UNIT-5 CORRELATION AND REGRE! SION ANALYSI: Correlation is a statistical measure that expresses the extent to which two variables are linearly related (meaning they change together at a constant rate). It's a common tool for describing simple tatemeent about eause and effect. ATION C NT: 1c * The correlation coefficient ‘r’ lies between -1 to +1. + The correlation coetti the variables + The correlation coefficient ‘r° is independent of change of origin i.e. the value of r is not affected even if each of the individual value of two variables is increased or decreased by some non-zero constant, + The correlation coefficient ‘ris independent of change of scale i.e. the value of r is not affected even ifeach of the individual value of two variables is multiplied o divided by some non-zero constant. lependent of the units of measurement of t “1° is the pure number and is i TYPE OF CORRELATION: ‘+ Positive Correlation: when the values of the two variables move in the same direction so that an inerease/decrease in the value of one variable is followed by an increase/decrease in the value of the other variable. + Negative Correlation: when the values of the two variables move in the opposite direction so that increase/decrease in the value of one variable is followed by decrease/inerease in the value of the other variable. ‘+ No Correlation: when there is no linear dependence or no relation between the two variables. year or non-linear corel: The nature of the graph gives us the idea of the linear type of correlation between two variables. If the graph is in a straight line, the correlation is called a “linear correlation” and if the graph is not in a straight line, the correlation is non-linear or curvi-linear. an METHODS OF DETERMINING CORRELATION: © Scatter Plot ‘+ Karl Pearson’s coefficient of correlation ‘* Spearman’s Rank-correlation coefficient. SCATTER PLOT (SCATTER DIAGRAM OR DOT DIAGRAM): Scatter Plots (also called scatter diagrams) are used to graphically investigate the possible relationship between two variables without calculating any numerical value. In this method, the values of the two variables are plotted on a graph paper. One is taken along the horizontal (X-axis) and the other along the vertical (Y-axis). By plotting the data, we get points (dots) on the graph which are generally scattered and hence the name ‘Scatter Plot’. The manner in which these points are scattered, suggest the degree and the direction of correlation. The degree of correlation is denoted by ‘r’ and its dire: positive and negative. n is given by the signs4G4 6% bbb 4 add GCUCEUSSEVUUUOUUOOOObCCOES “ef wed aoe oes! wes ae roe ee) ¢ oF / [fall pois tie ona rising # Wall points Hie on s filling sensi + Uthe points lie ia namow sip. ris ) Wethe points Ne ins namow scrip. fill Ree) ¢ Withe points are spread widely over a brood scrip, paxithe (se fi.) A the points are sqread widely over s brad sep. filling downward, the oomelasion is Low deere negative (se fis.) If the points are spread (Scstterad) withoot any specific pumem. the comeletion Bs absent Le r= OL degree of negate (Se upwards the comeistion is low Gee KARL PE ONS COEFFICIENT OF CORRELATION: It gives the precise numerical expression for the menare of comeistion, Eis denoted by ¥. The Nalue of °F gives the msgnitade of correistion snd its sign demvtes its dieedhon, The mechematical Semel for computing ris Where * DIRECT METHOD: this method over the cehers fn thet it can be ead even when the actual valoss of hams axe exiown, Foe exemple if you want know the corelasion hemes honesy sad wisdom of he bey use this method by giving ranks to the hays Irom sho be ed to find the dos Judgments of two examines oe two jodges The formals isS 666Gb GOS CbbbbEdd ddd dd bdde 6 when no value repeated. Ny when more than one valuc is repeated Problem 1: Calculate the correlation coefficient for the following xX 7 4 8 6 5 ys 6) 5 9 8 2 x y XY x ¥ 7 6 a2 49 36 4 5 20 16= [25 8 9 72 4 81 6 8 48 36 4 5 2 to. | 25 [4 30 30 192 190 | 210 From the table, ‘To find correlation coefficient: i NEXY-DXEY VNDX? ~ (2X) YNEY? = (Y)? 5(192) — (30)(30) Y5C190) — GO} V5210) = GoF 960 — 900 (70742247) 60 6.6025culate the correlation coefficient for the following date Xx os WO Bs Rt B26 Yo os7 96) tor 939 — x Y : 128. 87 aoe io [96 5216 135 | _101 a Todt | dos’ {9801 | 140 | 20440 _| 16 | 80972 Vy, 88Y WS From the table, ‘To find correlation coefficient: NEXY-EXTY YNEX? = (EX)PVNEV? = BY)? 6(80972) — (782)(616) Teci02450) — (782) [6(65036) — (616)? 485832 — 481712 Yeia700 — 611524V390216 — 379456 _ 4120 ~ Yar7evi0760 _ 4120 = ]6.356)(103.73) _ 4120 ~ 5845.60788 Problem r > > > v . > S Ss > ~S ~S ~»> ~»> ~~» ~~ ~> ~~» ~ ~ ~ ~> > = > > a) Calculate the correlation coefficient for the following da xX 2% 40 2 2M 2 46 yom 6 W 3 9 14 Solution:dE ECO CEI Eble ddd ddd dd dd ddsdddddds xX Y XY x ¥ a 7 196 784 49. 40 | 6 240, 1600) 36 25 10 250 625 too) a 3 63 441 9 32 9 198_|_ 484 81 6 ir ia_[ 2116 | 196 isz_ | 49591 | 6050 cul Wem the table, YN= 182 Yye49 CNY = 1591 YN? = 6050 Yvan N ‘Vo find correlation coefficient: NUXY-EXEY VNEXE= OXF YNEY? = (LY)? 6(1591) ~ (182)(49) J6(6050) = 182)" f6CA71) = 9)? 9546 - 8918 © V30300 — 331242026 — 2401 28 Vat76Va25 628 (G6356)(20.615) 628 161.7894 Calculate the correlation coefficient for the following data: X 25 26 (27803235, Y 2 22 2 25 2% 27 34 Solution: x Y XY x ¥ 25 20 300 | 625 | —a00 26 22 S72 | 676 | asa 27 24 648 | 9 | 576 28 25 700 | 784 | 625 30 26 780 | 900 | 676 32 27 864 [1024 [799 35 34 [1190 | ~1235 | ise 203 [178 | 5284 | 5963 | ~aeaeFrom the table, LV = 4646 7 To find correlation coefficient: —_NEXYTEXEY __ YNDX?— (2X) VND? — (Ly)? _ 7(5254) — (203)(178) © T5963) — 203)" /7(4646) — 178)? _ 36778 — 36134 "* Yatrat — 4120932522 — 31604 644 r= V532V838 644 r (723.065)(28.948) 644 667.68562 Cee eeeooe cee eeee”™ Problem S: ~ ai Calculate the correlation coefficient for the following data: e xX 2 mM 6 6% 27 «27 2% 28 > Y 18 2 20 2 2 27 24 2 sS Solution ~ x ¥ x¥ x ¥ ~ 22 18 396 48a_[ 324 24 20 480_| 576 | 400 | 26 20 320 [676 | 400 26 24 @24_| 676 576 =o 27 22 594 | 729 484 27 27 ne | ne | 129 = 28 24 672 784 576 28 21 588 784 [441 = 29 25 725 841 625 30 29 ‘370_| 900_| 8a = 267 | 230 | 6198 | 7179 [5396 = From the table, . = DX =267 = TY =230 IXY = 6198 o Ex = 7179 Le —) as mtLY? =5396 N=10 ‘To find correlation coefficient: NEXY-DXTY 61980 — 61410 5870 V501V1060 870 570 728.72331 Probl Calculate Karl Pearson’s coefficient of correlation for the foll (22.383)(32.557) eee INEX? — (LX)AYNEY? - (ZY)? 10(6198) — (267)(230) 1006198) = 267230) 10(7179) — (267)2/10(5396) — (230)? V71790 = 71289V53960 — 52900 X 6 8 2 15 18 2 24 yY 0 2 1 15 18 2 2 X-X | y=¥-¥ 3 5 cc exas| "sya | x y 6 | 10 =I2 5 108 144 31 3 | 12 =10 7 70 100 49 12 [15 6 4 24 36 16 15_|_15 3 4 12 9 16 18 | 18 0 =I 0 0 1 20 | 25 2 6 12 4 36 24 | 22 6 3 18 36 9 28 _|_26 10 7 70 100 49 31 | 28 3 9 17 169 81 162 | 171 0 0 431 308 | 338 From the table, Sxy=431 Ex = 598 Ly =338To find correlation coefficient: xy 431 V598V338 431 (24.454) (18.384) 431 449.562336 Problem 7: ‘Two judges in the beauty competition rank 12 entries as follows: xX 1 2 3 4 5 6 7 8 Yoreo9 6 0 3 5 4 7 Find the rank correlation? Solutio Xx [Ya [d= Xn- Ye @ 1 [12 =i 121 2 9 7 49 3 6 3 9 4-10 6 36 5 3 2 4 6 5 1 1 7 4 3 9 37 1 1 9 | 8 1 T 10 [2 3 64 [it 0 0 2p 1 iW 12h [yas] pa a6 To find rank correlation: 6ya? P= tla?) 6(416) o=1- aaa feel aaa —1) -[ 2496 | 2043). _ pa a 11776] ep — 14545 10 2 u 2Problem 7: Marksin Economies S060 65070750 40,7080 Marksin Statistics 80 716075 9D, 827050 X [| Y | Xe | Ya [d=Xx d so| 80 | 7 | 3 4 16 60 | 71 | 6 5 1 1 6s | 60 | 5 7 a 7o | 75 [35 [4 0.25 75_| 90 | 2 1 1 1 40 [sz [8 [2 6 70 | 70 | 35 | 6 25 so [so 1 | 8 =I d=0 When m=2 mim?-1) _ 202-1) _ 24-1) _ 20) _ 6 Da os To find rank correlation: (rar 6113.5 +05 -Faea 6(114) 1a] 684 ~ laces! 684} ~ Goa Regression is a statistical procedure that determines the equation for the straight line that best fits a specific set of data. FORMULA: + Regression equation of ¥ on X: Y-¥=byx(X-X) byx is called the regression coefficient of Y on X.AY= (ENN) Ny © Regression equation of Now Vs X=Kedy-) day iscaltad the rgeession evelticient of N ow Vy yYAY- (yy by yee nyve-(eyy? Peoblem 8 Vind the regression tine Von X fir the data: - } From the table, n equation Y on X: Y-Y=byx(X—X) where by = To find X.¥, byx bay = NEXY= OEY) 8 = NEE (EX) _ 546) = (15)(15) ~~5(85) — (15)? =x Nn alaelaYee byte 8) Yak = ve ay Yas QIN 08 Year on rd Ve naxnnny, Pooblom. ss Wind the vngeesston Hane Nowe Ve Be the data XW dW Yow wo mM 6 Ww Solutions xv y tw is 180) mw WW 2 Los tt Ww uM 12 576 33 6 12 iw 20 30. 7 900 0 Jo Toso 10 126 272 Pron the tables YNe uo Ye ns Yxve-27 Yar Neo Regression equation X on V! bee P= AD where yy = To find XY, Dyyt yx 5 © oH X= NON) NyVF= QV)? _ 6(2772) = (120) 126) (8276) = (20)? _ 62 — 15120 T9650 — 1HN7E AS12 3780,| edbbbdEEE vb dbf xX 2 6 29 30 31 31 34 35 og 20) 20 ee pole 29) cere eee ed Solution x ¥ XY x y! 2 20 wa0_| 484 400) 26 20 520 676 400 _| 29 21 609 841 aay 30 2 870 ‘900 84 31 27 837 961 29 31 24 744 961 576 34__|_27 918 | 1156 | 729 35 31 Toss_| 1225 | 961 238 199 | 6023 | 7204 | 5077 From the table, Regression equation X on Y: X—K = byy(¥—¥) where byy = "2 7 bey -NEXY- NEY) ND Y= EY) 8(6023) ~ (238)(199) 8(5077) — (199)? 8184 — 47362 ~ 40616 — 39601 822 1s= XAN X= 29.75 = 2927) by =) O.8O99(Y = 24.875) ONY = 20.1451 X= QROVY = 20.14514 29.75, XSEONIVOVA9.6049 Regression equation Yon XN: Y=¥ = dyg(X—N) where byy ‘Vo find byx: NEXY~(EXMEY) NEN (EN? _8(6023) ~ (238)(199) © 87204) = Gan) 48184 — 47362 © $7632 822 ~ oan yx = > Y-¥ Y= 24.875 Y= 24.875 Dyx(X = » Problem 1 Obtain both regress AMC) ny x o4 Soya? nS vy o6 +t 0 o 2 1 5 Solution: x T 5 3 2 1 1 T 3 2 From the table, px=23 DY=16 EXY=36SUG bEKOOOBUUVODEOOOEEEEEEHEOEHDEBHEE Regression equation X on Y X=X= day (VV) where bey = xy - (2) NEY?-(lyp _ 8(36) - (23)(16) 8(68) ~ (16)? _ 288 ~ 368 ~ 544—256 eeu NEVEG. To find KV: IX _ 23 ! uM Dyy(¥-¥) =0.278( — 2) -0.278Y +.0.556 -0.278Y + 0.556 + 2.875 X= $0.278Y4+3.431 Regression equation Y on X: Y—¥=byx(X—X) where byx = To find byx: NEON p,, -NEAY= GY) "x= NN (DN) 8(36) ~ (23)(16) 8(99) Y-Y¥ = by(X —X) 03 (x — 2.875) Y-2=-03xX4 o1bo5 Y= -03X 408685 +2 WECM AKA BOIS 1 -03%4 9.8695166466666668 ited. LV bt b&b & bb fy bb beddbbdddbddd444 é PP fF eel een oprernea ft carer gene fi sores F set 6 a u aaey Find the correlation coefficient and regression lines for the data: x1 Y ‘From the table, 1s, To. ind sorpelstion coefficient: \ NES ESEY aS Jeawe- Ow NEYE- OD \5(88) — (585), “YEO sy (sCS1\= 25)" 735 = 625 To find regression equation: Regression equation X on Y x= = byy(¥—¥) where by =~ To find X,¥, Byy: Ex 15dddGGOGCOCCEEEA” , G veddudddd bud Q Lhe ee @vaued Dyy = NEXY= ENE) wT ND Qy)? _ 5(@8) = (15)(25) ~8(151) = (25) Regression equation Y on Xz oo Y~YV = byx(X—X) where by, = ASEAN ONE EX To find by: by, = NEXY- NEY) eS NEX = (LX)? — 5(88) — (15)(25) -S65- a5" To Find worrreledion loth icaant _ 440-375 ow: + bry: byx Find the correlation coefficient between x and y when the lines of regression are 8X-10Y+66-0, 40X-18Y=214. Given: 8X-10Y466= 0 40X-1BY = 214 Suppose 8X-10Y466 = 0 is a regression equation of Y on X:ECEEEESLESS ESE E4G SO 8X — 107 + 66 -10Y 0 8X — 66 yo 28X86 =10” =10 Y=0.8X+6.6 = DRE ‘Suppose 40X-18Y = 214 is a regression equation of X on Y: 40X — 18Y = 214 40X = 214 + 18Y 214 | 18Y X= 5.35 + 0.45Y = by=O4s To find Correl: r= Dix Day coefficient: Equation of two lines of regression are 12X-15Y+99 = 0, 64X-27Y = 373. Find © Mean values of X & Y © Correlation coefficient of X & Y? Solution: Given: 12X-15Y+99 = 12X-15Y =-99 = AXSY =-33 64X-27Y = 373 ssseeeeen(3) To find mean value of X & Y: ‘To solve equation (2) & (3): (2) *16 @)¢ yRLsaunt , Ab 4 1 ~~ i ~. Nnbytitvete V vate by Rquathon QU: 1 aXny est ANS) = at a~ AY =e ht AY MES a~ Ay 82 s~ yee =~ Newent Therefore, :~ STM VALE OF NETS SMe vate of VEIT Vo Hud correlation coefficient 66 a> Suppose L2N-ISV199 © 0 iy regression equation of ¥ on Xt As. Lax 15V1.99 = 0 2 =I5Y = -12X-99 =12X 99 1 hg ad > DRO ad Suppose 64X-27V = 373 is a regression equation of X on Ys GAX —27Y = 373 bb bode r= Dy Day (OAzi0B) 375Theory Explain the relation between regression and correlation analysis interconnection oF a co-relationship betwee Variables. In Correlation, both the independent and dependent values have no difference. The primary objective of Correlation is, to find out a quantitative/numerical value expressing the association between the values. Correlation stipulates the degree to which both of the variables can move together. Correlation helps to constitute the connection between the two variables. Theory 2: Regression ‘Regression’ explains how an independent variable is numerically associated with the dependent variable: However, in Regression, both the dependent and independent variable are different. When it comes to regression, its primary intent is, to reckon the values of a haphazard variable based on the values of the fixed variable. However, regression specifies the effect of the change in the unit, in the known variable (p) on the evaluated variable (q). Regression helps in estimating a variable’s value based on another given value. ‘Show that Geometric mean of the coefficient of regression is the coefficient of correlation? Proof Proof Let, (xi.y1), (x2,y2! (XxYa) be the pairs of n observations. Then the correlation coefficient between x and y is denoted by ry and defined as, Lew -70 ty A)sS ~s ~s > ~> ~s ~> = ~> ~s ~s <3 5 Gb8 } aed = ty proved) Theory 3: Show that Arithmetic mean of the coefficient of regression is the coefficient of correlation? Statement? ‘The arithmetic mean of two regression cocfficient is greater than correlation event ie (22°) prot Let (sok Geypione Gage) he pi of m obsonations Then the aa eye Gnkey bean neepeson cote of 08 one by Thearithmeticmean of by andy is A.M ) andthe esomeri means GM= fb, We know, Correlation coefficient is the geometric mean of regression coefficients. ie. t= JB, =F, Since, A.M 2G.M on, (Pe) = oh, Theory 4: If 8 is angle between the two regression lines then show that tan@ = Proof:cave ver resins cate wake 0 bocause <6 9207 SO oto, Derive the formala Oyy = 07, 07y + 270,07 Proof. Coefficient of correlation of xandy is and 62, 0; and o?., are varience of x, yandx—y a, = Latn+(rst a, = 2-9 +0)+P dia-3)+0 +9F , of, = hea + +9) 28-3) D) af, = 2ia-8F]+ Ay -981-25-R7-9) Of, ,) = 08 + Gf — 21040, 20,6, = 08 +o? - of» _ oto =) ee thatr = vy