Welcome to Data Analysis and Evaluation!

You’ve finished your experiment, your lab coat is off, and you have a page full of numbers. What now? In this chapter, we learn how to be "Science Detectives." We will look at how to turn those numbers into meaningful conclusions and, more importantly, how to spot if something went wrong. This is a vital skill for your Pearson Edexcel International GCSE Science (Single Award) exams, as about 20% of your marks come from experimental skills!

1. Analysing Your Data

Analysing data means looking for patterns or trends in your results. You want to see if changing your independent variable actually caused a change in your dependent variable.

Finding the Pattern

When looking at a graph or a table, ask yourself: "As X gets bigger, what happens to Y?"

  • Positive Correlation: As one variable increases, the other increases (e.g., as light intensity increases, the rate of photosynthesis increases).
  • Negative Correlation: As one variable increases, the other decreases (e.g., as the distance from a light source increases, the light intensity decreases).
  • No Correlation: The points are scattered everywhere; there is no clear link.

Drawing a Conclusion

A conclusion is a simple statement that describes the relationship you found. Top Tip: Always use data from your results to support your conclusion. Instead of just saying "It got faster," say "The rate increased from \(20\text{ cm}^3/\text{min}\) to \(40\text{ cm}^3/\text{min}\) when the temperature was doubled."

Quick Review: A conclusion should always refer back to your original aim or hypothesis!

2. Spotting Anomalies

An anomaly (or outlier) is a result that does not fit the pattern of the rest of the data. It’s the "odd one out."

How to spot them:
  • In a table: Look for a number in a set of repeats that is much higher or lower than the others.
  • On a graph: Look for a point that is far away from the line of best fit.
What to do with them:

If you identify an anomaly, you should not include it when calculating your average (mean). If you include a "weird" result in your average, your whole set of data becomes less reliable.

Example: In three repeats, you get: \(12.1\text{s}\), \(12.3\text{s}\), and \(19.5\text{s}\). The \(19.5\text{s}\) is an anomaly.
Calculation for mean: \( \frac{12.1 + 12.3}{2} = 12.2\text{s} \)

3. Evaluating the Method: Accuracy, Reliability, and Validity

This is often the trickiest part for students, but it's very important. Don't worry if these terms seem similar at first—here is how to tell them apart:

A. Reliability

Reliability is about how much you can trust your results. Are they consistent?

  • How to improve it: Repeat the experiment at least three times and calculate a mean.
  • How to check it: If your repeats are very close together (e.g., \(5.1\), \(5.2\), \(5.1\)), your results are reliable. If they are spread out (e.g., \(5.1\), \(8.9\), \(2.4\)), they are unreliable.

B. Accuracy

Accuracy is about how close your result is to the "true" or "real" value.

  • How to improve it: Use better equipment. For example, using a gas syringe to measure gas volume is more accurate than counting bubbles by eye, because bubbles can vary in size.
  • Using a digital thermometer that reads to \(0.1^{\circ}\text{C}\) is more accurate than a glass one that only shows every \(1^{\circ}\text{C}\).

C. Validity

Validity asks: "Was it a fair test?" Did you actually test what you set out to test?

  • How to ensure it: You must control all your control variables. If you are investigating how light affects plant growth, but you forget to give all the plants the same amount of water, your experiment is invalid because the water might be what's causing the growth, not the light.

Memory Trick:
Repeats = Reliability
Apparatus = Accuracy
Variables = Validity

4. Suggesting Improvements

In the exam, you might be asked how to improve a specific experiment. Here are the most common "Science Student" answers that actually gain marks:

1. Improve Measurement Precision

Instead of "be careful," suggest using a piece of equipment with a smaller scale.
Example: "Use a burette or a measuring cylinder with smaller graduations to measure volume more precisely."

2. Reduce Human Error

Humans are slow at reacting.
Example: "Use a data logger or light gates to measure time instead of a stopwatch to remove human reaction time."

3. Control the Environment

Sometimes the room temperature or light changes during the day.
Example: "Use a water bath to keep the temperature constant throughout the experiment."

5. Summary and Key Takeaways

When you are evaluating a method or analysing data in your IGCSE exam, keep these points in mind:

  • Look for the trend: Describe what happens to the variables using data.
  • Circle anomalies: Spot them on graphs and ignore them in averages.
  • The "Mean" Rule: \( \text{Mean} = \frac{\text{Sum of results (minus anomalies)}}{\text{Number of results}} \)
  • Evaluation: Critique the Accuracy (equipment), Reliability (repeats), and Validity (control variables).

Common Mistake to Avoid: Never just say an experiment was "wrong" or "bad." Always explain why using scientific terms. Instead of "the results were bad," say "the results were unreliable because the repeats were not consistent."

You're now ready to tackle data analysis! Remember, science is just as much about thinking about the "how" and "why" as it is about doing the experiment itself.