Statistical skills — AQA A-Level Geography
Test yourself on Statistical skills with AQA A-Level practice questions.
7 days Premium · Then free forever · No card, no charge
Statistical skills explained
Measures of central tendency summarise a dataset by identifying its central or typical value.
Read the full explanation
The mean is the mathematical average, calculated by summing all values and dividing by the total number of observations. It uses all data but is easily skewed by extreme outliers. The median is the middle value when data is ordered sequentially; it is highly effective for skewed distributions, such as river discharge data during a flood event. The mode is the most frequently occurring value, useful for categorical data like dominant pebble lithology on a beach. Selecting the appropriate measure depends on the data type and the presence of anomalies, ensuring geographical analysis remains robust and representative of the true conditions.
Measures of dispersion – range, inter-quartile range and standard deviation.
Measures of dispersion describe how widely data values are spread. The range is maximum minus minimum, so it is quick but strongly affected by extreme values. The interquartile range (IQR) is upper quartile minus lower quartile and describes the middle 50% of ordered values, making it more robust for skewed data or outliers. Standard deviation is the square root of the mean squared deviation from the mean (using the required population or sample convention). It uses every observation and is useful with the mean, but extreme values can influence it. In context, a low standard deviation in pebble size indicates values clustered near the mean, while a high value indicates greater variation; interpretation still depends on units, sampling and the process being studied.
Inferential and relational statistical techniques to include Spearman’s rank correlation and Chi-square test and the application of significance tests.
Inferential and relational statistics test whether a sample pattern provides evidence against a null hypothesis. Spearman’s rank correlation measures the strength and direction of a monotonic association between paired ranked observations, such as river velocity and distance downstream. Chi-square compares observed and expected frequencies for categorical data, either for association in a contingency table or for a stated expected distribution. For a significance test, choose the test and significance level before interpreting results, calculate the statistic, and use the correct critical-value table. Spearman tables are entered with n, the number of valid pairs; Chi-square uses degrees of freedom, (rows − 1)(columns − 1) for association. If the result is significant, reject the null hypothesis; otherwise fail to reject it. Statistical association does not by itself prove causation.
Your focus
- Calculate the mean, median and mode for various geographical datasets.
- Select the most appropriate measure of central tendency based on data distribution and outliers.
- Evaluate the strengths and weaknesses of each measure of central tendency in geographical contexts.
Show all 9 objectives
- Calculate the range, inter-quartile range and standard deviation for geographical data.
- Interpret what different values of dispersion indicate about the physical or human geography context being studied.
- Compare the suitability of the range, inter-quartile range and standard deviation for datasets with and without anomalies.
- Formulate appropriate null and alternative hypotheses for geographical investigations.
- Execute Spearman's rank correlation and Chi-square tests on geographical datasets.
- Interpret calculated statistical values against critical value tables to determine significance.
Statistical skills exam tips
Marking Points
- Identify the correct measure of central tendency for a given geographical dataset.
- Calculate the mean by summing all values and dividing by the sample size.
- Determine the median by ordering data and finding the middle value, or the average of the two middle values if the sample size is even.
- Identify the mode as the most frequent value, recognising that datasets can be bimodal or have no mode.
- Evaluate the limitations of the mean when extreme outliers are present in the dataset.
- Calculate the range by subtracting the minimum value from the maximum value in a dataset.
- Determine the inter-quartile range by finding the difference between the upper quartile and lower quartile.
- Interpret standard deviation as a measure of how spread out numbers are from the mean, where a lower value indicates greater consistency.
- Evaluate the inter-quartile range as a more reliable measure of spread than the range because it ignores extreme outliers at either end of the dataset.
- Apply standard deviation to geographical contexts, such as comparing variability in climate measurements or the spread of sediment sizes.
- State a testable null hypothesis before analysis and identify the paired ranked data or categorical frequencies being tested.
- For Spearman's rank, rank each variable consistently, calculate the coefficient and use n, the number of valid pairs, to select the critical value.
- For Chi-square association, calculate expected frequencies and degrees of freedom as (rows − 1)(columns − 1), then compare the statistic with the correct critical value.
- State the chosen significance level, reject the null hypothesis only when the result is significant, and otherwise report that there is insufficient evidence to reject it.
- Interpret strength, direction and context without treating statistical association as proof of causation.
Examiner Tips
- 💡Always show your working when calculating the mean, as method marks may be awarded even if the final answer is incorrect.
- 💡When asked to justify a chosen measure, reference the presence or absence of extreme outliers in the provided data.
- 💡Check if the question asks for the answer to a specific number of decimal places or significant figures.
- 💡When comparing two datasets, always pair a measure of central tendency with a measure of dispersion to provide a complete analysis.
- 💡If asked to calculate standard deviation, follow the stated population or sample formula, square each deviation, obtain the variance and then take its square root.
- 💡Use the inter-quartile range when your data is represented on a box and whisker plot, as it directly corresponds to the box length.
- 💡Write the null hypothesis and significance level before calculating so the decision rule is clear.
- 💡For Spearman, check tied-rank handling and count the valid pairs; for Chi-square, show observed and expected frequencies and the degrees-of-freedom calculation.
- 💡Use 'significant at p ≤ 0.05' or 'not significant at p ≤ 0.05' as appropriate, then reject or fail to reject the null hypothesis.
- 💡Finish with a contextual interpretation and state that a significant association alone does not demonstrate causation.
Common Mistakes
- Failing to rank data in numerical order before finding the median; always sort the dataset from lowest to highest first.
- Assuming the mean is always the best average; use the median instead if the dataset contains extreme anomalies that would skew the result.
- Confusing the mode with the majority; the mode is simply the most frequent value, even if it does not represent over half the dataset.
- Using the range to compare datasets with extreme outliers; use the inter-quartile range instead to focus on the core data spread.
- Miscalculating quartiles by not ranking the data first; always order the dataset sequentially before finding the percentiles.
- Assuming a high standard deviation means the data is incorrect; it simply indicates a wide variation or spread in the geographical phenomena being measured.
- Writing a vague null hypothesis; correction: name the variables or categories and state that there is no statistically significant association, correlation or difference as appropriate.
- Treating a negative Spearman coefficient as weak; correction: the sign gives direction while the absolute value indicates strength.
- Using n − 2 as Spearman's degrees of freedom; correction: use the AQA critical-value table row for n valid pairs, while degrees of freedom apply to Chi-square.
- Saying the null hypothesis is accepted when a result is not significant; correction: report that you fail to reject it because the sample gives insufficient evidence.
- Using Chi-square on raw continuous measurements; correction: it requires observed frequencies in meaningful categories and suitable expected counts.