Revision notes

Statistics: Foundation

The key points, rules to remember, tables and diagrams from all 9 lessons, on one page for quick revision. To keep a copy, use your browser's Print and choose "Save as PDF".

Lesson 1

Mean, median and mode of raw data

Mean x̄ = Σx/n (or Σfx/Σf for a frequency table). Median = middle value of the ordered data (average the two middle values when n is even). Mode = most frequent value.

Number of books read by 20 students
Value x23456
Frequency f35741

Key takeaways. Add and divide for the mean, using Σfx/Σf for a frequency table. Arrange the data before finding the median and average the two middle values when n is even. The mode is the most frequent value, and there may be more than one.

Read the full lesson

Lesson 2

Mean, median and mode of grouped data

Mean = A + h(Σfu/Σf) with u = (x − A)/h. Median = l + ((N/2 − cf)/f)h. Mode = l + ((f₁ − f₀)/(2f₁ − f₀ − f₂))h.

Marks of 40 students
Class0-1010-2020-3030-4040-50
Frequency5101564
Class mark x515253545
u = (x − 25)/10−2−1012

Key takeaways. Grouped data are handled through class marks. The step-deviation method gives the same mean with smaller numbers. For the median, find N/2 and use the cumulative frequency of the class before the median class; for the mode, use the frequencies on both sides of the modal class.

Read the full lesson

Lesson 3

Mean deviation about the mean and the median

Mean deviation about a: MD(a) = Σ|x − a|/n, or Σf|x − a|/Σf for a frequency table. Use a = x̄ for the mean deviation about the mean and a = M for the mean deviation about the median.

x25681012
f2810785

Key takeaways. Deviations from the mean add to zero, so the mean deviation uses distances. Always say about what the deviation is measured. Find the centre first, then the distances, then their average, multiplying by frequencies when there are any.

Read the full lesson

Lesson 4

Variance and standard deviation

σ² = Σ(x − x̄)²/n = Σx²/n − x̄² (divide by n). For grouped data with u = (x − A)/h: σ² = h²[Σfu²/N − (Σfu/N)²], and σ = √σ².

u−2−1012Total
f510156440
fu−10−10068−6
fu²2010061652
  • Variance is never negative, and it is 0 only when every observation is the same.
  • Variance is in squared units (marks²); the standard deviation is back in the original units.
  • In the step-deviation formula, h² multiplies the whole bracket; forgetting it is the usual error.
  • Do not round the mean before squaring it in the shortcut formula.

Key takeaways. Variance is the mean of squared deviations, dividing by n. The shortcut Σx²/n − x̄² saves work, and the step-deviation formula handles grouped data with small numbers. The standard deviation is the square root of the variance.

Read the full lesson

Lesson 5

Coefficient of variation and the effect of changing the data

CV = (σ/x̄) × 100. Smaller CV means more consistent. Adding k to every value: mean + k, spread unchanged. Multiplying every value by k: mean × k, standard deviation and mean deviation × |k|, variance × k².

Effect of changing every observation
Change to every valueMeanVarianceStandard deviation
add kx̄ + kσ² (same)σ (same)
multiply by kk x̄k²σ²|k|σ

Key takeaways. Use the coefficient of variation, not the standard deviation, to compare consistency when the means differ; the smaller CV wins. A shift changes the averages but not the spread. Scaling by k scales the averages by k, the standard deviation by |k| and the variance by k².

Read the full lesson

Lesson 6

Range, quartiles and the quartile deviation

Range = largest − smallest. Q₁ = median of the lower half, Q₃ = median of the upper half (leave out the median if n is odd). Interquartile range = Q₃ − Q₁. Quartile deviation = (Q₃ − Q₁)/2.

Key takeaways: the range uses only the extremes and is easily distorted. The quartiles describe the middle half of the data; the quartile deviation is (Q₃ − Q₁)/2. Both are unchanged by a shift and scaled by |k| under multiplication.

Read the full lesson

Lesson 7

Mean, median and mode together: skewness and the empirical relation

ShapeDescriptionOrder of the averages
Symmetricbalanced on both sidesmean = median = mode
Positively skewed (right tail)a few very large valuesmean > median > mode
Negatively skewed (left tail)a few very small valuesmean < median < mode

Mode ≈ 3 median − 2 mean. In right-skewed data mean > median > mode; in left-skewed data the order reverses. Pearson's measure of skewness: (mean − mode)/σ — positive for a right tail, negative for a left tail, zero for symmetric data.

Key takeaways: the order of mean, median and mode shows the skew. The empirical relation mode ≈ 3 median − 2 mean links them. One extreme value moves the mean much more than the median or mode.

Read the full lesson

Lesson 8

Cumulative frequency tables: less-than, more-than and reading the median

ClassFrequencyLess than (upper limit)More than or equal (lower limit)
0-104below 10: 4from 0: 40
10-208below 20: 12from 10: 36
20-3012below 30: 24from 20: 28
30-4010below 40: 34from 30: 16
40-506below 50: 40from 40: 6

Less-than count at an upper limit = running total of the frequencies. More-than count at a lower limit = N − (less-than count at that limit). Within a class, assume the observations are spread evenly to estimate counts at values inside it.

Key takeaways: less-than tables use upper limits and running totals; more-than tables use lower limits and N minus the less-than count. The median and quartiles are read from the first class whose cumulative frequency reaches N/2, N/4 or 3N/4.

Read the full lesson

Lesson 9

Building frequency tables: tallies, inclusive classes and class marks

Inclusive classes (10-19, 20-29, ...) become continuous by moving each lower limit down by half the gap and each upper limit up by half the gap: 9.5-19.5, 19.5-29.5, ... The class mark is the same either way: (10 + 19)/2 = (9.5 + 19.5)/2 = 14.5.

Key takeaways: tally into equal-width classes and check the total. Correct inclusive classes by half the gap before using median or quartile formulas. Class marks are the same before and after the correction.

Read the full lesson

Privacy · Terms · Refunds · Delivery