Percentages totalling more than one hundred are not automatically an arithmetic mistake. First check whether categories overlap. A question allowing several selections measures something different from a question allowing exactly one.
Need help choosing a next step? Scan the contact QR code below or contact Hanlin on WeChat. Send CBPA MULTI SELECT 1031 and tell us your school year and question.
Begin with fictional respondent counts
In an original classroom example, twenty people answer a question allowing them to select every product feature they value. Twelve choose price, ten choose speed and eight choose design.
Using the twenty respondents as the denominator gives 60%, 50% and 40%. These percentages sum to 150%, because one respondent can appear in several feature counts.
Separate people from selections
The counts contain thirty selections, not thirty different respondents. Dividing by thirty would produce 40%, about 33.3% and about 26.7%: shares of all selections, rather than shares of people choosing each feature.
Both summaries have defined meanings, but their labels must differ. Calling 40% the proportion of people who selected price would be wrong in this example; that proportion is 12/20 = 60%.
Choose a display consistent with overlap
A bar chart can show the respondent percentage for each feature, with a note that multiple selections were allowed. A pie chart suggesting three exclusive parts of the same group would conceal the overlap.
The three totals alone cannot identify how many people chose all three features or at least one. Those questions require information about intersections or individual response patterns. Do not invent that detail to make the chart complete.
Preserve the wording with the results
Record the exact question, number answering, whether multiple selections were allowed and how incomplete answers were handled. If a different survey asks for one favourite feature, its categories follow a different counting rule and cannot be compared without explanation.
These figures are invented for general business-research learning relevant to CBPA interests. They are not customer evidence, an organiser's required survey or a completed student submission. Students should independently collect any permitted real data, keep identities protected and describe limitations rather than present these teaching numbers as findings.
Continue reading: CBPA Business Skills: A SWOT Label Is Not Evidence.
Contact Hanlin
Planning a first attempt or reviewing a topic you find difficult? Scan the QR code and send CBPA MULTI SELECT 1031. Include your school year, country or competition region, experience and the question you want to discuss.
On your phone, use the button or save the QR image and scan it within WeChat.
Official information and further reading: Source 1
