41297 Chap.6 Dispersion, Z-Scores and Distribution Shape
Dispersion, Z-Scores and Distribution Shape
Why spread matters as much as the average
Two sets of figures can share a mean and still behave very differently, so the second lecture follows the measures of centre with measures of variation: range, variance, standard deviation and the coefficient of variation, each with a population and a sample form.
From absolute to relative spread
This chapter computes variance by squaring deviations, turns it back into the data's units as a standard deviation, and then divides by the mean to compare variability across products, shares or variables measured in different units.
Position and shape
It closes with z-scores, which locate a single value in standard deviations from the mean, and with the coefficient of skewness, whose bands tell you whether a distribution is roughly symmetric or leans heavily to one side, and therefore whether the mean is a fair summary of the centre.
What this chapter covers
- 01
Range and its sensitivity to extreme values
- 02
Population and sample variance
- 03
Standard deviation in the units of the data
- 04
Coefficient of variation for relative spread
- 05
Standardised z-scores
- 06
Skewness bands and the position of the mean
Worked example · free
Compute the sample variance and standard deviation of weekly returns
- 1Mean: the total is 70, and 70 ÷ 5 gives an average of 14 returns a week.
- 1Deviations from 14 are −2, +1, −5, +4 and +2, which add to zero as they must.
- 1Squared deviations are 4, 1, 25, 16 and 4, giving a sum of squares of 50.
- 1Sample variance: divide by n − 1 = 4, so s² = 50 ÷ 4 = 12.5.
- 1Standard deviation: s = √12.5 ≈ 3.54 returns a week.
Key terms
- Range
- The largest value minus the smallest, the quickest but least robust measure of spread.
- Variance
- The average squared distance of the values from their mean, divided by N for a population or n minus 1 for a sample.
- Standard deviation
- The positive square root of the variance, which expresses spread in the original units.
- Coefficient of skewness
- A number summarising how far a distribution leans to the right or left of symmetry.
Dispersion, Z-Scores and Distribution Shape FAQ
Why does the sample variance divide by n minus 1?
A sample's own mean sits closer to its values than the unknown population mean would, so the squared deviations come out slightly too small. Dividing by one fewer than the number of observations corrects for that. When your data cover every member of the group you care about, use the population version and divide by N instead.
What does a coefficient of variation show that a standard deviation cannot?
It shows spread relative to the size of the data. A standard deviation of 25 cups may be modest for a product selling hundreds a week and enormous for one selling thirty. Expressing it as a percentage of the mean lets you rank consistency across products, prices or variables in different units.
How should I read a negative z-score?
A minus sign means the value sits below the mean, and the size tells you by how many standard deviations. A score of minus 0.8 is moderately below average; minus 2.5 would be unusually low. Z-scores carry no units, so you can compare a sales figure with a delivery time on the same scale.
How can I tell whether my data are skewed?
Compute the coefficient of skewness with Excel's SKEW function or the descriptive statistics output in Minitab or SPSS. The lecture treats values between minus 0.5 and 0.5 as relatively symmetric and values beyond 1 in either direction as highly skewed. Compare the mean and median too: a long right tail drags the mean upward.
Is the range ever a sensible measure to report?
Yes, as a quick first look or when you need the full span, for instance the lowest and highest price in a market. It should not stand alone, though, because one unusual observation can stretch it a long way while saying nothing about how the bulk of the data is spread.
Assessment move
Build a small spreadsheet for each method in this chapter. Type five to ten values, compute the mean, the deviations, their squares and both variance versions by hand-written formulas, then check against the built-in functions until the two agree. Add a column of z-scores and confirm that their mean is zero.
For your assessment dataset, report each quantitative variable's mean, standard deviation and coefficient of variation together, and add the skewness coefficient so you can say whether the mean or the median is the fairer centre. Write one sentence per variable on what the spread means for the business problem.
Working through Dispersion, Z-Scores and Distribution Shape in 41297? Sia is AskSia’s AI Statistics tutor — ask any 41297 Dispersion, Z-Scores and Distribution Shape question and get a clear, step-by-step explanation grounded in how 41297 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.