Data accuracy and validity
Big Idea: Accuracy = measured correctly.
Validity = appropriate for purpose.
| Term | Meaning |
|---|---|
| Accurate | Correct measurement (no systematic error) |
| Valid | Appropriate for the research question |
| Reliable | Consistent if measurement repeated |
Worked example
Scale always reads 2 kg too high.
Are readings accurate?
Valid?
Solution
- Systematic error of 2 kg → NOT accurate
- Relative comparisons still work (order preserved) → VALID for ranking
Final answer
Accurate? NO. Valid? Depends on purpose.
Free preview
This is the free notes preview
You're reading the free notes. Aimnova Pro unlocks the full study experience — and you can try it with your first topic free to keep:
- FlashcardsLock in vocabulary and key terms with spaced repetition.
- Practice questionsAnswer exam-style questions and get instant AI marking.
- Mock exams & past-paper vaultSit full mocks and see exactly how examiners award marks.
- Personalised study planA daily plan built around your exam date and weak areas.
Bias and measurement error
Systematic error: Consistent mistake in one direction (e.g., scale always 2 kg high).
Random error: Unpredictable variation (e.g., 0.1 kg variation each weighing).
Bias: Systematic error in data collection (e.g., surveying only happy people).
| Error type | Cause |
|---|---|
| Bias | Flawed sampling or question |
| Systematic | Instrument error |
| Random | Measurement variation |
Know your predicted grade
Take timed mock exams and get detailed feedback on every answer. See exactly where you're losing marks.
Minimizing errors in data collection
Before: Clear definitions Calibrated instruments Random sampling Careful question design
During: Train collectors Regular calibration Consistent method Multiple measurements
After: Check for outliers Document issues Report error margins
Practical improvements
Worked example
Survey on student happiness done at cafeteria.
What biases?
How improve?
Solution
- Location bias: Cafeteria students may be happier
- Sampling bias: Misses absent students, others
- Improvements: Random time/location + anonymous
Final answer
Result: Representative sample, less biased.
Get feedback like a real examiner
Submit your answers and get instant feedback — what you did well, what's missing, and exactly what to write to score full marks.
Two problems in real data: Missing data: a value was never recorded (a blank cell). Recording error: a value was written down wrongly (a slipped decimal point, the wrong unit, two digits swapped).
Worked example — race times
A coach records the 5 km times, in minutes, of 8 runners:
24, 27, 25, (blank), 26, 260, 23, 28
(a) How should the blank be dealt with when finding the mean?
(b) The value 260 was meant to be 26.0. Find the mean with the 260 and with it corrected.
Step by step
- (a) Do not invent a value: leave that runner out and say so.
- (b) With the recording error.
- (b) Corrected to 26.
Final answer
(a) leave the missing value out (use the 7 recorded times) and state it (b) 59 minutes with the error, 25.6 minutes corrected
For "explain the bias"
- Name the problem (leading question, non-response …).
- Say which answers it pushes up or down.
- Give a fix if asked: neutral wording, follow up non-replies, anonymous answers.
Learn what examiners really want
See exactly what to write to score full marks. Our AI shows you model answers and the key phrases examiners look for.
Exam-style question — a library survey [4 marks]
A library emails all 2400 members the question "Do you agree that our excellent new opening hours should be kept?" 310 members reply and 84 % of them say yes. The library says that most members support the new hours.
(a) Identify one problem with the wording of the question. [1]
(b) Explain why the 310 replies may not represent all members. [2]
(c) Suggest one improvement to the survey. [1]
Step by step
- (a) The word "excellent" suggests the answer.
- (b) Who replies, and which way it pushes.
- (c) A fair fix.
Final answer
(a) leading question ("excellent") (b) only 13 % replied and those who reply may hold stronger views, so the 84 % may not hold for all members (c) e.g. neutral wording, or contact a random sample and follow up those who do not reply