Descriptive statistics and statistical tests

Welcome to the World of Marine Data!

In Marine Science, we don't just observe the ocean; we measure it. Whether you are counting mangrove seedlings or measuring the salinity of a rock pool, you will end up with a pile of numbers. Descriptive statistics help us summarize that pile into a clear picture, and statistical tests help us decide if our results are actually meaningful or just a happy accident of "luck."

Don’t worry if math isn’t your favorite subject! For Papers 2 and 4, you don’t need to be a mathematician. You just need to know which tool to use and what the numbers are telling you. Let's dive in!

1. The Basics: Descriptive Statistics (AS & A Level)

Descriptive statistics are used to summarize data. These are the tools you likely already know, but they are vital for Paper 2.

A. Measures of Central Tendency

These tell us where the "middle" of our data lies:

Mean (\(\bar{x}\)): The arithmetic average. Add all values and divide by the number of samples (\(n\)). It’s great for normal data but can be "skewed" by one or two very weird, extreme numbers (outliers).
Median: The middle value when data is put in order. It’s useful if you have outliers.
Mode: The most frequent value. This is helpful for categorical data (e.g., the most common type of substrate on a beach).

B. Measures of Spread

These tell us how "stretched out" or "clustered" our data is:

Range: The difference between the largest and smallest values. It’s simple but limited because it only looks at the extremes.

C. Special Marine Indices

The syllabus requires you to use specific formulas for ecology. These are usually provided in the exam, but you must know how to plug in the numbers:

Lincoln Index: Used to estimate population size (\(N\)) of mobile animals (like crabs) using mark-release-recapture.
\(N = \frac{n_1 \times n_2}{m_2}\)
Where \(n_1\) is the first sample caught/marked, \(n_2\) is the second sample caught, and \(m_2\) is the number of marked animals in the second sample.

Simpson’s Index of Diversity (\(D\)): Used to measure the biodiversity of a habitat. A higher value usually means a more diverse and stable ecosystem (like a coral reef).

Quick Takeaway: Descriptive stats summarize what happened in your specific experiment, but they don't prove a general rule.

2. Understanding Variation (A Level Only)

In Paper 4, you need to go deeper into how much your data varies. This tells us how "reliable" our mean is.

Standard Deviation (\(s\))

The Standard Deviation tells us how much the data spreads around the mean.
- Small \(s\): Data points are close to the mean (high precision/consistency).
- Large \(s\): Data points are spread out (low consistency).

Standard Error (\(SE\)) and 95% Confidence Intervals (CI)

Standard Error (\(SE = \frac{s}{\sqrt{n}}\)) helps us estimate how close our sample mean is to the "true" population mean.
From this, we calculate the 95% Confidence Interval:
\(\text{95% CI} = \text{mean} \pm (2 \times SE)\)

This tells us that we are 95% confident that the real average of the entire ocean (the population) falls within this range. If this range is small, your experiment was very reliable!

Error Bars on Graphs

When you draw a bar chart or line graph, you often add error bars representing the \(SE\) or 95% CI.
The Golden Rule of Overlap:
- If error bars overlap: The difference between the means is likely due to chance (not significant).
- If error bars do NOT overlap: There is likely a significant difference between the groups.

Memory Aid: "Overlap = Oh well, nothing special." "No overlap = Noticeable difference!"

3. Statistical Tests: Searching for Significance

Statistical tests help us test a null hypothesis (which usually states that there is no relationship or difference).

A. Spearman’s Rank Correlation (\(r_s\))

Use this when you want to see if two variables are linked (e.g., does water temperature affect the rate of photosynthesis?).
- Values range from \(+1\) (perfect positive correlation) to \(-1\) (perfect negative correlation).
- A value of \(0\) means no correlation at all.
Important: Correlation does not prove causation. Just because two things change together doesn't mean one caused the other!

B. Chi-squared (\(\chi^2\)) Test

Use this for "categorical" data where you compare Observed (O) results with Expected (E) results (e.g., do fish prefer to swim near the surface or the bottom?).
The formula is: \(\chi^2 = \sum \frac{(O-E)^2}{E}\)
(Note: You only need to handle data in one row or one column for this syllabus).

4. Critical Values and the "Magic Number"

Summary Checklist for Success

1. Identify the Goal: Are you comparing means (Error bars/95% CI), looking for a link (Spearman's), or comparing counts (Chi-squared)?
2. Check the Units: Always use the correct notation (e.g., \(ppt\) for salinity, \(kg \cdot m^{-3}\) for density).
3. Significant Figures: Match your answer to the same number of significant figures as the raw data provided (or one more).
4. Command Words: If the question says "Evaluate," look at those error bars! If they overlap, the data might not support the conclusion.

Don't worry if this seems tricky at first! The exam often provides the formulas and even parts of the calculation for you. Your job is to be the "detective" who interprets what the final number actually means for the marine environment.

Next Step: Practice using these tools in the "Analysis, interpretation and conclusions" chapter to see how they fit into a full investigation!