Introduction to Evaluating Methods
In Biology Unit 6, you aren't just expected to carry out experiments; you are expected to be a science detective. This chapter focuses on how we look back at an investigation and ask: "How much can I trust these results?" Evaluating a method involves identifying where things might have gone wrong (errors), how precise our tools were (uncertainty), and how we can make the experiment better next time.
Don't worry if this seems technical at first! Most of it comes down to common sense and being observant about the tools you use in the lab.
1. Understanding Uncertainties and Resolution
Every piece of equipment has limits. If you use a ruler to measure a leaf, and the smallest marks are millimeters, you cannot be 100% certain about anything smaller than a millimeter.
Resolution is the smallest change an instrument can detect. For example:
• A standard thermometer might have a resolution of \(1^{\circ}\text{C}\).
• A digital balance might have a resolution of \(0.01\text{g}\).
• In Unit 6, you must also consider dimensions, such as the area of a quadrat or the volume of a beaker, and how these affect your measurements.
Uncertainty is the range of values within which the "true" value is expected to lie. Usually, the uncertainty is considered to be plus or minus half of the smallest graduation on your tool. If a syringe is marked every \(1\text{ cm}^{3}\), the uncertainty is \(\pm 0.5\text{ cm}^{3}\).
2. Types of Errors
When evaluating a method, you need to distinguish between two main types of errors:
Systematic Errors: These are "built-in" errors that happen every single time you take a measurement. They shift all your results in the same direction.
Example: A weighing balance that isn't zeroed properly (a zero error) will make every mass reading exactly \(0.05\text{g}\) too heavy.
How to fix: Calibration of instruments (e.g., using a buffer solution to calibrate a pH probe) helps eliminate systematic errors.
Random Errors: These are unpredictable and vary from one measurement to the next. They cause results to be spread around the true value.
Example: Human reaction time when stopping a stopwatch, or slight fluctuations in room temperature.
How to fix: Take repeat readings and calculate a mean. This helps "cancel out" the effect of random errors.
3. Calculating Percentage Error
To understand how much an uncertainty actually matters, we calculate the percentage error. A small error on a huge measurement is fine, but a small error on a tiny measurement can be a big problem!
The formula is:
\(\text{Percentage Error} = \frac{\text{Uncertainty}}{\text{Measured Value}} \times 100\)
Example: If you measure \(10\text{ cm}^{3}\) of a solution using a pipette with an uncertainty of \(0.1\text{ cm}^{3}\):
\(\text{Percentage Error} = \frac{0.1}{10} \times 100 = 1\%\)
Quick Tip: To reduce percentage error, try to measure larger quantities. Measuring \(100\text{ cm}^{3}\) with the same pipette would drop the error to \(0.1\%\)!
4. Evaluating the Experimental Design
In exam questions, you may be asked to criticise a method. Here is a checklist of what to look for:
1. Range and Number of Readings: Did the researcher test enough values of the independent variable? Usually, at least five different values are needed to see a trend. Did they repeat each one at least three times?
2. Control Variables: Were factors like temperature, pH, or light intensity actually kept constant? If the method says "the experiment was done at room temperature," that is a weakness because room temperature can change throughout the day.
3. Measurement Technique: Was the method for measuring the dependent variable accurate? For example, using a colorimeter is much more objective than just "looking at the color" with your eyes (which is subjective).
4. Significant Figures: In Unit 6, you must report results only to the limits of the least accurate measurement. If your equipment only measures to one decimal place, your final answer shouldn't have four!
5. Calibration and Ethical Issues
Instrument Calibration: This is a specific requirement for Unit 6. It involves checking an instrument against a known standard. For example, ensuring a microscope graticule is calibrated using a stage micrometer so that your size measurements are actually correct.
Ethical Issues: When evaluating methods involving living organisms (like brine shrimp or germinating seeds), always consider their welfare. This includes using the minimum number of organisms necessary and returning them to their natural habitat if possible.
Key Takeaways for the Exam
• Random errors are reduced by repeats and means; systematic errors are reduced by calibration.
• Resolution is the smallest division on your tool; uncertainty is usually half that division.
• When asked to suggest improvements, think about using equipment with higher resolution or better ways to control "messy" variables like temperature.
• Always check your significant figures—don't provide more detail than your tools allowed you to measure!
Note: For help with choosing statistical tests or drawing the right graphs for your data, see the chapter on Data Analysis, Statistics and Graphs (Unit 6).