tutorial08

tutorial08 - 1 ISMT 111 B USINESS S TATISTICS – T UTORIAL...

Info iconThis preview shows pages 1–4. Sign up to view the full content.

View Full Document Right Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: 1 ISMT 111 B USINESS S TATISTICS – T UTORIAL 8 created by Andrew Yam Simple Linear Regression Model § Aim: Use X (Independent Variable, or Predictor Variable) to predict Y (Dependent Variable, or Response Variable) after drawing n pairs of observations, ( 29 ( 29 n n y x y x , ,..., , 1 1 . § Actual value i i i x Y e b b + + = 1 , n i ,..., 2 , 1 = (also shown as i i i x Y e b a + + = ) where i e is the random error § Predicted value i i x Y 1 ˆ ˆ ˆ b b + = , n i ,..., 2 , 1 = § Assumptions of Regression Model - Linearity ( 29 i i i x x X Y E 1 | b b + = =- Constant Variance; Homoscedasticity 2 2 | s s = = i i x X Y i 2200- Normality ( 29 2 , ~ s e N i i 2200- Independence i e and j e are independent j i ≠ 2200 Estimating β and β 1 We use the least square (LS) method to find b and 1 b by minimizing the total sum of square errors ( SSE ), i.e. minimize ( 29 ( 29 ( 29 ( 29 ∑ ∑ ∑ = = = = +- =- = n i i n i i i n i i i e x y y y SSE 1 2 1 2 1 1 2 ˆ ˆ ˆ b b xy yy S S 1 ˆ b- We get 2 ( 29 ( 29 ( 29 ( 29 n x x n y x y x x x y y x x S S i i i i i i i i i xx xy 2 2 2 1 ˆ ∑ ∑ ∑ ∑ ∑ ∑ ∑-- =--- = = b x y 1 ˆ ˆ b b- = Estimating σ MSE n SSE f d SSE s =- = = = 2 . . ˆ 2 2 s , where 2 ˆ s is an unbiased estimator of 2 s . Testing the Model § Hypothesis Testing for the Slope, 1 b Null Hypothesis 10 1 : b b = H Test Statistics: xx S s t / ˆ 10 1 b b- = ~ t- distribution with 2 d.f.- = n i) when 10 1 : b b < a H , H is rejected when a t t- < , ii) when 10 1 : b b a H , H is rejected when a t t , iii) when 10 1 : b b ≠ a H , H is rejected when 2 / a t t- < or 2 / a t t . Note: If : 1 = b H , rejecting H means X is useful for the prediction of Y . A ( 29 % 100 1 a- C.I. for 1 b is xx n S s t 2 ; 2 1 ˆ a b- ± § Hypothesis Testing for the Population Correlation Coefficient, r Sample Correlation Coefficient, r 3 ( 29 ( 29 ( 29 ( 29 yy xx xy i i i i S S S y y x x y y x x r =---- = ∑ ∑ ∑ 2 2 which is used to measure the linear relationship between X and Y . Note: - r lies between – 1 and 1, i.e. 1 1 ≤ ≤- r . - | | r closes to 1 implies a strong linear relationship between X and Y . - | | r closes to 0 implies a weak linear relationship between X and Y . - r is positive. ⇔ X and Y are positively linearly related. - r is negative. ⇔ X and Y are negatively linearly related. If we want to test the relation between X and Y , we can test about the population correlation coefficient, r : Null Hypothesis : = r H Test Statistics: 2 1 2 r n r t-- = ~ t- distribution with 2 d.f.- = n i) when : < r a H , H is rejected when a t t- < , ii) when : r a H , H is rejected when a t t , iii) when : ≠ r a H , H is rejected when 2 / a t t- < or 2 / a t t ....
View Full Document

This note was uploaded on 09/15/2008 for the course ISMT 111 taught by Professor Wanxuhu during the Fall '08 term at HKUST.

Page1 / 12

tutorial08 - 1 ISMT 111 B USINESS S TATISTICS – T UTORIAL...

This preview shows document pages 1 - 4. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online