Statistics in the Punjab Police SI paper is narrow and very predictable. All 19 questions asked since 2016 are about three numbers — the mean, the median and the mode — and about one relationship between them. That relationship, Mode = 3 × Median − 2 × Mean, carries six questions on its own. Five more are plain mode-spotting, four need a median found by counting, and four read a frequency table. Nothing here requires standard deviation, variance or any formula beyond those on this page.
Three averages, three jobs
Know what each one measures before you compute it.
Find the mode of the following data. 15, 17, 26, 28, 16, 10, 28, 16, 28 and 16 (SI 15 Jul 2023)
Show answer
Count carefully: in 15, 17, 26, 28, 16, 10, 28, 16, 28, 16 the value 16 appears three times and 28 also appears three times. The data as printed is genuinely bimodal, which a single-answer MCQ cannot express — the paper's key is 16, so mark that, but note that 28 is equally frequent and that this is a defect in the question rather than in your counting. If a mode question ever gives you two values tied at the top, recount once, then take the one that appears among the options.
The empirical relation
Six questions from a single line.
For a moderately skewed distribution the three averages sit roughly in line, with the median one-third of the way from the mean to the mode. Rearranged, the same line gives you the other two: Median = (Mode + 2 × Mean) ÷ 3 · Mean = (3 × Median − Mode) ÷ 2 You only ever need to remember the first form; the other two follow in one step of algebra. Every SI question on this topic gives you two of the three and asks for the third.
Reading a frequency table
Four questions, and the same two traps every time.
First, giving the frequency instead of the value. The mode of a frequency table is the observation with the highest frequency, not the frequency itself. If the mark 15 was scored by 16 students and that is the biggest count, the mode is 15, not 16. Second, averaging the row of values and ignoring the frequencies entirely. That only happens to work when the frequencies are symmetric, which examiners deliberately arrange in one or two questions so that the lazy method appears to succeed.
- Giving the frequency when the question asked for the observation. The mode of a table is the value in the top row, not the count below it.
- Taking the median without sorting first. The data in the stem is almost never in order.
- Stopping at 2 × Mean or at x instead of finishing the last step. Several SI options are exactly the half-done value.
- Mixing up the empirical relation. It is Mode = 3 Median − 2 Mean, with the 3 on the median — never 3 Mean − 2 Median.
- Averaging a frequency table's value row and ignoring the frequencies. It works only when the frequencies are symmetric.
- Mode = 3 × Median − 2 × Mean
- Median = (Mode + 2 × Mean) ÷ 3
- Mean = (3 × Median − Mode) ÷ 2
- Odd n: median is the (n + 1)/2-th value
- Even n: average the two middle values
- Grouped mean = Σfx ÷ Σf, x = class mid-point
- Table median: use the running total
- The mode is a value, never a frequency
Mo-Me-Mn, three-two. Write the three averages in that order — Mode, Median, Mean — and the formula reads straight off it: the first equals three times the second minus two times the third. The coefficients 3 and 2 go with the two words you did not start from.