Introduction to Quantitative Skills in Geography

In Geography, we don't just describe what we see; we use numbers to prove it. Whether you are studying the size of pebbles on a beach for Component 01 or analyzing social inequality in Component 02, you need tools to summarize your data and test if your findings are actually meaningful or just a coincidence. This chapter covers the essential statistical tools you need for your A Level exams and your Investigative Geography (NEA).

1. Descriptive Statistics: Summarizing Data

Descriptive statistics help us describe the main features of a dataset. They are broken down into measures of central tendency (the "middle") and dispersion (the "spread").

Measures of Central Tendency

  • Mean: The arithmetic average. You add all the values together and divide by the number of values (\(n\)). It is useful because it uses all the data, but it can be "pulled" by extreme outliers.
  • Median: The middle value when the data is placed in order. It is great for skewed data because it isn't affected by extreme outliers.
  • Mode: The most frequently occurring value. This is useful for categorical data (e.g., the most common type of sea defense).

Measures of Dispersion

  • Range: The difference between the highest and lowest values. It is simple to calculate but can be misleading if there is one unusually high or low result.
  • Interquartile Range (IQR): This measures the spread of the middle 50% of the data. It ignores the extremes at either end, making it a more reliable measure of spread than the range.
  • Standard Deviation (\(\sigma\)): This shows how much the data "deviates" from the mean. A low standard deviation means the data points are clustered closely around the mean; a high standard deviation means the data is spread out.

Quick Review: If you are comparing the wealth of two neighborhoods, the median is often better than the mean because one billionaire living in a poor area would make the "average" (mean) wealth look much higher than it really is!

2. Sampling and Measurement

In fieldwork, you rarely have the time to measure everything. Instead, you take a sample. However, no sample is perfect.

  • Measurement Errors: These happen due to human mistakes (like misreading a tape measure) or faulty equipment (like a digital thermometer with low batteries).
  • Sampling Bias: This occurs if your sample doesn't accurately represent the whole "population" you are studying.

For more on how to choose your sample, see the chapter on Fieldwork and the Independent Investigation.

3. Correlation and Lines of Best Fit

When we want to see if two things are related (e.g., does distance from the CBD affect house prices?), we look for correlation.

  • Positive Correlation: As one variable increases, the other increases.
  • Negative Correlation: As one variable increases, the other decreases.
  • Line of Best Fit: A line drawn through the center of the points on a scatter graph to show the general trend.

Important Note: Correlation does not always mean causation. Just because two things happen at the same time doesn't mean one caused the other!

4. Significance Tests: Proving Your Patterns

A significance test helps you decide if a pattern in your data is statistically "significant" or if it could have happened by pure chance. To do this, we always start with a Null Hypothesis (\(H_0\)), which states that there is no relationship or difference.

Spearman’s Rank Correlation Coefficient (\(r_s\))

Used to test the strength and direction of a relationship between two sets of ranked data.
- The result (\(r_s\)) will always be between \(-1\) (perfect negative) and \(+1\) (perfect positive).
- A result of \(0\) means no relationship at all.

Chi-squared (\(\chi^2\))

Used to see if there is a significant difference between the observed frequencies (what you counted) and the expected frequencies (what you would expect if everything was random).
- Example: Are certain types of shops more common in one part of the city than another?

Mann-Whitney U Test

Used to test for a significant difference between the medians of two separate sets of data.
- It is "non-parametric," meaning it's great for data that isn't perfectly bell-shaped.
- Example: Comparing the roundness of pebbles on two different beaches.

Student’s T-test

Used to test for a significant difference between the means (averages) of two sets of data.
- Unlike Mann-Whitney U, this requires the data to be "normally distributed" (bell-shaped).
- Example: Comparing the average infiltration rates of two different soil types.

Don't worry if this seems tricky: In the exam, you are often asked to interpret the results of these tests (using a significance table) rather than calculating them from scratch!

5. Understanding "Significance" Levels

In Geography, we usually look for a 95% confidence level (written as \(p = 0.05\)).
- If your result is significant at this level, it means there is only a 5% chance that the pattern happened by accident.
- If your result is higher than the "critical value" in a statistical table, you can reject the Null Hypothesis and say your findings are significant.

Key Takeaways
  • Descriptive stats (mean, SD, IQR) summarize your data.
  • Significance tests (Spearman's, Chi-squared, etc.) prove if your data matters.
  • Always check for measurement errors and sampling bias when evaluating your results.
  • Use lines of best fit to visualize trends before running complex tests.