In: Statistics and Probability
| Heater | Wattage | Area | ||
| 1 | 1,500 | 131 | ||
| 2 | 1,500 | 142 | ||
| 3 | 750 | 111 | ||
| 4 | 1,500 | 264 | ||
| 5 | 1,250 | 237 | ||
| 6 | 1,750 | 267 | ||
| 7 | 1,500 | 61 | ||
| 8 | 2,000 | 135 | ||
| 9 | 2,000 | 263 | ||
| 10 | 2,000 | 263 | ||
| 11 | 750 | 125 | ||
| 12 | 1,500 | 76 | ||
| 13 | 2,000 | 77 | ||
| 14 | 750 | 170 | ||
| 15 | 1,000 | 165 | ||
| 16 | 1,000 | 152 | ||
| 17 | 1,000 | 238 | ||
| 18 | 1,750 | 209 | ||
| 19 | 1,250 | 167 | ||
| 20 | 1,750 | 75 | ||
Compute the correlation between the wattage and heating area. Is there a direct or an indirect relationship? (Round your answer to 4 decimal places.)
Conduct a test of hypothesis to determine if it is reasonable that the coefficient is greater than zero. Use the 0.010 significance level. (Round intermediate calculations and final answer to 3 decimal places.)
Develop the regression equation for effective heating based on wattage. (Negative value should be indicated by a minus sign. Round your answers to 3 decimal places.)
Which heater looks like the “best buy” based on the size of the residual? (Round residual value to 3 decimal places.)
1)

| correlation r='Sxy/(√Sxx*Syy) = | 0.1256 | |
direct relationship
b)
| null hypothesis: Ho: ρ'= | 0 | |
| Alternate Hypothesis: Ha: ρ> | 0 | |
| Decision rule: reject Ho if test statistic t>2.552 | 
| test stat t= | r*(√(n-2)/(1-r2))= | 0.537 | 
| P value = | 0.2988 | (from excel:tdist(0.5373,18,1) | 
Fail to reject the null hypothesis
c)
| sample size n= | 20 | |
| x̅ = | 1425.0000 | |
| y̅ = | 166.4000 | |
| Sxx= | Σ(Xi-X̅)2= | 3637500.000 | 
| Sxy = | Σ(Xi-X̅)(Yi-Y̅)= | 73350.000 | 
| slope= β̂1 = | Sxy/Sxx= | 0.0202 | 
| intercept= β̂0 = | y̅-β1x̅= | 137.6649 | 
| Least square line equation: ŷ =137.665+0.020*x | ||
heater 4 is appear to be best buy since provide highest positive residual of 96.088