The following table gives the number of mistakes made by 16 data entry clerks who enter medical data from case report forms. The column Entered indicates the number of values entered, and the column Errors gives the number of coding errors that were detected among these. Entered | Errors 4434 | 35 4841 | 42 6280 | 15 1958 | 28 7749 | 36 2829 | 42 4239 | 18 3303 | 64 4706 | 34 3770 | 40 3363 | 36 1740 | 23 3404 | 27 1640 | 26 3803 | 56 1529 | 20 (i) Make a scatter-plot of these data. Which do you choose for the response and which for the explanatory variables? Describe any patterns. (ii) Does the scatter-plot indicate any relation between the two variables? (iii) One analyst concluded, "The clerks who enter more values make more mistakes. Evidently they become tired as they enter more values." Is this an appropriate conclusion?
Choose the variables, describe how to plot and read the scatter, measure the association, and judge the analyst's causal claim.
Common trap: Students often accept the analyst's claim. Compute r (about 0.07), note the error rate actually falls as volume rises, and point out that fatigue was never measured.
SyllabusOld-pattern GS Paper II (2006–12): Statistical analysis, graphs and diagrams — scatter diagrams and correlation
Model answer
9 marks · 90-word limitHere the number of values Entered is the explanatory variable (x-axis) and the number of Errors is the response (y-axis), because errors may depend on workload, not the reverse.
Given
- Data: Sixteen clerks; Entered ranges from 1,529 to 7,749 (total 59,588, mean 3,724.25), Errors from 15 to 64 (total 542, mean 33.875).
Method / formula
- Scatter: Plot sixteen points, x from 1,000 to 8,000 in steps of 1,000 and y from 0 to 70 in steps of 10.
- Correlation: Karl Pearson's r = [nΣxy − ΣxΣy] / √{[nΣx² − (Σx)²][nΣy² − (Σy)²]}, with Σxy = 2,043,000, Σx² = 265,071,144 and Σy² = 21,000.
Working
| Measure | Value |
|---|---|
| Pearson r | about 0.07 |
| r squared | about 0.005 (0.5 per cent of variation explained) |
| Least-squares line | Errors ≈ 31.8 + 0.00057 × Entered |
| Errors per 1,000 values, clerks under 2,000 entries | about 13 to 16 |
| Errors per 1,000 values, clerks over 6,000 entries | about 2.4 to 4.6 |
Result
- (i) Pattern: Points form a shapeless cloud with no trend; 3,303 entries with 64 errors and 6,280 with only 15 errors stand out as unusual.
- (ii) Relation: With r of about 0.07, there is practically no linear relationship between values entered and errors made.
- (iii) Conclusion: Not appropriate: the data show no association, the error rate actually falls for high-volume clerks, and fatigue was never measured, so no causal claim follows.
Read more — sources
- BookS.C. Gupta and V.K. Kapoor, Fundamentals of Mathematical Statistics (Sultan Chand) — correlation and regression
- NCERTNCERT Class XI, Statistics for Economics — Ch. 8 Correlation
More from GS Paper II (old) 2012
Model answer by the EdiShorts editorial team. It is one well-argued treatment of the question, written to the official word limit — not an official UPSC answer key.