Control groups and matched pair tests (Edexcel GCSE Statistics): Revision Notes
Control groups and matched pair tests
When conducting scientific investigations, researchers need to ensure their results are reliable and valid. One of the biggest challenges is dealing with extraneous variables - factors that could affect the results but aren't the main focus of the study. Two important methods for controlling these variables are control groups and matched pair tests.
What are control groups?
A control group is a fundamental part of experimental design used to test how effective a treatment really is. Here's how it works:
The researcher uses random selection to divide participants into two groups:
- Test group (experimental group): Receives the actual treatment being tested
- Control group: Receives either no treatment at all or a placebo (an inactive treatment that looks identical to the real treatment)
The key point is that participants don't know which group they're in. This is called "blinding" and is crucial because it ensures both groups experience the same psychological conditions, preventing bias from affecting the results.
By comparing the results between these groups, researchers can determine whether any observed effects are genuinely due to the treatment rather than other factors.
What are matched pair tests?
Matched pair tests offer another approach to controlling extraneous variables. Instead of randomly assigning participants to groups, this method carefully pairs individuals who are similar in key characteristics.
The fundamental difference from control groups is that matched pair tests use careful pairing based on characteristics rather than random assignment. This provides more precise control over specific variables that might influence the results.
Here's how matched pair tests work:
- Each person in the treatment group is paired with someone in the control group
- The paired individuals share important characteristics (like age, gender, or initial condition severity)
- The only significant difference between paired individuals is whether they receive the treatment
- This pairing helps ensure that any differences in results are due to the treatment, not to differences between the people in each group
Worked example: Testing arthritis tablets
Worked Example: Pharmaceutical Testing of Arthritis Medication
A pharmaceutical company wants to test a new tablet for treating arthritis. They set up their experiment as follows:
Control group approach:
- One group receives the actual medication
- The control group receives tablets that look identical but contain no active ingredient (placebo)
- Arthritis severity is measured on a scale from 1 to 100
Using matched pairs for this study:
The researchers could pair participants based on important characteristics:
| Group 1 (medication) | Group 2 (no medication) |
|---|---|
| Pair A: Male, age less than 30, severity level less than 15 | Male, age less than 30, severity level less than 15 |
| Pair B: Female, age 30 to 40, severity level 15 to 25 | Female, age 30 to 40, severity level 15 to 25 |
This matching ensures that factors like age, gender, and initial arthritis severity are controlled across both groups.
Advantages and disadvantages of matched pairs
Advantages:
- Excellent control of extraneous variables like age, gender, and initial condition severity
- Results are more likely to show genuine treatment effects
- Reduces the impact of individual differences on the outcome
Disadvantages:
- Very time-consuming to find enough participants with matching characteristics
- Can be difficult to identify and match all relevant characteristics
- May limit the size of the study if suitable matches cannot be found
Practice scenario: The gardener's fertiliser test
Consider this example: A gardener wants to test whether a new fertiliser affects the growth of rose bushes. He divides his rose bushes into two groups - one receives the fertiliser and the control group receives no fertiliser.
Purpose of the control group: The control group allows the gardener to compare the growth of fertilised plants against plants that grow under normal conditions. Without this comparison, he couldn't determine if any observed growth differences were due to the fertiliser or other factors like weather, soil conditions, or natural plant variation.
Identifying extraneous variables: One major extraneous variable could be the location of the rose bushes in the garden. Some areas might receive more sunlight, better drainage, or different soil quality.
How to control this variable: The gardener could control location effects by ensuring both fertilised and unfertilised plants are distributed evenly across different areas of the garden, or by using matched pairs where each fertilised bush is paired with an unfertilised bush in the same location.
Summary
Key Points to Remember:
- Control groups help determine if a treatment actually works by comparing it against no treatment or a placebo
- Matched pair tests control extraneous variables by pairing similar individuals who differ only in the treatment they receive
- Both methods aim to ensure that observed differences in results are due to the treatment being tested, not other factors
- Control groups use random selection, while matched pairs use careful pairing based on key characteristics
- Matched pairs provide excellent control but can be time-consuming and difficult to set up properly