In: Statistics and Probability
Cost and Customer Satisfaction
A researcher is interested whether there is a linear correlation between cost (in dollars) of internet service per month and customer satisfaction on a scale of 1 - 10 (with a 1 being extremely dissatisfied and 10 extremely satisfied). The researcher only includes internet service with similar download speed. The sample he collected is below.
Cost (x) | 15 | 18 | 17 | 11 | 9 | 11 | 12 | 19 | 18 | 22 |
Satisfaction (y) | 6 | 8 | 10 | 4 | 9 | 6 | 3 | 5 | 2 | 10 |
Find the best predicted value for the level of
Customer Satisfaction given that the Cost is 13. Use a significance
level of 0.05.
Round your answer to 1 decimal place, if needed.
ΣX | ΣY | Σ(x-x̅)² | Σ(y-ȳ)² | Σ(x-x̅)(y-ȳ) | |
total sum | 152.00 | 63.00 | 163.60 | 74.10 | 24.40 |
mean | 15.20 | 6.30 | SSxx | SSyy | SSxy |
Sample size, n = 10
here, x̅ = Σx / n= 15.200
ȳ = Σy/n = 6.300
SSxx = Σ(x-x̅)² = 163.6000
SSxy= Σ(x-x̅)(y-ȳ) = 24.4
estimated slope , ß1 = SSxy/SSxx =
24.4/163.6= 0.1491
intercept,ß0 = y̅-ß1* x̄ = 6.3- (0.1491
)*15.2= 4.0330
Regression line is, Ŷ= 4.033 +
( 0.149 )*x
best Predicted Y at X= 13
is
Ŷ= 4.03301 +
0.14914 *13= 5.972
≈ 6.0