All subjects · Geography · WGE03 Contested Planet

Unit 3 skills: scattergraphs, Spearman's rank and chi-squared

Edexcel IAL Geography · WGE03 Contested Planet

Unit 3 expects you to interpret correlation using scattergraphs and Spearman's rank (including calculation) and to use chi-squared to test for significant differences between countries in development or superpower status (skills 15 and 16). The figure questions in Section A and the 5-mark questions in Sections B and C also reward precise use of data.

Practise this in the app: instant marking and a fix for every mistake. Free, no email needed to start.

Practise this lesson

Scattergraphs and correlation

rs = 1 - (6 x sum of d squared) / (n(n squared - 1))

rs runs from +1 (perfect positive) to -1 (perfect negative). Compare the value with the critical value for n at a chosen significance level (usually 0.05). If rs is greater than or equal to the critical value, the correlation is significant: there is less than a 5% probability it happened by chance. For n = 10 the two-tailed critical values are 0.648 (0.05) and 0.794 (0.01) (Pearson tables).

Worked data: income and life expectancy

UNDP 2023 data for ten countries (rank 1 = highest):

CountryGNI per capita (PPP $)RankLife expectancyRankd squared
Iceland69,117382.730
Norway112,710183.321
United Kingdom54,372481.340
United States73,650279.359
South Korea49,726584.3116
China22,029678.060
Brazil18,011775.870
India9,047972.090
South Africa13,694866.1104
Bangladesh8,4981074.784

UNDP Human Development Report 2025. Sum of d squared = 34.

Chi-squared

The chi-squared test compares observed counts with the counts expected if there were no difference between categories:

chi-squared = sum of (O - E) squared / E

Degrees of freedom = number of categories - 1 (for a single row). If the result is greater than or equal to the critical value, the difference is significant. Critical values (Pearson tables): df 1: 3.84 (0.05); df 2: 5.99 (0.05) and 9.21 (0.01).

Example: operational nuclear reactors at the end of 2024 were 94 in the USA, 57 in China and 57 in France (IAEA). If the three powers had equal numbers, each would have (94 + 57 + 57) / 3 = 69.33. Use counts only (not percentages or money), and expected values should be at least 5.

Worked example

Calculate Spearman's rank for GNI per capita and life expectancy in the table and test its significance at the 0.05 level.

  1. Sum of d squared = 34, n = 10.
  2. rs = 1 - (6 x 34) / (10 x (10 squared - 1)) = 1 - 204 / 990 = 0.794.
  3. Critical value for n = 10 at 0.05 (two-tailed) is 0.648. 0.794 > 0.648, so the positive correlation is significant at 0.05. (Unrounded, rs = 0.79394, just below the 0.01 value of 0.794, so it is not significant at 0.01.)
  4. Interpretation: richer countries tend to have longer life expectancy, but anomalies exist (South Korea has the longest life expectancy without the highest income; South Africa's is low for its income).

Answer: rs = 0.794: a strong, significant positive correlation at the 0.05 level.

Key idea: Scattergraphs: describe direction, strength, anomalies, best fit; correlation is not causation. Spearman: rank, d squared, rs = 1 - 6 sum d squared / n(n squared - 1); compare with the critical value (n = 10: 0.648 at 0.05). Chi-squared: sum (O - E) squared / E with df = categories - 1; compare with the critical value (df 2: 5.99 at 0.05). Use counts, not percentages.

Check yourself

  1. Using the table of ten countries, calculate Spearman's rank correlation coefficient between GNI per capita and life expectancy. Give your answer to 3 decimal places.
    Show answer

    0.794. rs = 1 - 6 x 34 / 990 = 0.794.

  2. What is the sum of d squared in the table?
    Show answer

    34. Adding the last column gives 34.

  3. In the reactor example, what is the expected number of reactors per country if the three were equal? Give your answer to 2 decimal places.
    Show answer

    69.33. 208 / 3 = 69.33.

  4. Calculate chi-squared for the reactor counts (USA 94, China 57, France 57). Give your answer to 2 decimal places.
    Show answer

    13.16. (94 - 69.33) squared / 69.33 + 2 x (57 - 69.33) squared / 69.33 = 13.16.

  5. With 2 degrees of freedom, the critical value at 0.05 is 5.99. The reactor chi-squared is 13.16. What can you conclude?
    • The difference between the countries is significant at the 0.05 level
    • There is no significant difference
    • The test cannot be used with counts
    • The USA has fewer reactors than expected
    Show answer

    The difference between the countries is significant at the 0.05 level. 13.16 > 5.99, so the difference is unlikely to be due to chance.

  6. A significant positive correlation between GDP per capita and CO2 emissions per capita shows that
    • higher GDP definitely causes higher emissions in every country
    • there are no anomalies
    • the two variables tend to rise together, but it does not prove one causes the other
    • the relationship is negative
    Show answer

    the two variables tend to rise together, but it does not prove one causes the other. Correlation is not causation.

  7. Why should chi-squared not be calculated using percentages or money values?
    • The test is designed for frequencies (counts); percentages and money distort the result
    • Percentages are always too small
    • It only works with negative numbers
    • Money values have no degrees of freedom
    Show answer

    The test is designed for frequencies (counts); percentages and money distort the result. Chi-squared uses observed frequencies.

Practise with instant feedback

Sources

Last updated 8 November 2026. Spotted a mistake? Tell us.