calculating the least squares regression line

17
Calculating the Calculating the Least Squares Least Squares Regression Line Regression Line Lecture 49 Lecture 49 Secs. 13.3.2 Secs. 13.3.2 Fri, Apr 28, 2006 Fri, Apr 28, 2006

Upload: garth-burris

Post on 30-Dec-2015

27 views

Category:

Documents


0 download

DESCRIPTION

Calculating the Least Squares Regression Line. Lecture 49 Secs. 13.3.2 Fri, Apr 28, 2006. The Least Squares Regression Line. The equation of the regression line is y ^ = a + bx . Thus, we need to find the coefficients a and b . The formulas are. or. Example. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Calculating the Least Squares Regression Line

Calculating the Calculating the Least Squares Least Squares

Regression LineRegression LineLecture 49Lecture 49

Secs. 13.3.2Secs. 13.3.2

Fri, Apr 28, 2006Fri, Apr 28, 2006

Page 2: Calculating the Least Squares Regression Line

The Least Squares The Least Squares Regression LineRegression Line

The equation of the regression line is The equation of the regression line is

yy^̂ = = aa + + bxbx.. Thus, we need to find the Thus, we need to find the

coefficients coefficients aa and and bb.. The formulas areThe formulas are

xbya

xx

yyxxb

2 xbya

xxn

yxxynb

22or

Page 3: Calculating the Least Squares Regression Line

ExampleExample

Consider again the data setConsider again the data set

x y

2 3

3 5

5 9

6 12

9 16

Page 4: Calculating the Least Squares Regression Line

Method 1Method 1

Compute the means and deviations Compute the means and deviations for for xx and and yy..

x y x –x y –y

2 3 -3 -6

3 5 -2 -4

5 9 0 0

6 12 1 3

9 16 4 7

x = 5

y = 9

Page 5: Calculating the Least Squares Regression Line

Method 1Method 1

Compute the squared deviations, Compute the squared deviations, etc.etc.

x y x –x y –y (x –x)2 (y –y)2

(x –x)(y –y)

2 3 -3 -6 9 36 18

3 5 -2 -4 4 16 8

5 9 0 0 0 0 0

6 12 1 3 1 9 3

9 16 4 7 16 49 28

Page 6: Calculating the Least Squares Regression Line

Method 1Method 1

Find the sums of the last three Find the sums of the last three columns.columns.

x y x –x y –y (x –x)2 (y –y)2

(x –x)(y –y)

2 3 -3 -6 9 36 18

3 5 -2 -4 4 16 8

5 9 0 0 0 0 0

6 12 1 3 1 9 3

9 16 4 7 16 49 2830 110 57

Page 7: Calculating the Least Squares Regression Line

Method 1Method 1

Compute Compute bb::

Then compute Then compute aa::

9.130

57b

5.059.19 a

Page 8: Calculating the Least Squares Regression Line

Method 2Method 2

Consider again the dataConsider again the data

x y

2 3

3 5

5 9

6 12

9 16

Page 9: Calculating the Least Squares Regression Line

Method 2Method 2

Compute Compute xx22, , yy22, and , and xyxy for each row. for each row.

x y x2 y2 xy

2 3 4 9 6

3 5 9 25 15

5 9 25 81 45

6 12 36 144

72

9 16 81 256

144

Page 10: Calculating the Least Squares Regression Line

Method 2Method 2

Then find the sums of Then find the sums of xx, , yy, , xx22, , yy22, and , and xyxy..

x y x2 y2 xy

2 3 4 9 6

3 5 9 25 15

5 9 25 81 45

6 12 36 144

72

9 16 81 256

144

155 515 28225 45

Page 11: Calculating the Least Squares Regression Line

Method 2Method 2

Then find the sums of Then find the sums of xx, , yy, , xx22, , yy22, and , and xyxy..

x y x2 y2 xy

2 3 4 9 6

3 5 9 25 15

5 9 25 81 45

6 12 36 144

72

9 16 81 256

144

155 515 282

x = 25y = 45x2 = 155y2 = 515xy = 282

25 45

Page 12: Calculating the Least Squares Regression Line

Method 2Method 2

Compute Compute bb::

Then compute Then compute aa::

9.1150

285

251555

452528252

b

5.059.195

259.1

5

45

a

Page 13: Calculating the Least Squares Regression Line

ExampleExample

The second method is usually easier.The second method is usually easier. By either method, we get the By either method, we get the

equationequation

yy^̂ = -0.5 + 1.9 = -0.5 + 1.9xx..

Page 14: Calculating the Least Squares Regression Line

TI-83 – Regression LineTI-83 – Regression Line

On the TI-83, we On the TI-83, we couldcould use 2-Var Stats use 2-Var Stats to get the basic summations. Then use to get the basic summations. Then use the formulas for the formulas for aa and and bb..

For our example, 2-Var Stats LFor our example, 2-Var Stats L11, L, L22 reports thatreports that nn = 5 = 5 xx = 25 = 25 xx22 = 155 = 155 yy = 45 = 45 yy22 = 515 = 515 xyxy = 282 = 282

Page 15: Calculating the Least Squares Regression Line

TI-83 – Regression LineTI-83 – Regression Line

Or we can use the LinReg function.Or we can use the LinReg function. Put the Put the xx values in L values in L11 and the and the yy values values

in Lin L22.. Select STAT > CALC > LinReg(a+bx).Select STAT > CALC > LinReg(a+bx). Press Enter. LinReg(a+bx) appears in Press Enter. LinReg(a+bx) appears in

the display.the display. Enter LEnter L11, L, L22.. Press Enter.Press Enter.

Page 16: Calculating the Least Squares Regression Line

TI-83 – Regression LineTI-83 – Regression Line

The following appear in the display.The following appear in the display. The title LinReg.The title LinReg. The equation y = a + bx.The equation y = a + bx. The value of The value of aa.. The value of The value of bb.. The value of The value of rr22 (to be discussed later). (to be discussed later). The value of The value of rr (to be discussed later). (to be discussed later).

Page 17: Calculating the Least Squares Regression Line

TI-83 – Regression LineTI-83 – Regression Line

To graph the regression line along with To graph the regression line along with the scatterplot,the scatterplot, Put the Put the xx values in L values in L11 and the and the yy values in L values in L22.. Select STAT > CALC > LinReg(a+bx).Select STAT > CALC > LinReg(a+bx). Press Enter. LinReg(a+bx) appears in the Press Enter. LinReg(a+bx) appears in the

display.display. Enter LEnter L11, L, L22, Y, Y11

Press Enter.Press Enter. Press Y= to see the equation.Press Y= to see the equation. Press ZOOM > ZoomStat to see the graph. Press ZOOM > ZoomStat to see the graph.