Understand the relationship
The reasoning behind the result
Nondetection and missingness are different observations
A zero records that a survey took place without detecting the species. NA records that no survey outcome is available. Replacing NA with zero would increase apparent effort and may distort both site and visit summaries.
A site with only NA entries is listed but excluded from the denominator of surveyed sites. Its occurrence is unknown, not negative.
One detection changes a site's observed status
Observed share = detected sites/surveyed sites
A surveyed site is classed as observed if any entered visit has a detection. Repeated visits to the same site contribute additional evidence but do not become additional independent sites in this proportion.
The visit detection percentage uses completed visits as its denominator and is reported separately. It generally differs from the fraction of sites where the species was ever detected.
Observed share is not detection-adjusted occupancy
Even an occupied site may produce no detections when detection is imperfect. Consequently the observed share should not be presented as an estimated true occupancy rate without an appropriate model and assumptions.
This tool does not fit detection probability, closure, covariate effects or false-positive rates. Retaining the full history makes the record inspectable and suitable for separate analysis rather than silently claiming those quantities have been estimated.
Follow the numbers
Three detections sites among four surveyed sites
- Pond histories 101, 000, 010 and 111 contain at least one detection at three of four sites.
- The observed site detection share is 3/4 = 75%.
- There are six detection visits among twelve completed visits, giving 50% at the visit level.
75% describes where detection occurred in this record. It is not a fitted probability that a site is occupied.
Quick guide
How to use this calculator
- Give each site a unique name and each scheduled visit its own column.
- Enter 1 for a detection, 0 for a completed nondetection and NA for an uncompleted survey.
- Inspect site coverage and visit-level detections before reading the observed site share.
- Use a separate occupancy model if you need to estimate occurrence while accounting for imperfect detection.
Calculation method
Calculation and interpretation
Describe where a species was observed and how much survey effort supports that record.
Observed site detection share = sites with ≥1 detection / sites with ≥1 completed visit
Worked example
Three detections sites among four surveyed sites
75% describes where detection occurred in this record. It is not a fitted probability that a site is occupied.
Observed site detection share = sites with ≥1 detection / sites with ≥1 completed visit
Supported inputs
Precision and limits
Observed records only
Entered records do not establish unbiased observation, independent sampling or a biological cause. Missing observations must remain distinguishable from measured zeros. No diagnosis or intervention is inferred.
Continue calculating
Related calculators
