A sample limitation is a reason why the group a business asked may not give a trustworthy picture of its whole market. The questions to ask are: how many, who, where, when and how chosen? In IGCSE Business you are often asked to judge how reliable research is. The strongest answers name a specific weakness in the case and say what it does to the conclusion.
This lesson builds on wording survey items fairly. Fair wording and a fair sample work together. The topic belongs to the market research module.
What can go wrong with a sample?
- Too small: a few answers can be changed a lot by one or two people.
- Not representative: the group differs from the target market in age, income, area or habits.
- Convenient rather than fair: the business asks whoever is easy to reach, such as friends or people at one spot.
- Self-selected: people choose to reply, often because they feel strongly.
- Low response rate: many people asked did not reply, so the replies may not match those who did not.
- Wrong time or place: a weekday morning crowd is not a Saturday night crowd.
How to assess a sample, step by step
- Define the target market. Who does the decision affect?
- Describe the sample: how many, from where, at what time, chosen how.
- Compare the two. Which kinds of people in the market are missing?
- Check the response rate, if questionnaires were sent out.
- State the effect on the result, then say what could be done to improve it.
Worked example
Pandai Pastry plans to open a dessert shop in Petaling Jaya for students and young families across the town. The owner surveys 40 people outside a shopping mall on a Saturday afternoon. Of these, 28 say they would buy a RM12 dessert box weekly.
Step 1, the target market: students and young families across the town.
Step 2, the sample: 40 mall visitors, one place, one weekend afternoon, chosen by approaching whoever walked past.
Step 3, compare: weekday visitors, people who do not shop at malls and families from other areas are missing. Mall visitors may also spend more freely.
Step 4, the result: 28 ÷ 40 = 0.7, which is 70% of the sample. Only 70% of these 40 people, not 70% of the town.
Step 5, the effect: the sample is small and convenient, so the figure is a weak guide to the whole town. It does not show the shop will fail, but it should not be the only evidence. A stronger plan would ask more people across different places and days, then compare.
Now suppose the owner also emails 200 past customers, and 25 reply. The response rate is 25 ÷ 200 = 0.125, which is 12.5%. The 25 who replied may be the most enthusiastic or the most annoyed, so the other 175 are unknown.
The mistake to watch for
A common slip is to treat the sample percentage as a fact about the whole market.
Mistaken answer: “70% of people in Petaling Jaya would buy a dessert box, so the shop will succeed.”
This applies a result from 40 mall visitors to the entire town, and it jumps from “would buy” to “will succeed”.
The correction is to word the claim to match the evidence. For example: “In a survey of 40 mall visitors, 70% said they would buy a dessert box weekly. But the sample is small and from one place and time, so it may not represent the whole town”.
You can work out such percentages with the ratios tool. Remember that the sample’s limits remain even when the arithmetic is right.
Check yourself
Try these, then open each answer.
1. A stall owner asks 30 customers at 6 am on a Monday. 15 say they like the new breakfast set. State the percentage and one limitation.
Show answer
15 ÷ 30 = 0.5, which is 50%. A limitation: all 30 were early-morning customers, so people who visit at other times, or do not visit at all, are missing.
2. A shop sends 60 questionnaires and receives 9 replies. Calculate the response rate and say what it suggests.
Show answer
9 ÷ 60 = 0.15, which is 15%. It is low, so the 9 replies may come from people with strong views and may not reflect the 51 who did not reply.
3. A business can survey either 30 friends or 120 shoppers picked on different days in four neighbourhoods. Which sample is more reliable, and why?
Show answer
The 120 shoppers are more reliable. The group is four times larger, and drawing from different days and places makes it more likely to reflect the wider market. Friends may be similar to the owner and may want to be kind.
Where this leads next
Once you can judge the sample, you can decide how much weight a result deserves. Continue with using research evidence without treating opinion as fact, then test yourself on the practice set.
If your evaluation points tend to say “the sample is small” without applying it to the case, our teachers can help in online one-to-one Business tuition.