Four steps, four chances for bias
A club wants to know what activities its members want. It asks, collects, tidies and reports. At each step something can tilt the answer.
If you know where the lean gets in, you can block it. Today you learn where, and how.
Key words
| Word | What it means |
|---|---|
| Bias | an unfair lean towards one side |
| Demographic | a group of people described by age, sex, area, background or similar facts |
| Inclusive | making sure that all the groups of people are counted |
| Objective | judging by facts and not by feelings or favourites |
| Random sample | people picked by chance, so each has the same chance of being chosen |
| Leading question | a question worded to push people towards one answer |
| Margin | the gap you accept before shares count as different |
The cells students fill in are left blank in this preview.
By the end of this lesson
Note
I can explain clearly how bias is created inside a source and how to avoid it.
Starter
Have a go
Starter
3 questionsYour answers are marked here and are not kept. A college account keeps every student’s marks.
Where bias gets in
| Step | How bias is created | Example |
|---|---|---|
| Choosing who | The sample leaves a group out or over-counts one | A survey handed out only at the sports hall |
| Asking | The question pushes towards an answer | "Do you agree that the great new gym is a success?" |
| Recording | A form cannot hold some people's answers | A form with only two choices for a question that needs more |
| Cleaning | Records are removed because the analyst dislikes them | Deleting every low rating |
| Reporting | Only the helpful results are shown | A chart that leaves out the groups that disagreed |
The cells students fill in are left blank in this preview.
Three ways to avoid bias
First, stay based on fact and evidence. Record where each figure came from and show the method.
Second, be inclusive. Make sure every demographic is fairly counted. Third, be objective. Judge by the numbers, and let someone else check your work.
Actions that block bias
| Aim | Action | Blocks |
|---|---|---|
| Fact and evidence | Cite sources, publish the method, keep a note of removed records | Cleaning and reporting bias |
| Inclusive and fully representative | Pick a random sample, check each group's share, ask in more than one way | Selection bias |
| Objective | Test questions for leading words, ask a second person to review | Asking bias and confirmation bias |
The cells students fill in are left blank in this preview.
Run it: which shares are unfair?
groups = ["Y7", "Y8", "Y11"]
sample = [5, 20, 60]
school = [20, 20, 20]
margin = 10
for place in range(len(groups)):
gap = sample[place] - school[place]
if gap < 0:
gap = 0 - gap
if gap > margin:
print(groups[place], "is off by", gap)