In: Statistics and Probability
Graded problem #2. The table below gives the cost per unit of a manufactured piece as a function of the number of units produced. Take the number of units produced as the explanatory variable, x, and the cost per unit as the response variable y.
Units 10 20 50 100 150 200
Cost 140 155 153 150 180 160
a. Predict the cost per unit when 80 units are produced
b. Find a 90% confidence interval for the slope b1
c. Find a 90% confidence interval for the predicted value you found in answer b.
X | Y | X * Y | X2 | Ŷ | Sxx =Σ (Xi - X̅ )2 | Syy = Σ( Yi - Y̅ )2 | Sxy = Σ (Xi - X̅ ) * (Yi - Y̅) | |
10 | 140 | 1400 | 100.0000 | 147.339 | 6136.106 | 266.777 | 1279.441 | |
20 | 155 | 3100 | 400.0000 | 148.488 | 4669.440 | 1.778 | 91.109 | |
50 | 153 | 7650 | 2500 | 151.932 | 1469.442 | 11.111 | 127.776 | |
100 | 150 | 15000 | 10000 | 157.673 | 136.112 | 40.111 | -73.889 | |
150 | 180 | 27000 | 22500 | 163.414 | 3802.782 | 560.113 | 1459.447 | |
200 | 160 | 32000 | 40000 | 169.155 | 12469.452 | 13.445 | 409.448 | |
Total | 530 | 938 | 86150 | 75500 | 672.000 | 28683.333 | 893.333 | 3293.333 |
X̅ = Σ (Xi / n ) = 530/6 = 88.3333
Y̅ = Σ (Yi / n ) = 938/6 = 156.3333
Equation of regression line is Ŷ = a + bX
b = ( n Σ(XY) - (ΣX* ΣY) ) / ( n Σ X2 - (ΣX)2
)
b = ( 6 * 86150 - 530 * 938 ) / ( 6 * 75500 - ( 530
)2)
b = 0.1148
a =( ΣY - ( b * ΣX ) ) / n
a =( 938 - ( 0.1148 * 530 ) ) / 6
a = 146.1912
Equation of regression line becomes Ŷ = 146.1912 + 0.1148
X
Part a)
= 146.1912 + 0.1148 X
= 155.3752
Estimated Error Variance (σ̂2) =
S2 = ( 893.3333 - 0.1148 * 3293.3333 ) / 6 - 2
S2 = 128.814659
S = 11.3497
Part b)
Confidence Interval
= 2.132 ( Critical value from t table )
90% confidence interval is -0.0281 <
< 0.2577
Part c)
Confidence Interval of
= 146.1912 + 0.1148 X
= 155.3752
= 2.132 ( From t table )
X̅ = (Xi / n ) = 530/6 = 88.3333
= 155.3752
90% confidence interval is ( 145.4252 <
< 165.3252 )