Skip to content
Tutorials Center
My learning

Topics

Lesson 8 of 40 ยท Life and the Scientific Method

Variables and Experimental Design

Plan a fair test: independent, dependent and control variables, control groups, replicates, sample size, accuracy versus precision, and a worked example.

An experiment is a question asked in a very controlled way. If you change one thing and watch what happens, you can say that the change probably caused the effect. If you change many things at once, you cannot tell which one mattered. This lesson shows how to design experiments so that the results mean something.

The three kinds of variable

A variable is any factor that can change.

Variable Meaning Example (growing bean plants)
Independent The one you deliberately change Amount of fertilizer
Dependent What you measure as the result Height of the plant after 21 days
Control (or controlled) variables Everything you keep the same Light, water, temperature, pot size, seed type, soil

A helpful memory aid: the dependent variable depends on the independent one. On a graph, the independent variable goes on the x-axis and the dependent variable on the y-axis.

If you leave a control variable uncontrolled, it can become a confounding variable. Suppose the fertilized plants sit on a sunnier windowsill. Then you cannot tell whether fertilizer or light caused any difference.

Control groups

A control group is treated exactly like the test group except that it does not receive the factor being tested. In the bean experiment, the control group receives plain water. Without a control, you cannot know what would have happened anyway.

Medical trials use the same idea. One group receives a new drug, another receives a placebo, a harmless pill that looks the same. A double-blind trial means neither the patients nor the people measuring the result know who received what, which reduces bias.

Replicates and sample size

Living things vary. Two seeds from the same packet will not grow identically. So you need repeats (replicates): several plants per group, not one. The more measurements you have, the more reliable the mean (average). Many school experiments use at least three repeats; research often uses far more.

  • Repeatable: you get similar results if you do the experiment again.
  • Reproducible: other people get similar results with their own equipment.
  • Reliable: results are consistent across repeats.
  • Valid: the experiment really tests what it claims to test (for instance, by controlling variables).

Accuracy and precision

These are different:

  • Accuracy means close to the true value.
  • Precision means measurements are close to each other, whatever the true value is.

A balance that always reads 2 g too high gives precise but inaccurate results. Using the right instrument, measuring carefully, and calibrating equipment improve accuracy. The resolution of an instrument also matters: a ruler marked in millimeters cannot measure to a tenth of a millimeter.

A worked example

Question: Does the temperature of water affect how quickly yeast produces gas?

  • Independent variable: temperature of the yeast mixture (for example 10, 20, 30, 40, 50 ยฐC).
  • Dependent variable: volume of gas collected in 5 minutes, in cmยณ.
  • Control variables: mass of yeast, volume and concentration of sugar solution, time, type of container, same kind of yeast.
  • Replicates: three tubes at each temperature.
  • Safety: hot water can scald; use a thermostatic water bath and wear eye protection.

The design is described here but results are not invented; you would run it for real. The next TryIt shows how to organize such an experiment as a table with a random order, which helps avoid bias, such as always running the hottest tube last.

Python Plan the experiment table

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

The blank column is where you would write your real measurements.

Choosing a good measurement

  • Use numbers and units where you can.
  • Pick a range of the independent variable that is wide enough to show a pattern, with at least five values for a graph.
  • Decide in advance how long to measure and what counts as "finished".
  • Record raw data directly in a table with a title, column headings, and units in the headings, not repeated in every cell.
  • Spot anomalies (results that do not fit). Check them, repeat if possible, and mention them; do not silently delete them.
flowchart LR
    A["Question"] --> B["Choose independent variable"]
    B --> C["Choose dependent variable and how to measure it"]
    C --> D["List control variables"]
    D --> E["Plan repeats and controls"]
    E --> F["Check safety and ethics"]
    F --> G["Run the experiment"]

A planning route for a fair test.

Why repeats help: a simulation

The second TryIt simulates measuring the height of plants that share the same real average, with random variation. It shows that the mean of repeats is steadier than a single measurement. This is invented random data for illustration only.

Python Mean of repeats is steadier

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

The spread of means of 5 should come out noticeably smaller, close to the single spread divided by the square root of 5 (about 0.9 instead of 2.0), which is a general result in statistics.

Do not change two things at once and then claim to know which one caused an effect. If you want to test temperature and light, run two separate experiments, or use a planned design where each combination is tested.

Check your understanding

Quick check ยท 4 questions

of 4 answered You scored / 4 ยท perfect!

  1. 1. In an experiment on how light intensity affects the rate of photosynthesis, what is the independent variable?

    It is the factor you deliberately change.
  2. 2. What is the purpose of a control group?

    A control group is treated the same except for the tested factor, giving a baseline.
  3. 3. A balance always reads 3 g too high, but gives the same reading each time. The readings are

    The readings agree with each other (precise) but not with the true value (not accurate).
  4. 4. Why take several repeats of each measurement?

    Repeats reduce the effect of chance variation and let you spot anomalies.

Summary

  • The independent variable is changed, the dependent variable is measured, and control variables are kept the same.
  • A control group gives a baseline; placebos and double-blind designs reduce bias.
  • Repeats and a good sample size make results reliable.
  • Accuracy is closeness to the true value; precision is closeness of repeated values.
  • Plan the table, safety, and number of values before you start.

Next: Graphs and Data in Biology