7SSGN110 Chap.3 Describing Centre, Spread and Shape
Describing Centre, Spread and Shape
Describe the distribution before modelling it
A numerical summary compresses many observations, so it should be chosen only after inspecting what is being compressed. The mean uses every value and responds strongly to extremes. The median depends on rank and is more resistant.
A defensible description pairs the centre with a compatible measure of spread and keeps the environmental unit beside every reported value.
Match summaries to shape
Mean and standard deviation are informative for a roughly symmetric distribution whose tails and unusual values have been inspected. Median and interquartile range usually communicate a skewed central pattern more clearly.
That choice does not make extreme observations irrelevant. High flood losses, concentrations or counts can be the events of greatest scientific interest, so show the distribution and explain whether each unusual record was verified, retained or excluded under a rule.
Calculate spread with its unit intact
Sample variance adds squared deviations from the sample mean and divides by n minus one.
Its unit is squared, which makes it useful inside formulas but difficult to interpret beside the original measurements. Standard deviation takes the square root and returns to the observation unit. The interquartile range subtracts the first quartile from the third and describes the middle half of ordered values.
Stating units exposes many arithmetic and transcription mistakes.
Use comparison rather than automatic output
A five-number summary, histogram and boxplot reveal features that a mean alone cannot. Compare group-specific summaries when sites or habitats may arise from different processes; a pooled result can conceal distinct centres and spreads. Report the sample size behind each set of values.
A sensitivity calculation can show how strongly the mean depends on a verified extreme, but it must not become a pretext for deletion.
Treat models and transformations as choices
A normal curve is not a default picture for every variable. Physical bounds, counts, detection limits and mixed processes can create asymmetry or clusters. Standardising values changes centre and scale but preserves shape and dependence.
A logarithm or square root may stabilise variation when scientifically justified, yet the transformed scale changes interpretation. Retain original-unit summaries and explain any back-transformation so the reader can reconnect analysis with the environmental measurement.
Write a complete descriptive result
A results sentence should state the variable, sampled group or period, centre, spread, unit and sample size.
Add one distribution feature that explains the selected summaries, such as right skew or a verified extreme. When groups are compared, present their summaries in the same order and with compatible precision. Do not label an observation an outlier without naming the rule or context used.
A compact table and one diagnostic plot can then support the prose without repeating every raw value.
For revision, calculate both resistant and non-resistant summaries for a short ordered sample, then alter one extreme value. Track which quantities move and which remain stable.
Finish by writing one sentence that chooses the lead summary from the observed shape while preserving the environmental importance of the extreme.
What this chapter covers
- 01
Mean
- 02
Median
- 03
Mode
- 04
Variance
- 05
Standard deviation
- 06
Interquartile range
- 07
Skewness
Worked example · free
Describe a skewed turbidity sample
- 2Calculate the mean as 42 divided by 7.
- 3Identify the median and both quartiles from ranked values.
- 3Explain why median and IQR lead for the skewed sample.
Key terms
- Median
- The Median is the middle ranked observation or the midpoint of the middle pair.
- Interquartile Range
- The Interquartile Range is the distance from the first to the third quartile.
- Standard Deviation
- Standard Deviation measures spread around the mean in the original measurement unit.
Describing Centre, Spread and Shape FAQ
When should the median lead?
Use the median when skew, extreme observations or an ordinal scale make the mean a poor description of a typical value. Pair it with the interquartile range and show the distribution so unusual but valid events are not silently hidden.
Why does variance have squared units?
Variance averages squared deviations from the mean, so its unit is also squared. Taking the square root produces standard deviation in the original unit, making it directly comparable with the observations and their mean.
Does standardising make data normal?
No. A z-score subtracts the mean and divides by standard deviation, changing location and scale but preserving distribution shape. Skew, clusters and dependence remain and must be addressed separately rather than concealed by standardisation.
Assessment move
Practise on short lists until mean, median, quartiles, sample variance and standard deviation can be checked by hand. Before using software output, predict how one extreme value would move each statistic and which unit the result should carry.
Working through Describing Centre, Spread and Shape in 7SSGN110? Sia is AskSia’s AI Statistics tutor — ask any 7SSGN110 Describing Centre, Spread and Shape question and get a clear, step-by-step explanation grounded in how 7SSGN110 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.