How many mistakes get made during data entry? The following table gives the number of mistakes made by 15 data entry clerks who enter medical data from case report forms. These forms are submitted by doctors who participate in studies of the performance of drugs for treating various illnesses. The column Entered indicates the number of values entered, and the column Errors gives the number of coding errors that were detected among these.
EnteredErrors
4,434......35
4,841......42
6,280......15
1,958......28
7,749......36
2,829......42
4,239......18
3,303......54
5,706......34
3,770......40
3,363......36
1,740......23
3,404......27
1,640......26
3,803......56
1,529......20
(a) Make a scatterplot of these data. Which did you choose for the response and which for the explanatory variable? Describe any patterns.
(b) Find the correlation for these data.
(c) Suppose we were to record the counts in the table in hundreds, so 4,434 became 44.34. How would the correlation change? Why?
(d) Write a sentence or two that interprets the value of this correlation. Use language that would be understood by someone familiar with data entry rather than correlations.
(e) One analyst concluded, “It is clear from this correlation that clerks who enter more values make more mistakes. Evidently they become tired as they enter more values.” Explain why this explanation is not an appropriate conclusion.