| Method | Description | Example |
|---|---|---|
| Survey/questionnaire | Written or online questions | Online poll about social media use |
| Interview | Face-to-face or phone questions | Structured interview with customers |
| Observation | Directly watching and recording | Counting cars at an intersection |
| Experiment | Controlled manipulation of variables | Testing a new drug vs placebo |
| Existing records | Using published/administrative data | ABS census records, hospital data |
Bias occurs when the sample does not accurately represent the population, leading to systematic errors in conclusions.
KEY TAKEAWAY: A biased sample produces results that consistently over- or under-estimate the true population value — no matter how large the sample is.
| Bias Type | Cause | Example |
|---|---|---|
| Self-selection bias | People volunteer to participate | Online reviews (only very happy/unhappy respond) |
| Convenience sampling bias | Selecting whoever is easiest to reach | Surveying only your friends |
| Non-response bias | People who don’t respond differ from those who do | Low survey return rate |
| Undercoverage bias | Some groups of the population are excluded | Phone survey misses people without phones |
| Question wording bias | Leading questions influence responses | “Don’t you agree that…?” |
Scenario: A magazine runs a phone-in poll asking readers: “Do you support the new shopping centre development?”
Identify the bias: Self-selection bias — only people with strong opinions (mostly opposed or strongly supportive) are likely to call. The result does not represent the general population.
Better method: Randomly select names from the electoral roll and conduct structured telephone interviews.
| Sampling Error | Bias | |
|---|---|---|
| Cause | Random chance in sample selection | Systematic flaw in method |
| Direction | Varies (can be above or below) | Consistently one direction |
| Fix | Increase sample size | Change sampling method |
EXAM TIP: VCAA often asks you to identify the type of bias and explain why it makes the sample unrepresentative. Always link your answer to the specific context given.
VCAA FOCUS: Know the difference between sampling error (unavoidable, random) and sampling bias (systematic, avoidable with better design). Increasing sample size reduces sampling error but does NOT fix bias.
