In: Statistics and Probability
We suspect that Method 1 for detecting impurities in iron ingots yields higher results than Method 2 for finding such impurities. We want to give a 95% confidence interval for the difference in the average number of impurities detected by the two methods. In order to obtain the confidence interval we measured the impurities of 3 ingots, employing both methods for each ingot. The results are presented in the table below. Assume the difference scores are normally distributed.
We let μ1 denote the population average number of defects for Method 1 and μ2 denotes the population average number of defects for Method 2. We wish to develop a confidence interval for μd = μ1-μ2. Give the margin of error for the 95% confidence interval for μd.
Ingot 1 | Ingot 2 | Ingot 3 | |
Method 1 | 16 | 32 | 12 |
Method 2 | 12 | 26 | 10 |