2/27/2017 discretemathematicsProve$\sum\binomnk2^k=3^n$usingthebinomialtheoremMathematicsStackExchange

signup login tour help

_
MathematicsStackExchangeisa Here'showitworks:
questionandanswersiteforpeople
studyingmathatanyleveland
professionalsinrelatedfields.Join
themitonlytakesaminute:
Anybodycanask Anybodycan Thebestanswersarevoted
Signup aquestion answer upandrisetothetop

Prove ( usingthebinomialtheorem
n k n
)2 = 3
k

I'mstudyingforamidtermandneedsomehelpwithprovingsummationusingthe
binomialtheorem.
n
n k n
( )2 = 3
k
k=0

ThisiswhatI'mthinkingsofar:Inthebinomialtheoremwesetx=0,andy=2,so:
n
n n n k
3 = (x + y) = ( )y
k
k=0

n
n k
= ( )2
k
k=0

AmIgettingthiscorrectsofarorcompletelywrong?Anyhelpwouldbeappreciated.
Thanks.

(discretemathematics ) (summation) (binomialcoefficients ) (binomialtheorem)

editedDec15'14at14:40 askedOct14'13at0:46
MartinSleziak StackPWRequirmentsAr
39.1k 5 93 202 eCrazy
342 1 3 12

n
Use(1 + 2) .FelixMarinOct14'13at1:26

2Answers

Notquiteifx = 0 andy = 3thenyouwouldhave3 n

= y
n
whichisnotwhatyouwant.If
yousetx = 0 andy = 2youwouldhave2 = y . n n

Comparethiswiththegeneralbinomialtheorem,
n

n
n nk k
(x + y) = ( )x y
k
k=0

Noticethatifx = 1 andy = 1 wehave,

n

n
n
2 = ( )
k
k=0

Noticethatifx = 6 andy = 4 wehave,

n

n
n nk k
(2) = ( )6 (4)
k
k=0

Whatvaluescouldxandyhaveinordertogeta3 ontheleft? n

Whatvaluescouldxandyhaveinordertogeta2 inthesummation? k

Ifyoucanfindvaluesforxandywhichsatisfyboththosequestionsthenyouhavesolvedthe
problem.

FurtherMotivation:

http://math.stackexchange.com/questions/525266/provesumbinomnk2k3nusingthebinomialtheorem 1/3
2/27/2017 discretemathematicsProve$\sum\binomnk2^k=3^n$usingthebinomialtheoremMathematicsStackExchange
Youcorrectlyguessedx=1andy=2.

AsfarasIknowthereisnoformalalgorithmicapproachthatwillsolvethistypeofproblem.
Thisisinfactagreattoolforteachingproblemsolvingskillsbecauseitinvolvesthinkingabout
aproblemfromdifferentangleswithoutaninitiallyclearpathtothesolution.

Whenlookingataproblemyoushouldfirstwriteitdownandexamineeverypiece.Wehad
n

n
n k
3 = ( )y
k
k=0

Thefirstthingwhichstrikesmewhenlookingatthisisthatwehavebinomialcoefficients.This
suggeststhatwemayfindgreaterinsightbylookingatthebinomialtheorem.
n

n
n nk k
(x + y) = ( )x y
k
k=0

Comparingthestatementofthebinomialtheoremtoourproblemwenoticethatbothhavea
summationwithbinomialcoefficientsequaltothepowerofanumber.Firstletuscomparethe
resultsoftherespectivesummations.
n n
3 (x + y)

Weseethatinonecasewehavethenumber3raisedtothen'thpowerandintheothercase
wehavethenumberx + y raisedtothen'thpower.Wesuspectthatthesenumberswillhave
tobethesameforthebinomialtheoremtobeusefulinsolvingtheproblem.Howeverthereare
twodegreesoffreedominthesecondnumberintheformofthevariablesxandythismeans
thereisnotaunique(x, y)pairwhichwillproducea3whenaddedtogether.Note(1,4),(0,3),
(1,2),(103,100),etc.alladdtothree.

Thismeansthatwehaveonlypartiallydeterminedthesolutionbycomparingtheresultsofthe
summations.Tonarrowitdowntoasolutionwecomparethesummands.

n k
n nk k
( )2 ( )x y
k k

Bothtermshaveidenticalbinomialcoefficientswhichmeansthatwecanignorethem.Onehas
powersofasinglenumber,2,theotherhaspowersofourvariablesxandy.Wenoticethat2
andyarebothraisedtothekwhereasxisraisedtothen k .Thissuggeststousthatitmay
beusefultoassociateywith2.Ifwedidthiswecouldsaythatwebelievey = 2butarenotyet
confidentofthevaluexshouldhave.

Thinkingbackwerememberthatoneofthepairsofpossiblexandyvalueswas(1, 2)thisis
hopefulsinceitidentifiesy = 2aswedesire.Ifx = 1 doesn'tcauseanyproblemswewillhave
solvedtheproblem.

Examiningthebinomialtheoremwith(1, 2)weseethat,
n

n
n nk k
(x + y) = ( )x y
k
k=0

n
n nk k
(1 + 2) = ( )(1) (2)
k
k=0

n
n k
3 = ( )(2)
k
k=0

Whichistheidentitywewantedtoestablish.

Atthispointweknowthepaththroughthewoods.Thesolutionis:Evaluatethebinomial
theoremforx = 1andy = 2andtheresultisthedesiredidentity.Thisislogicallyimpeccable
butcontainsnonofthethoughtthatwasnecessarytoproduceit.Oneadvantageofthisisthat
ifweareluckyenoughtoguesstheright(x,y)wecansolvetheproblemevenifwedidn'thave
agoodreasonforcomingupwiththeorderedpair.Thedisadvantageisthatwedon'tlearn
anythingaboutsolvingotherproblemswhenweseejustthebaresolution.

Youlearntothinkthiswaybysolvingalotofproblemswithoutclearlymarkedpathstothe
solution.Ifyouareinterestedinlearningmoreabouthowtothinkinthiswayyoushouldread
"HowtoSolveIt"byGeorgePolya.

editedOct14'13at1:36 answeredOct14'13at0:51
Spencer
7,183 11 43

Thanksforthereply.Woulditbex=1andy=2?Ifitis,Isimplyguessedthat,isthereaproperwayto
calculatethis? StackPWRequirmentsAreCrazy Oct14'13at1:09

http://math.stackexchange.com/questions/525266/provesumbinomnk2k3nusingthebinomialtheorem 2/3
2/27/2017 discretemathematicsProve$\sum\binomnk2^k=3^n$usingthebinomialtheoremMathematicsStackExchange
Yourguessisright.Idon'tthinkthereisa"formal"algorithmtocalculatethat.Thisproblemdoesrequire
somelateralthinking.I'lladdsomemoremotivationtotheexplanationnowthatyou'veguessedtheright
value.Theimportantthingtotakeawayintermsofproblemsolvingistolearntoaskyourselfthetypesof
questionsIhavebulletedabove.SpencerOct14'13at1:11

Youarealmosttherethevalueofxshouldbe1ratherthan2,sothatweget(essentiallywhat
youalreadyhavewritten):
n n

n n
n nk k
n k
3 = (1 + 2) = ( )1 2 = ( )2 .
k k
k=0 k=0

answeredOct14'13at6:23
AG.
1,041 3 13

http://math.stackexchange.com/questions/525266/provesumbinomnk2k3nusingthebinomialtheorem 3/3