Top Banner
1 Chapter 10 Correlation and Regression Chapter 10.1 Correlation
11

Chapter 10 Correlation and Regression Chapter 10.1 ...

Mar 02, 2023

Download

Documents

Khang Minh
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Chapter 10 Correlation and Regression Chapter 10.1 ...

1

Chapter 10Correlation and Regression

Chapter 10.1Correlation

Page 2: Chapter 10 Correlation and Regression Chapter 10.1 ...

2

Define the following terms in your notes:  correlation, regression, simple relationship, multiple relationship (multiple regression), correlation coefficient, PPMCC, lurking variable, population correlation coefficient

Then  read p. 531 (starting at 10­1) to example 10­1 on p. 532.

Page 3: Chapter 10 Correlation and Regression Chapter 10.1 ...

3

­simple correlation studies = collecting data on 2 numerical or quantitative variables to see possible relationship

ex:  Create scatterplot to see relationship of the data:

Student Hrs. of Study Grade (%)A     6     82B     2     63C     1     57D     5     88E     2     68F     3     75

Page 4: Chapter 10 Correlation and Regression Chapter 10.1 ...

4

Correlation Coefficient­range from ­1 to 1

0­1 1

strong negativelinear 

relationship

strong positivelinear 

relationshipno linear

relationship

*examine p. 534, figure 10­3 for examples

Page 5: Chapter 10 Correlation and Regression Chapter 10.1 ...

5

Computing Correlation Coefficient  (r)­determines strength of linear relationship­no units, always round to 3 places­formula:

r =  _________________________  n( Σxy) ­ ( Σx)( Σy)√[n( Σx2) ­ ( Σx) 2] [n( Σy2) ­ ( Σy) 2]

Page 6: Chapter 10 Correlation and Regression Chapter 10.1 ...

6

ex:  Compute correlation coefficient for yesterday's hours of studying vs. grade

Hrs. of    GradeStudent  Study      (%) xy x2 y2

A 6 82B 2 63C 1 57D 5 88E   2 68F 3 75

Page 7: Chapter 10 Correlation and Regression Chapter 10.1 ...

7

­If significant relationship exists and null can be rejected for specific  α, consider that:

1. there's a direct cause­and­effect relationship 2. there's a reverse cause­and­effect relationship3. the relationship may be caused by a 3rd variable4. there may be a complexity of interrelationships     among variables5. the relationship may be coincidental

Page 8: Chapter 10 Correlation and Regression Chapter 10.1 ...

8

Population Correlation Coefficient  ( ρ)­sample coeff. can be used to estimate pop. coeff. if:

­x and y are linearly related­variables are random

­to test  ρ, null and alternative will always be these:H0:   ρ = 0H1:   ρ ≠ 0 always trying to see if sig. 

linear relationship exists, so use 0

Page 9: Chapter 10 Correlation and Regression Chapter 10.1 ...

9

t­Test for Correlation Coefficient­formula:

t = r where d.f. = n ­ 2, n = # of ordered pairs (x, y)

*use t­table in Appendix C, Table F

ex:  Test the significance of the correlation coefficientof studying hrs. vs. grade from earlier, with  α = 0.05

n ­ 21 ­ r 2

Page 10: Chapter 10 Correlation and Regression Chapter 10.1 ...

10

Additional Way to Test Corr. Coeff.­Appendix C, Table I­if r falls outside of  ± the # from table (closer to  ±1), reject null

ex:  Using Table I, test  α = 0.01 of corr. coeff. of r = ­0.176 if sample size is 8.

Page 11: Chapter 10 Correlation and Regression Chapter 10.1 ...

11

ASSIGNMENTp. 539­40 Applying the Concepts

p. 541 #12, 14