  
miles  from  the  United  States!  

A   comparison   with   other   earthly   distances—for   example,   the   3000   miles   across   the   entire   continental   United  
States—would  have  shown  my  class  neighbor  that,  when  forced  to  choose  between  90  and  90,000  miles,  the  
correct  distance  had  to  be  90  miles.  However,  this  powerful  reasoning  tool  of  comparison  had  not  been  part  
of   his   mathematical   education.   Handling   the   numbers   without   understanding   or   meaning,   he   picked   a  
distance  almost  halfway  to  the  moon.  

He  lacked  a  powerful  ability:  using  numbers  to  understand  and  make  sense  of  the  world.  If  in  your  mind  Cuba  
could   as   well   be   halfway   to   the   moon,   numbers   are   not   helping   you   understand   the   physical   layout   of   the  
world.  Columbus,  despite  all  his  geography  mistakes,  had  a  more  accurate  picture  of  the  world  500  years  ago.  

Mathematical   understanding   is   crucial   for   high   performance   in   our   personal,   public,   and   work   lives.   At   home,  
we  may  want  to  understand  the  results  of  a  medical  test,  or  rekindle  our  child's  love  of  math.  At  work,  we  may  
need  to  estimate  the  money,  time,  and  employees  for  a  large  project.  As  citizens,  we  may  want  to  judge  the  
rise   in   carbon-­‐dioxide   levels   in   the   air,   or   the   proportion   of   tax   dollars   that   should   go   to   health,   education,   or  
war.  Finally,  mathematics  underlies  our  science,  technology,  and  engineering.  

A   society   where   a   small   elite   has   the   legitimacy   to   govern   has   little   need   for   mathematics   as   a   social   language  
for   resolving   questions.   The   members   of   the   elite   share   a   worldview   and   trust.   In   democratic   societies,   the  
elite   is   diverse,   the   shared   trust   is   absent,   and   it   is   replaced   by   numbers—hence   the   prevalence   of   cost-­‐
1
benefit  analyses.  Indeed,  geometry  and  deductive  argument  arose  in  ancient  Athens  because  of  the  need  for  
2
them   in   the   democratic   Athenian   courts   and   juries.   Mathematics   became   the   only   universal   and   impartial  
language,  and  understanding  mathematics  becomes  essential  for  "full  participation  in  society".  

1
Porter, T. (1995). Trust in Numbers. Princeton University Press.
2
Lloyd, G. (1990). Demystifying Mentalities, Cambridge University Press; G.E.R. Lloyd (1992), "Democracy, philosophy, and science in ancient
Greece," in John Dunn (ed.), Democracy: The Unfinished Journey, 508 BC to 1993, Oxford University Press, pp. 41-56.
 

Mathematics  for  the  21st  Century  

  © CCR – [Link]   1  
PISA's goal is widely applicable real learning
3
The  breadth  of  mathematical  application  is  reflected  in  the  four  contexts  assessed  by  PISA:  

1. personal:  self,  family  and  peer  groups  

2. societal:  one's  community  
3. occupational:  the  world  of  work  
4. scientific:  application  to  science  and  related  issues  and  topics  

These   contexts   are   outstanding   choices.   Furthermore,   by   weighting   them   equally,   PISA   ameliorates   the  
misconception  that  mathematics  is  useful  only  in  the  scientific  context.  

Furthermore,   PISA   aims   to   assess   whether   students   can   “extrapolate   from   what   they   know,”   not   simply  
4
reproduce   knowledge.   Without   making   meaning   and   sense   of   mathematics,   students   cannot   extrapolate   or  
transfer   their   knowledge   to   new   domains.   Rote   knowledge—learning   without   understanding   or   meaning—is  
inert.   It   is   the   fundamental   obstacle   to   PISA's   goal,   that   mathematics   help   prepare   students   for   full  
participation  in  society.  

Origin of Rote Learning

It   is   understandable   that   the   traditional   mathematics   curriculum   would   produce   rote   learning,   because   our  
mathematics   curriculum   grew   up   in   an   era   with   different   needs.   At   first,   arithmetic   was   necessary   for  
monetary   calculations   needed   by   the   rising   merchant   class   and   Fibonacci's   Liber   Abaci   (1202)   was   the  
5
manual   of   instruction.   Even   through   much   of   the   twentieth   century,   if   an   exact   calculation   was   needed,   a  
human   had   to   do   it.   The   Enigma   code-­‐breaking   effort   and   the   Manhattan   project   even   relied   on   human  
computers.  Pressed  by  this  need,  we  developed  precise  algorithms  for  long  multiplication  and  division,  even  
for  extracting  square  roots,  and  in  mathematics  students  learned  these  procedures.  

However,  with  the  advent  of  electronic  calculators,  and  computational  technology  in  general,  the  human  and  
the   computer   could   specialize.   In   the   21st   century,   calculators   and   computers   can   make   the   tedious  
calculations.   Humans   will   keep   the   rest:  making   understanding   and   meaning.   Mathematics   teaching   can   and  
should  change  accordingly.  

The   contrast   between   human   and   computer   thinking   is   illustrated   in   an   old   experiment   by   the   Gestalt  
psychologist   Max   Wertheimer,   who   describes   observing   students   learning   the   fundamental   arithmetic  
6
operations.  The  students  were  solving  the  problems  set  by  the  teacher,  seemingly  understanding  what  they  
were  doing.  Curious  about  the  depth  of  understanding,  Wertheimer  gave  the  class  the  following  question:  

243 + 243 + 243 + 243

------------------------------- = ?
4

3
OECD. (2013). PISA 2012 Assessment and Analytical Framework: Mathematics, Reading, Science, Problem Solving and Financial Literacy. OECD
Publishing, p. 37.
4
Schleicher, A. (2012). "Use data to build better schools", TEDGlobal 2012.
5
Devlin, K. (2011). The Man of Numbers: Fibonacci's Arithmetic Revolution. Walker.
6
Wertheimer, M. (1945/1959). Productive Thinking (Michael Wertheimer, ed.), Harper.
 

Mathematics  for  the  21st  Century  

  © CCR – [Link]   2  
There   are   three   kinds   of   response:   laughter,   despair   ("that's   too   hard"),   and   computation.   The   best   answer   is  
laughter;  these  students  understood  the  connection  between  addition  and  multiplication.  The  worst  answer  
is   to   do   the   calculation.   This   answer   reveals   that   the   students,   even   if   they   could   do   arithmetic   manipulations  
with  great  skill,  did  not  have  any  clue  what  the  arithmetic  operations  meant.  

Imagine  trying  Wertheimer's  experiment  with  computers  as  the  students.  The  computers  cannot  laugh.  Nor  
can  the  computers  throw  up  their  hands  in  despair.  Rather,  the  computers  would  simply  compute  correctly.  
The  natural  computer  method  is  the  worst  human  method.  The  best  human  answer,  laughter,  is  beyond  the  
comprehension   of   any   computer.   Computer   and   human   reasoning   are   opposites.   The   kind   of   learning   that  
humans  and  computers  need—real  and  rote  learning,  respectively—are  opposites.  

Rote learning is the fundamental obstacle

In   all   four   PISA   contexts,   the   fundamental   barrier   to   high   performance   is   rote   or   parrot   learning.   Davydov  
distinguishes  rote  versus  real  learning  as  empirical  versus  theoretical  thinking,  with  empirical  thinking  (rote  
7
learning)  focusing  on  the  surface  features  of  a  problem.  As  a  result,  rote  learning  is  far  less  efficient  than  real  
learning.  

8
The  classic  comparison  is  Galperin  and  Pantina's  study  of  teaching  children  to  write  Russian  characters.  In  
comparison   to   children   who   simply   practiced   from   model   letters,   children   taught   a   theoretical  
understanding—the   underlying   structure   of   the   letter   strokes   and   how   to   describe   them—learned   with   far  
fewer   repetitions   and   greater   accuracy   and   transfer   to   new   domains   (such   as   copying   blueprints).   The   lesson  
for   mathematics   education   is   that   whenever   students   need   a   great   deal   of   drill,   it   is   likely   that   they   are  
learning  by  rote.  

The   canonical   example   of   rote   learning   comes   from   the   third   National   Assessment   of   Education   Progress  
(NAEP)   in   the   United   States,   students   performed   exact   long   multiplication   with   high   success   rates   (for   13-­‐  
9
and   17-­‐year   olds,   they   were   57-­‐   and   72-­‐percent,   respectively).   However,   when   students   were   asked   to  
"Estimate   3.04   *   5.3"   and   given   choices   of   1.6,   16,   160,   and   1600,   “the   13-­‐year   olds'   responses   were   almost   at  
the  level  of  guessing,  and  the  performance  by  17-­‐year-­‐olds  was  not  much  better.”  Students  do  not  understand  
what  multiplication  means,  even  though  they  can  perform  the  procedure.  

This   problem   has   been   studied   for   almost   a   century.   In   the   1920s,   Louis   Benezet,   who   had   just   become  
superintendent   of   schools   in   Manchester,   New   Hampshire,   noted   that   the   traditional   teaching   of   arithmetic  
10
was   “chloroform[ing]   the   child's   reasoning   faculties.”   As   an   example,   he   gave   students   in   a   traditionally  
taught  classroom  the  following  problem:  

7
Davydov , V. (1990). Types of Generalization in Instruction: Logical and Psychological Problems in the Structuring of School Curricula. Soviet Studies
in Mathematics Education. Volume 2. National Council of Teachers of Mathematics (original work published in 1972 in Russian). See also
Karpov, Y. (1995). "L. S. Vygotsky as the founder of a new approach to instruction.” School Psychology International 16(2):131-142.
8
Galperin, P. & Pantina, N. (1965). "The Dependency Of The Motor Habit On The Type Of The Task Orientation," Russian Monographs on Brain
and Behavior 3:425-433 (original work published in 1957 in Russian).
9
Carpenter, T., et al. (1983). "Results of the third NAEP mathematics assessment: Secondary school," The Mathematics Teacher 76(9), 652-659.
10
Benezet, L. (1935, 1936). “The teaching of arithmetic I, II, III: The story of an experiment," Journal of the National Education Association 24(8):241-
244 (I, Nov 1935), 24(9):301-303 (II, Dec 1935), 25(1):7-8 (III, Jan 1936), available at the Benezet Centre. The "chloroform" quote is from part I. For
a study of Benezet's curriculum, showing very positive results, see Etta Berman (1935), The Result of Deferring Systematic Teaching of
Arithmetic to Grade Six as Disclosed by the Deferred Formal Arithmetic Plan at Manchester, New Hampshire, Masters Thesis, Boston
University, USA.
 

Mathematics  for  the  21st  Century  

  © CCR – [Link]   3  
Three   steamers   leave   Boston   simultaneously   for   Portland   [Maine].   One   takes   12   hours;   the   other   takes   15  
hours;  and  the  third  takes  17  hours.  How  long  before  all  the  steamers  are  in  Portland?  

In   the   ninth   grade   (roughly   age   14),   only   6   out   of   29   students   solved   the   problem   correctly.   (As   a   result   of  
such   performances,   Benezet   instituted   a   mathematics   curriculum   in   grades   1-­‐7   that   emphasized  
understanding   and   using   mathematics,   with   the   procedural   aspects   mastered   in   a   few   months   in   the   7th  
grade.)  

The  problem  of  lacking  meaning  is  worldwide  and  is  separate  from  computational  fluency.  An  example  comes  
11
from  the  famous  NAEP  school-­‐bus  problem:  

1128  soldiers  have  to  be  bused  to  a  base;  each  bus  holds  36  soldiers.  How  many  bus  are  needed?  

Even  though  the  correct  answer  is  32,  and  even  though  most  students  can  set  up  the  division  correctly,  the  
most   common   answer   by  US  students  (on  the  NAEP)  was  the  meaningless  "31  R  12"  (31  with  a  remainder  of  
12).  Unfortunately,  the  failure  to  make  meaning  of  mathematics  affects  students  in  China  as  well,  despite  the  
12
greater  computational  fluency  of  Chinese  students.  

Verschaffel   and   colleagues,   studying   students   in   Belgium,   Japan,   and   beyond,   found   that   students   are  
13
reluctant  to  use  real-­‐world  information  to  solve  or  even  make  sense  of  mathematics  problems.  For  example,  
when   asked   for   the   water   temperature   after   combining   40-­‐   and   120-­‐degree   water,   students   answer   with   160  
degrees—an   answer   divorced   from   any   meaning.   In   contrast,   if   you   ask   students   to   describe   the   result   of  
mixing  hot  and  cold  water—that  is,  if  the  question  doesn't  directly  involve  mathematics—they  correctly  say  
warm   or   lukewarm.   When   the   problem   becomes   mathematical,   thinking   too   often   gives   way   to   rote  
procedures.  

The good news is that testing drives education

Rote   learning   is   so   fundamental   and   widespread   that   one   might   despair.   Fortunately,   assessment   drives  
14
learning  and  teaching.  That  observation,  described  extensively  in  medical  education,  provides  the  lever  to  
guide  mathematics  teaching  toward  21st  Century  Skills.  The  "students"  of  PISA  are  the  OECD  and  associated  
countries.   When   PISA   sets   a   high   standard,   countries   follow   in   their   teaching   practices.   The   four   PISA  
contexts,  developed  in  the  2012  assessment,  have  solved  the  problem  of  “What  should  mathematics  be  for?”  
If,  in  each  area,  the  measure  is  understanding,  countries  will  shift  their  education  away  from  rote  learning  and  
toward  real  understanding.  

11
Carpenter (1983), supra.
12
Cai, J. & Silver, E. (1995). "Solution processes and interpretations of solutions in solving a division-with-remainder story problem: Do Chinese
and US students have similar difficulties?" Journal for Research in Mathematics Education 26(5), 491-497.
13
Verschaffel, L., Greer, B., & de Corte, E. (2000). Making Sense of Word Problems, Swets and Zeitlinger; Hajime Yoshida, Lieven Verschaffel, and
Erik de Corte (1997), "Realistic considerations in solving problematic word problems: Do Japanese and Belgian children have the same
difficulties?" Learning and Instruction 7(4), 329-338.
14
Newble, D. & Jaeger, K. (1983). "The effect of assessments and examinations on the learning of medical students," Medical Education 17(3),
165-71. See also Marjolein Heijne-Penninga (2008), "Influence of open- and closed-book tests on medical students' learning approaches,"
Medical Education 42(10), 967-974.
 

Mathematics  for  the  21st  Century  

  © CCR – [Link]   4  
Test for understanding
Thus,  the  marking  of  questions  needs  to  have  the  correct  sign,  in  that  answers,  or  justifications,  that  reveal  
15
lack  of  understanding  “contaminate  the  whole  answer.”  Vinner  has  called  this  kind  of  rote  thinking  “pseudo-­‐
16
conceptual”  or  “pseudo-­‐analytic”  because  it  contains  only  a  facsimile  of  analysis  or  conceptual  knowledge.  
Being   able   to   detect   whether   analysis   is   real   or   fake   requires   the   right   kind   of   question,   one   where   rote  
learning   and   understanding   lead   to   different   conclusions,   or   where   rote   learning   fails   entirely.   As   Vinner  
illustrates,  if  the  question  is,  “A  rectangle  has  side  lengths  7  and  5  cm.  What  is  its  area?”  then  understanding  
or   rote   learning   will   lead   to   the   answer   of  35   cm^2.   However,   if   the   question   is,   “The   perimeter   of   a   rectangle  
is   24   cm,   and   one   side   has   length   7   cm.   What   is   the   rectangle's   area?”   then   the   rote   approach   will   often   give   7  
*  24  =  168,  and  the  student  will  write  down  168  cm^2.  

17
An  example  from  PISA  is  the  "Climbing  Mount  Fuji"  unit.  The  question  asked  students  to  find  the  step  length  
of  a  person  who  walked  22,500  steps  up  Mount  Fuji  covering  9  km.  A  small  proportion  of  students  got  partial  
credit   for   giving   0.4   cm   or   4000   cm   as   answers.   The   0.4   cm   can   be   understood   as   meaning   meters   and  
forgetting   to   include   the   units.   But   4000   cm   has   to   be   nonsense,   whether   as   meters   of   centimeters.   That   it  
could  be  written  down  shows  that  the  student  has  no  conception  of  mathematics  as  describing  the  world.  

18
Similarly,   the   "Pizzas"   unit   allows   students   to   calculate   the   answer   without   understanding   proportional  
reasoning.  They  can  mindlessly  compute  the  area  of  the  pizzas,  divide  the  areas  by  the  respective  cost,  and  
compare   the   results—a   very   low-­‐level   skill.   In   contrast,   if   the   pizzas   share   the   same   irregular   shape   such   as   a  
kidney  bean  (at  different  sizes),  then  proportional  reasoning  is  the  only  way  to  solve  the  problem.  

With  a  repertoire  of  questions  that  distinguish  rote  learning  from  understanding,  the  marking  scales  in  PISA  
could   reliably   emphasize   measuring   what   students   understand   about   mathematics   as   a   tool   for   making   sense  
of  the  world.  

 
Mathematics for transfer in the 21st century
The  mathematics  that  students  need  for  full  participation  in  society,  in  all  four  contexts,  can  be  divided  into  
reasoning   tools.   The   tools   fall   into   two   groups:   tools   that   build   connection   among   quantities   (and,   more  
generally,  mathematical  objects)  and  tools  that  simplify  the  world.  

(A) Tools  for  connection  

(1) comparison  
(2) proportional  reasoning  
(3) multiplicative  scales  (including  exponential  growth)  
(4) probabilistic  (including  logical)  reasoning  

15
Rogers, E. (1969). "Examinations: Powerful agents for good or ill in teaching," American Journal of Physics, 37, 954-962.
16
Vinner, S. (1997). "The pseudo-conceptual and the pseudo-analytical thought processes in mathematics learning," Educational Studies in
Mathematics 34:97–129. Similarly, Skemp distinguishes between instrumental and relational understanding [Richard Skemp (1976),
"Relational Understanding and Instrumental Understanding," Mathematics Teaching 77:20-26]. In contrast to instrumental understanding (the
analog of rote learning), relational understanding promotes itself, and leads to more relational understanding, as the network of
understanding grows and thickens.
17
PISA 2012 Assessment and Analytical Framework, p. 48.
18
PISA 2012 Assessment and Analytical Framework, pp. 49-51.
 

Mathematics  for  the  21st  Century  

  © CCR – [Link]   5  
 (B) Tools  for  simplification  
(5) divide  and  conquer  
(6) lumping  
(7) easy  cases  

Tools for connection

Man,   said   the   Greeks,   is   a   social   animal:   Without   any   conection   to   others,   we   are   not   fully   human,   and   life  
loses   meaning.   Quantities   are   similar.   Alone,   a   quantity   has   no   meaning;   it   is   just   an   isolated   fact.   Meaning  
comes   from   connections   to   other   quantities.   In   contrast   to   computers,   we   humans   are   meaning-­‐making  
creatures.  We  need  reasoning  tools  to  build  connection.  Thus,  this  first  set  of  tools—tools  for  connection—is  a  
first  major  step  toward  using  mathematics  to  represent  the  complexity  of  the  world.  

1) Comparison.   No   quantity   is   meaningful   by   itself,   and   thus   must   be   compared   to   another,   related  
quantity  to  give  it  meaning.  For  example,  a  monetary  rate  of  $1  trillion  dollars  per  year  means  little  
by  itself.  As  a  fraction  of  a  government  budget  or  a  GDP,  it  acquires  meaning  for  us.  The  essential  skill  
is  making  meaningful  dimensionless  ratios.  

19
Comparison   a   fundamental   part   of   critical   thinking.   For   example,   reading   a   bare   number   in   a  
newspaper   article,   such   as   "1   million   fans   came   into   the   streets   to   celebrate   their   sports   team"   (as  
appears   regularly   in   Boston   newspapers)   conveys   little   information.   Only   when   the   1   million   is  
compared  to  the  total  population  of  Boston  and  the  inner  suburbs,  also  approximately  1  million,  does  
the  number  acquire  meaning  (as  being  highly  implausible).  

2) Proportional   Reasoning.   Comparison   naturally   produces   ratios.   Understanding   how   ratio   changes   are  
related  (e.g.  doubling  lengths  quadruples  areas)-­‐-­‐-­‐the  essence  of  proportional  reasoning—is  essential  
for  engineering  and  science.  For  example,  the  quadratic  proportionality  between  diffusion  times  and  
distance  means  that  large  animals  must  have  a  circulatory  system.  

Proportional  reasoning,  explained  the  renowned  mathematics  educator  Hans  Freudenthal,  deserves  
“priority  above  algorithmic  computations  and  applications  of  formulae  because  it  deepens  the  insight  
20
and   the   rich   context   in   the   naive,   scientific,   and   social   reality   where   it   operates.”   Without  
proportional   reasoning,   the   world   makes   no   sense.   With   it,   we   find   the   world   rich   in   quantities  
connected  through  ratio.  Seeing  the  world  through  ratio  is  hardwired  into  our  brains'  approximate  
21
number  system.  Our  teaching  should  build  on  this  inherent  and  powerful  ability  to  make  sense  of  
the  world.  

3) Multiplicative   scales   (including   exponential   growth).   Once   ratio   is   seen   as   essential,   the   next   step   is  
counting   ratios—e.g.   by   how   many   factors   of   2   (or   10)   is   the   sun   larger   in   mass   than   the   earth?  
Counting  ratios  leads  to  logarithmic  and  exponential  scales.  They  are  equally  essential  for  the  social  
and  physical  worlds.  Logarithmic  scales  organize  the  huge  dynamic  range  of  lengths,  times,  energies  
in   the   universe.   In   the   social   world,   exponential-­‐growth   models   help   understand   the   growth   of  
populations,  economies,  or  resource  use.  Without  understanding  and  taming  exponential  growth,  the  

19
Critical thinking is one of the 4 C's in Trilling, B. & Fadel, C. (2009). 21st Century Skills: Learning for Life in Our Times, Jossey-Bass.
20
Freudenthal, H. (1983). Didactical Phenomenology of Mathematical Structures, Reidel, p. 401.
21
Dehaene, S. (2011). The Number Sense: How the Mind Creates Mathematics, Oxford University Press.
 

Mathematics  for  the  21st  Century  

  © CCR – [Link]   6  
 22
world  has  no  chance  of  using  its  resources  wisely.  Proportional  reasoning  and  exponential  growth  
need   to   be   understood   at   a   gut   level,   simply   as   part   of   how   students   perceive   the   world  
23
quantitatively.  Both  ideas  take  a  short  time  to  describe,  but  a  long  time  to  master.  

(This   mastery   would   come   much   quicker   if   our   teaching   of   ratio,   proportionality,   and   logarithmic  
scales  were  aligned  with  our  innate  perception  of  quantity.    Instead,  students  are  untaught  our  innate  
multiplicative  scale,  and  then  retaught  it  much  later  as  a  series  of  seemingly  arbitrary  and  therefore  
meaningless  rules  for  manipulating  logarithms.    Teachers  of  physical  science  and  engineering  can  list  
numerous  examples  of  students'  difficulties  with  logarithms.)  

Within  the  scope  of  this  tool  falls  the  Rule  of  72  for  estimating  the  doubling  time  of  an  exponential  
24
process  given  the  growth  rate.  As  an  illustrative  assessment,  though  one  unlikely  to  be  adopted  by  
many  governments,  students  (without  a  calculator)  could  be  asked,  "Your  national  government  plans  
to  underestimate,  by  2  percent  per  year,  the  rate  of  inflation.  If  you  will  retire  in  40  years,  by  roughly  
what  factor  will  your  public  pension  be  devalued?"  

4) Probabilistic   (including   logical)   Reasoning.   Our   knowledge   of   the   world   is   necessarily   incomplete.   We  
represent   our   incomplete   knowledge   through   probability-­‐-­‐-­‐in   particular,   the   probabilities   of  
hypotheses.   Collecting   data   and   evidence   changes   our   knowledge   and   thus   our   probabilities  
25
according  to  Bayes  theorem.  Bayes  theorem  connects  evidence  and  belief.  It  is  a  tool  for  connection.  

If  students  understand  Bayes  theorem,  they  do  not  fall  into  the  prosecutor's  fallacy  or  the  defense-­‐
26
attorney's   fallacy.   They   can   judge   the   strength   of   clinical   evidence,   and   acquire   a   way   of   thinking  
that  matches  strength  of  conviction  to  evidence.  This  tool  is  essential  for  critical  thinking.  

Probabilistic  reasoning  contains  logical  reasoning:  In  the  extreme  case  where  probabilities  are  either  
0  (false)  or  1  (true),  probabilistic  reasoning  simplifies  to  propositional  logic  (logical  reasoning).  The  
best  example  of  logical  reasoning  taught  for  transfer  is  Harold  Fawcett's  geometry  course  from  the  
27
1930s.   Instead   of   of   reproducing   geometric   constructions,   which   are   anyway   performed   more  
precisely  by  computers,  students  learned  proof  in  order  to  deconstruct  advertising  copy  or  improve  
the   framing   of   laws-­‐-­‐-­‐true   critical   thinking.   An   illustrative   problem   from   the   course   was   the  
following:  

22
Bartlett, A. (2004). The Essential Exponential! For the Future of Our Planet, Center for Science, Mathematics and Computer Education, University of
Nebraska, Lincoln. Al Bartlett's lecture on the importance of understanding exponential growth has over 5 million views on YouTube. The
subsequent seven parts each have more than 500,000 views.
23
On the difficulties student face in mastering proportional reasoning, see for example Dirk de Bock et al. (2002), “Improper use of linear
reasoning: An in-depth study of the nature and the irresistibility of secondary school students' errors,” Educational Studies in Mathematics
50(3): 311-334. On the long time to master multiplicative structures, see OECD (2009), Learning Math for Life: A Perspective from PISA, p. 177
(p.179 of the PDF file).
24
Alberta Education. (2008). The Alberta 10--12 mathematics program of studies with achievement indicators. Alberta, Canada, p. 92.
25
Jaynes, E. (2003). Probability Theory: The Logic of Science. Cambridge University Press.
26
Thompson, W. & Schumann, E. (1987). "Interpretation of Statistical Evidence in Criminal Trials: The Prosecutor's Fallacy and the Defense
Attorney's Fallacy," Law and Human Behavior 2(3):167-187. See also Leila Schneps and Coralie Colmez (2013), Math on Trial: How Numbers
Get Used and Abused in the Courtroom, Basic Books.
27
Fawcett, H. (1938). The Nature of Proof, Bureau of Publications, Columbia University. Reprinted by the NCTM in 1995 and 2001. See also Erica
Lane (2004), “The Nature of Proof in Today's Classroom,” The Montana Mathematics Enthusiast (TMME) 1(2):58-65; and Fred Flener (2001, April
6), “A geometry course that changed their lives: The guinea pigs after 60 years,” Annual Conference of the NCTM, Orlando, Florida.
 

Mathematics  for  the  21st  Century  

  © CCR – [Link]   7  
 While   reading   a   magazine   Helen's   attention   was   drawn   to   the   picture   of   a   beautiful   girl   with   an  
attractive   smile   who   was   represented   as   saying:   “I'd   wished   a   thousand   times   for   a   brighter   smile.  
One   tube   of   Colgate's   gave   it   to   me.   It   was   so   annoying   to   see   other   girls   with   lovely   smiles   get   all   the  
dates.  Then  I  tried  Colgate's.  Now  my  smiles  are  bright  too.”  

This   is   really   an   argument   for   using   Colgate's   Dental   Cream.   The   argument   is   based   on   certain  
assumptions.  What  are  these  assumptions?  

It  is  hard  to  imagine  a  more  powerful  or  transferable  ability.  

Tools for simplification

The  world  is  filled  with  quantities  and  other  mathematical  objects.  Thus,  giving  them  meaning,  which  is  the  
purpose   of   the   first   group   of   tools,   can   be   dangerous.   The   meanings   will   overwhelm   us.   Thus,   we   cannot   shut  
our   toolbox   yet.   The   second   group   of   tools—tools   for   simplification—enable   us   to   fit   the   complexity   of   the  
world  into  our  limited  minds.  

1) Divide   and   Conquer.   We   often   split   hard   problems   into   manageable   pieces.   However,   this   essential  
problem-­‐solving   skill   needs   naming,   so   that   students   realize   when   they   are   using   it,   and   have   a  
method  to  begin  even  difficult  problems.  This  tool  is  valuable  from  the  earliest  ages.  For  example,  in  
adding  24  +  38,  students  can  divide  38  into  6  and  32,  and  then  add  24+6=30  and  30+32=62.  

For   PISA-­‐age   students,   a   class   of   problems   that   use   divide-­‐and-­‐conquer   reasoning   is:   "At   what   rate  
(passengers   per   hour)   can   a   busy   train   track   carry   passengers?"   Students   must   divide   the   problem  
into   simpler   estimates:   how   many   trains   per   hour,   how   many   cars   on   each   train,   and   how   many  
passengers  in  each  train.  (This  problem  also  tests  their  ability  to  be  comfortable  with  imprecision,  an  
ability  developed  by  the  next  reasoning  tool  of  lumping.)  

This  tool  is  essential  to  solving  interview  questions  for  management-­‐consulting  positions  at  leading  
firms  such  as  McKinsey.  The  task  of  a  consultant  is  to  understand  quickly  the  essential  attributions  of  
a  problem.  Skills  essential  for  consulting  are  essential  in  the  occupational  context.  

2) Lumping.  Rigor  leads  to  rigor  mortis.  When  we  round  numbers  or  approximate  complicated  shapes  
with  simple  shapes,  we  throw  away  less  important  details,  in  order  to  free  our  minds  to  master  the  
essential  parts  of  a  problem.  Only  in  this  way  can  be  master  complexity.  A  student  who  has  mastered  
lumping  could  never  answer  that  3.04  *  5.3  is  approximately  1600.  

3) Easy   Cases.   When   a   problem   is   still   too   hard,   we   look   at   its   special   cases,   ones   that   we   can  
understand.  For  example,  in  evaluating  a  mathematical  conjecture,  we  first  test  it  with  n=0  and  n=1.  
Or,  in  the  personal  and  societal  contexts,  mortgages  can,  at  one  extreme,  be  like  annuities:  
interest  rate  *  loan  period  >>  1  
At  the  other  extreme,  they  can  like  installment  loans:  
interest  rate  *  loan  period  <<  1  
In   each   extreme   the   payment   is   much   easier   to   calculate,   and   the   calculation   is   much   easier   to  
understand,  than  in  the  general  case.  Thinking  about  easy  cases  promotes  insight.  

Mathematics  for  the  21st  Century  

  © CCR – [Link]   8  
 This   kind   of   reasoning   is   particularly   valuable   in   the   scientific   context.   For   example,   in   Liping   Ma's  
28
study  of  American  and  Chinese  mathematics  teachers,  teachers  were  asked  to  propose  answers  to  a  
student   who   says,   "I   discovered   that   the   bigger   the   perimeter   of   a   rectangle,   the   bigger   the   area."  
Many   teachers   did   not   know   how   to   decide   whether   the   claim   was   true.   The   method   of   easy   cases  
provides  an  insightful  route:  Make  the  rectangle  thinner  but  longer.  In  that  way,  the  perimeter  can  be  
made  as  large  and  the  area  can  be  made  as  close  to  zero  as  one  wants.  Thus,  the  conjecture  must  be  
false.   If   teachers   and,   more   importantly,   their   students   can   apply   this   reasoning,   they   gain   a   valuable  
tool  for  testing  and  developing  new  conjectures.  

Conclusion:
Reasoning tools fight rote learning and promote transfer
Reasoning   tools,   by   their   nature,   transfer   across   domains.   They   are   therefore   a   powerful   antidote   to   rote  
learning.   By   emphasizing   real   understanding,   PISA's   assessment   will   push   the   mathematics   curriculum  
worldwide   toward   competence   in   the   four   PISA   contexts:   personal,   societal,   occupational,   and   scientific.   By  
offering  reasoning  tools  as  the  organizing  curriculum  principle,  PISA  can  help  countries  reach  this  promised  
land.  

28
Liping M. (1999). Knowing and Teaching Elementary Mathematics: Teachers' Understanding of Fundamental Mathematics in China and the United
States. Lawrence Erlbaum Associates.
 

Mathematics  for  the  21st  Century  

  © CCR – [Link]   9  
Center for Curriculum Redesign
Boston, Massachusetts
[Link]

Sanjoy Mahajan, Ph.D. With the generous support of:

Copyright © 2014 Center for Curriculum Redesign. All Rights Reserved.