Identify the quantity behind the percentage
Open each survey's methods and result notes. Write the numerator and denominator in words before copying the percentage into a comparison chart. Do not infer the unit from a title that simply says “colony losses.”
Determine whether the result counts colonies, respondents or operations. If the publication summarizes each respondent's percentage and then averages those values, retain that method. If it pools reported colonies into a single fraction, retain the colony counts. Neither figure should be renamed as the other because both appear with a percent sign.
Two correct calculations can answer different questions
Consider a fictional survey with two respondents and a fixed observation period. One begins with ten eligible colonies and reports four lost. The other begins with one hundred eligible colonies and reports ten lost. For this arithmetic example, there are no additions or other changes to the defined population.
Their individual percentages are forty percent and ten percent. Giving each respondent equal weight produces an average of twenty-five percent. Pooling the reported colonies produces fourteen divided by one hundred and ten, or about 12.7 percent.
Both calculations use the same invented responses, but they describe different quantities. A comparison saying that one survey's twenty-five percent contradicts another's 12.7 percent would be misleading if the first reports the respondent average and the second reports the pooled colony fraction. Preserve the method before interpreting the difference.
Match the observation windows
Record the start and end definitions used by each survey. A named winter period, calendar year and interval between two visits should not be treated as interchangeable labels. Also distinguish the period described from the date a questionnaire was returned or a report was published.
Do not divide an annual percentage by two to invent a six-month result. Nor should two period percentages be added when their denominators or membership differ. Use the published result within its actual interval, or obtain a justified comparable measure from the underlying documentation.
Keep losses separate from the closing inventory
An opening and closing count do not necessarily describe everything that occurred between them. In another fictional record, an operation starts with one hundred colonies, records ten losses and adds ten replacement colonies. The closing count is again one hundred. The unchanged inventory does not erase the recorded losses.
Read how the survey handles additions, departures, replacement units and incomplete responses. Preserve its definitions rather than quietly substituting your own. If those details are unavailable, identify the comparison as unresolved instead of reconstructing an exact loss figure from the final inventory alone.
Make a comparison sheet before writing the conclusion
| Comparison field | Detail to retain from each report |
|---|---|
| Unit | Colonies, respondents or operations |
| Calculation | Pooled fraction, respondent average or another stated method |
| Period | Defined start, end and date basis |
| Eligibility | Who could participate and which responses were included |
| Population changes | Treatment of additions and departures |
| Uncertainty | Reported intervals, missing information and exclusions |
| Edition | Publication identity and any correction |
Keep voluntary respondents distinct from every operation in a region unless the survey's method supports that wider inference. Preserve any reported weighting and uncertainty without inventing a missing interval or treating it as zero uncertainty.
Write a useful comparison, including its limits
The conclusion can explain that two percentages use different units, cover different periods or cannot yet be reconciled. That is a substantive answer. It is more useful than ranking regions or operations using incompatible measures.
Use the mortality-denominator guide for basic count distinctions and the prevalence-claim guide for population scope. These examples concern report arithmetic and interpretation. They establish no real colony-loss estimate, cause of loss or management recommendation.
Questions before comparing headline percentages
Can both percentages be correct when they differ greatly?
Yes. As the fictional example shows, a respondent average and a pooled colony fraction can differ while using the same responses. Describe which quantity each survey reports before interpreting the difference as a disagreement.
Does an unchanged closing inventory mean no losses occurred?
Additions and departures can change the closing total. Keep the survey's actual handling of population changes visible. A final count alone does not reconstruct every loss or replacement during the period.
Can a seasonal result be converted into an annual rate?
Do not invent the missing interval by multiplication or division. The result belongs to the observation window and population defined in the source. A comparable measure needs an appropriate basis in the underlying documentation.
What if the methods are too incomplete to reconcile?
State the unresolved difference and identify the missing information. That is a useful finding, not a failed comparison. Avoid ranking regions or management systems from measures whose units or eligibility rules remain unclear.
How should the chart be labeled?
Name the unit, period and calculation beside each result. Put a concise limitation near the comparison rather than hiding it in a distant note. A reader should understand whether a bar represents colonies, operations or respondents without having to infer the denominator from the title.
Cite this reference
LivestockDefense. “Compare colony-loss percentages across apiary surveys.” Updated 2026-11-03. https://livestockdefense.com/guides/comparing-colony-loss-percentages-across-two-apiary-surveys/
Add the date you accessed the page. This identifies the current reference; keep a permitted saved copy if your work depends on its exact wording. A citation alone does not establish support for a claim.
How to cite a page and its revision ↗