Scribd
Upload
50 views20 pages

Regression Analysis

Uploaded by

Ramandeep Kaur Brar
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
50 views20 pages

Regression Analysis

Uploaded by

Ramandeep Kaur Brar
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
REGRESSION ANALYSIS (9) tincar R Linear Regression ‘The lea or dictionary meaning of the word “Repression” is ‘stepping hack returning 10 he verge sale’ The term wa at ted by rsh bomrican St Fant Galion othe tp he ‘aces in commen wath ses he mao ting text oc sae of so of tall pets rovers ortegieaes Sack to the mean sare of the population. He stated the ‘Sicip etveu chip of chute cd Ets nd sms pobted te rs pope "egress nara Mery n Heredia Stare The nese eats is ty were (Teta fates Rave als nd sha thers ae sh sone (i) The average igh of the sons of group fl fathers ss than hat fhe fates andthe average night sons of tuo sh fas ne than ha oe fates Troe words, Galo’ sis reveled that the of ings of sora oho pret tn ever a tp heck the avenge Delgo the populate phenomenon which Gaon escbed a Regrennon Medic He comida if the average eight of certain grup of Tate i cms. above (below) the sel arr ght at ergs nectar as Se (op ca shore ir pl ‘Serge ph whe he cela splice teen te bight oft ges poupo flee ad ‘arse tn ths csc conan pontine and cel rts we nee aT opr tein Uhabove Bat ny te word epson me inSain mich wider penpective witout any ference tw ome Repsson sala ine gens. mean he osiato pc ol waa alarms veel Eos toe al of ear vee. Wie of it vy etd ata fio whch cuemely edn ano all ince ~ trl acl a pipes opi en ‘Siinlr an economist study te retannip between to or mr ate that aia aualy {nn cstmation of den spi oes So mtn, rn and cnsomption tne Prion retina noe ofthe ahr poems is smot all pores of aman ati, THe ‘timation or prediction of ee pectin. cemsamption, pie, lvenes alk. fs. nce ‘Se-ofparamoune importance to a busnesxman fect. Population estimates and popuaion [jects ar ndspensble fr ecient plano of an economy The plarmacetial cncems re [orn in stdhing cman the eer frre on pce Regeon aa ne he ‘enya echniqus fr mang sich prin Inthe wer f MIM Dlr "Represent “utematical mesure ofthe sere tansy Between moo mit varies nm of the ginal rt fhe dat ‘We come actos mb of ntr-elatelevet nou di 1 yie.Fo inst the ed of sop depends on te cual the ont or pct fs Pree depen ce the pram an aberting ‘xpi, the dean for tier roi eprom rice, een fers depends on [iS icime. and Soon The fessionals confined the st) of ely two vars at ie termed ae simple regresion. Bt gute stent vas fa partir cane may be feted by maple of fac The ze anal Yor staying ethan fo vals a es known a ep regresion However tn har weal tae omelnen sole ere 92 busmess sransncs tn sper analyte ar top of vrs The vara wow ale ie intend te pried in clled dependent sari ad the variable which nen the salut o te fot ‘relictin lied Inne variate In reprvsi aa nspeniet vara a wnat ‘ere pec pa we ne he aba Ke eed ped }2. LINEAR AND NON-LINEAR REGRESSION 9° uses sranisnics {a reprnsion analy here a two type of vrs The vara whan ale einen i> te preted i called dependent voile a the variable which influence the value or ted ot {rede calledindeenden vari. tn repression aa inne vara lo kona [eres oper or eptantorwhedepenket Yaris pan reese et mpl 92. LINEAR AND NON-LINEAR REGRESSION the given bivariate dia ae pote on a graph he pins baie on the sete diagram wil more or lt concent nd cae, called the seo eresson “Of nich crv yt ict nde comeing sd sometses complicated. The mathe egetion he eee eee ‘thle he ereom euatncables sot The aerge change hh ae the depen Sable foray yen valu nent are the epson crv tsa ine, ney tha her ina repression twee the vas der sty. Tv equ fsck acu the eon of ah nar dope ean Sn ‘he satble rand ncn of na epson tales of he Spenser by Sonat ‘Seotnc amount fr a wit change athe ve ofthe llependen vars, However, Ifthe cuve of tezreina ota aight ne, the regrenon scree as cued or n-lnar repos, Te reson ‘Seon willbe funcinal claton tween aly sveving term andy of dee tan ow 12g tr of te pes? ya However hs chapter mesa confine our Scion © Hncaneressn hte cles nl 93. LINES OF REGRESSION Line of epson ethene mich pv the et esi fae aril for any piven ale of the ‘atv vhs 6 Varies and, Weal hn ones repress Bo fy a he ‘erotica. Delniton. Lie of rgrstion of yo. the ie which ive the be! exit forthe vl of for any pected vane of Sit no erin font Bie which ses the Bes estima forthe vale af ay speced vac. “he term bes finer in asodance wih the Principe of Least Sree which omits in ‘mann emo the aguas of he eas orth ror of exit, he devon vce he ‘em obverse lcs of he arial and er coresponding earned esa sien te ie of est {iW may mins sm fhe ue he ere rel toa par as the feme ‘ermiiming te smo quae of ror pul y-0x) giver equi of he eof egreion ‘fom and tear mnnsing eS fares teres prall vans ies he eton ‘fein of repression oon ‘Weshallexpain Below te eto of dessin he ation feline of eeremian a ya P34. Derivation of Line of Regreslon of yom x. Le (5 Ye ye Yao Bm als oF sures on wove aj une sty. Le yaaeoe on eth in of eres (et of y08 or any given plat Py 9) the seater agra cor fete al 2 en then of best (1) FH, Now the cordate fd) same a tat of Pvc ance (he nth in the yeni of Wyte, AS ven by hy), Hence, he rf estimate fr P ‘givens PPM y+) oy Paes LINEAR REGRESSION ANALYSIS 9° ‘This the enor ural whe ya) ote th poi, We wil have sch rf al he ols seater dagram Forte pois wich sbove the nether ol be ponte fre os wh ietelow th nthe cor mou negate, ‘Acorn he priciest ures, we have determine the const anf 1) sh ‘hs teste thee of he ro esti mnie ther wonder Pad raed o age oy bjt to vaio ina. "Wem ao write Eas Ea Yy-y= 20 -0-d07, REGRESSION o Regression Analysis measures the nature and extent of the relationship between two or more variables, thus enables us to make predictions. o Regression is the measure of the average relationship between two or more variables. | UTILITY OF REGRESSION o Degree & Nature of relationship o Estimation of relationship o Prediction o Useful in Economic & Business Research DIFFERENCE BETWEEN CORRELATION & REGRESSION o Degree & Nature of Relationship » Correlation is a measure of degree of relationship between X & Y « Regression studies the nature of relationship between the variables so that one may be able to predict the value of one variable on the basis of another. o Cause & Effect Relationship » Correlation does not always assume cause and effect relationship between two variables. » Regression clearly expresses the cause and effect relationship between two variables. The independent variable is the cause and dependent variable is effect. ALLOd “Sossqoug TERsrssy "ysuNg sopuL DIFFERENCE BETWEEN CORRELATION & REGRESSION o Prediction « Correlation doesn’t help in making predictions « Regression enable us to make predictions using regression line o Symmetric « Correlation coefficients are symmetrical i.e. r,, = r,,. « Regression coefficients are not symmetrical i.e. b,, # by,. o Origin & Scale » Correlation is independent of the change of origin and scale « Regression coefficient is independent of change of origin a) AuLDd “tossqjoag wee sissy “yung sopanig but not of scale TYPES OF REGRESSION ANALYSIS o Simple & Multiple Regression o Linear & Non Linear Regression o Partial & Total Regression SIMPLE LINEAR REGRESSION Simple Linear Regression Regression Regression Regression Lines Equations Coefficients & ALd “sossqjosg istesy “ysuTg sopuLngy REGRESSION LINES © The regression line shows the average relationship between two variables. It is also called Line of Best Fit. o If two variables X & Y are given, then there are two regression lines: * Regression Line of X on Y Regression Line of Y on X o Nature of Regression Lines Ifr = +1, then the two regression lines are coincident, » Ifr=0, then the two regression lines intersect each other at 90°. « The nearer the regression lines are to each other, the greater will be the degree of correlation. « Ifregression lines rise from left to right upward, then correlation is positive. ALOd “2ossjUY JURSIESy “‘ysUIE sopUL REGRESSION EQUATIONS o Regression Equations are the algebraic formulation of regression lines. o There are two regression equations: » Regression Equation of Y on X oY=a+bX o Y-Ff = byx (xX -%) o¥-Yer. ox = -X) « Regression Equation of X on Y oX=a+bY oX-¥ = bry (Y -¥) oX-f=r, =(¥ -f) 6 BLLOd Jossqqay TUR “Ysug JopULIG REGRESSION COEFFICIENTS o Regression coefficient measures the average change in the value of one variable for a unit change in the value of another variable. o These represent the slope of regression line o There are two regression coefficients: 6, » Regression coefficient of Y on X: b,, =r. = a » Regression coefficient of X on Y: by, = 7. y ‘SLLOd Yossqgaig weyissy ‘ystutg JepuLig PROPERTIES OF REGRESSION COEFFICIENTS o Coefficient of correlation is the geometric mean of the regression coefficients. i.e. r = Vbxy - byx © Both the regression coefficients must have the same algebraic sign. o Coefficient of correlation must have the same sign as that of the regression coefficients. © Both the regression coefficients cannot be greater than unity. o Arithmetic mean of two regression coefficients is equal to or greater than the correlation . : bxy+byx coefficient. i.e. = >r BLLOd “20esajouy IUMstey “YsuIg pug o Regression coefficient is independent of change of © origin but not of scale OBTAINING REGRESSION EQUATIONS Regression Equations OR shared Regression Coefficients Using Normal Equations REGRESSION EQUATIONS IN INDIVIDUAL SERIES USING NORMAL EQUATIONS o This method is also called as Least Square Method. © Under this method, regression equations can be calculated by solving two normal equations: * For regression equation Y on X: Y=a+bX o LY = Na + bX o EXY = aX + bx? » Another Method _ NLEXY ~EXBY @ -7_ ob, = NEX? (EK) a=¥ —bX ALOd “Joss THEstssy “yung sopuLng >» Here a is the Y — intercept, indicates the minimum value of Y for X = 0 o & b is the slope of the line, indicates the absolute increase in Y for a unit increase in X. REGRESSION EQUATIONS USING REGRESSION COEFFICIENTS REGRESSION EQUATIONS USING REGRESSION COEFFICIENTS (USING ACTUAL VALUES) o Regression Equation of Y on X + Y-V=b,, K-%) where b,, = == NIX? (Ex) o Regression Equation of X on Y » X-X=b, (Y-7) where b,, = SEAT Q3: Calculate the regression equation of Y on X & X on Y Y= 138X411, X=0.5+0.5Y @ ALD “sossqjosg ws esy “yung sopaLng REGRESSION EQUATIONS USING REGRESSION COEFFICIENTS (USING DEVIATIONS FROM ACTUAL VALUES) o Regression Equation of Y on X « Y-Y=b,, X-X) where by Se o Regression Equation of X on ee * X-X=b,, (Y—¥) where b,, ==> = Q4: Calculate the regression equation of Y on X & X on Y using method of least squares: Y= 0,26X + 3.2, X = 4.75 + 0.45Y Xx 3 i i r nt ¥ 4 2 5 10 3 6 q REGRESSION EQUATIONS USING REGRESSION COEFFICIENTS (USING DEVIATIONS FROM ASSUMED MEAN) o Regression Equation of Y on X N Edxdy -Zdx Ed © ¥-P=b, &-2) whereb, = TSee o Regression Equation of X on Y + X-X=by (Y-¥) where b,, fl texiym ex Bey N

You might also like