Designing investigations (AQA GCSE Statistics): Revision Notes

Designing investigations

When you want to test a hypothesis through data collection, there are several important constraints and considerations you must think about carefully. Planning a good investigation requires balancing practical limitations with the need to gather reliable, valid data.

Key considerations when designing investigations

Understanding these constraints will help you create more effective and realistic investigations:

Time and budget constraints Every investigation has practical limits. Consider how much time you have available to set up your study, collect data, and analyse results. Budget constraints might affect the size of your sample or the methods you can use.

Ethical responsibilities You must always respect people's dignity and rights when collecting data. This means getting proper consent, being honest about your research purposes, and ensuring participants can withdraw at any time.

Maintaining confidentiality It's crucial to keep all collected data secure and confidential. Personal information should be protected, and you should consider how you'll store and handle sensitive data throughout your investigation.

Practical convenience Think about how convenient it will be to collect your data. Local data collection is often more manageable, but consider whether this might introduce bias into your results.

Population and sampling methods You need to clearly identify your target population and decide on the best method to collect a representative sample. Consider whether your sampling method might exclude certain groups.

Planning for non-response

Non-response is when people don't participate in your survey or study. This is a common problem that can significantly affect your results, so you need to plan for it from the start.

Determining response requirements First, decide how many responses you actually need to conduct a valid analysis of your data. This depends on your research question and the level of accuracy you require.

Conducting pilot surveys A pilot survey is a small-scale trial run of your main survey. Send out a small number of questionnaires to test how many responses you're likely to receive. This helps you understand the response rate you can realistically expect.

Using pilot data for planning Take the proportion of responses from your pilot survey and use this to calculate how many surveys you'll need to send out for your full investigation. This mathematical approach helps ensure you gather enough data.

Calculating sample sizes

Here's how to work out how many surveys you need to send to get your target number of responses:

The basic principle If your pilot survey shows that 4 out of 5 people respond, then this proportion ($\frac{4}{5}$) should be the same for your main survey.

Step-by-step calculation method

• Calculate the response proportion from your pilot: responses received ÷ surveys sent
• Set up an equation where this proportion equals your target responses ÷ surveys to send
• Solve for the number of surveys to send

Worked example: tea or coffee preference study

Let's look at a practical example of these calculations in action.

Zeedan wants to investigate whether people in the UK prefer tea or coffee. He conducts a pilot survey with 270 people and receives 180 completed responses. Now he wants at least 400 completed surveys for his main study.

Dealing with sensitive topics

Some investigations involve sensitive issues where people might be reluctant to participate or provide honest answers. For example, questions about personal finances, health issues, or potentially embarrassing situations require extra care.

Choosing data collection methods Consider whether interviews or questionnaires would be more appropriate for your topic. Questionnaires might feel more anonymous and encourage honest responses, while interviews allow for more detailed exploration of complex topics.

Advantages and disadvantages of different approaches Anonymous questionnaires can encourage honest responses but limit your ability to follow up on interesting answers. Interviews provide rich detail but might make people uncomfortable discussing sensitive topics.