> ## Documentation Index > Fetch the complete documentation index at: https://docs.sightlinebehavior.com/llms.txt > Use this file to discover all available pages before exploring further. # Results explained > Look up what each number in Results means, how it is calculated, and which observations or scores it includes. Results for every method except Anecdotal opens with a Summary card: one number per behavior, with a line under it saying what the number is out of. Rates and percentages of time divide by the time the recording clock ran, so paused time is left out. Anecdotal observations have no computed measures. Any count you report from an anecdotal log is your own, so say so. ## Frequency | Measure | How it is calculated | How to read it | | - | - | - | | Rate | Occurrences ÷ minutes observed, shown as 1.2 /min. | 12 occurrences in 10 minutes is 1.2 per minute. Rates compare observations of different lengths only when definitions and conditions match. | | Occurrences | Every recorded count, shown under the rate, such as "12 occurrences". | Report it with the observation length. | | Peer rate | The peer's occurrences ÷ the same minutes, shown as "Peer 0.4/min". | With Peer comparison on, the peer was watched, so "Peer 0/min" is a real result. If you did not watch the peer, select ⋯ → Remove peer comparison. | | Gaps | Event Record → Gaps, once there are two or more occurrences: the time between them, with Average gap and Longest gap. | Whether the behavior clusters or spreads out. A gap marked ≥ ("at least") reaches the start or end of the observation or activity, so it may be longer. The average leaves it out. | ## Interval methods | Measure | How it is calculated | How to read it | | - | - | - | | Percent of intervals | Intervals marked ✓ ÷ scored intervals, shown with "8 of 20 scored". | 40% in Partial Interval means the behavior occurred in 40% of scored intervals, not for 40% of the time. | | Peer percent | The peer's intervals marked ✓ ÷ the peer's scored intervals, shown as "Peer 80%". Until the peer has 3 scored intervals, it shows Peer with a dash in place of the percentage. | Peer intervals replace student intervals, so each percentage has its own count. Report both, such as 28 of 80 and 16 of 20. | What ✓ means depends on the method: * Partial Interval: the behavior occurred at some point in the interval. It says nothing about how often or how long. * Whole Interval: the behavior occurred throughout the interval. An interval you left untouched is scored ✓. * Momentary Time Sampling: the behavior was occurring at the sampling moment. Intervals with no score, such as ones you cleared, are left out of both numbers. "No scored intervals" means there is no percentage, which is not the same as 0%. Name the method and interval length whenever you report a percentage. Interval Record also offers Gaps (Partial Interval: stretches between ✓ intervals) and Runs (Whole Interval: consecutive ✓ intervals). Both are counted in whole intervals, not exact times, and a run is a minimum. A value marked ≥ reaches an observation boundary or a break in scoring and is left out of the average. ## Duration | Measure | How it is calculated | How to read it | | - | - | - | | Total time | The behavior's episodes added together, such as 4:30. | How long the behavior went on during the observation. | | Percent of time | Total time ÷ time observed, shown as "23% of time". | Behaviors can overlap, so don't add percentages across behaviors. | | Episodes | The number of separate episodes, such as "3 episodes". | Read it with total time: one long episode and several short ones can add to the same total. | The card header shows the time it divides by, such as "20 min observed". ## Latency | Measure | How it is calculated | How to read it | | - | - | - | | Average latency | The mean time from cue to response, across trials with a response, shown as "6.5 s average latency". | Check Latency by Trial for one long wait pulling the average up. | | Trials | All trials, shown as "8 trials · 2 no response": 8 trials in total, 2 of them with no response. | If a trial is still open when you select End, Sightline asks to record it as No response. | | No response | Trials you marked No response, left out of the average. | Not a zero-second response. Report the number alongside the average. | | Range | The shortest and longest response. Sightline's written summaries and exports state it. | Report it with the average when responses varied. | Latency times one behavior per observation. ## ABC | Measure | How it is calculated | How to read it | | - | - | - | | Events | Each entry counts once overall and once for each behavior it holds. Behaviors with none show 0. | How many events you recorded, not every time the behavior occurred. Behavior counts can add up to more than the total. | | Share of events | Behavior's events ÷ all ABC events, shown as "of 10 events (40%)". | 4 of 10 events is 40% of events, not 40% of observation time. | | Shared events | Counts events with more than one behavior, plus each behavior’s “alone” and “with” counts. “With” counts can overlap. | “3 of 10 events included more than one behavior” describes the overall total. “7 alone · 3 with Throwing materials” means 7 events with only this behavior and 3 that also included Throwing materials. | | Tag percentages | In the antecedent and consequence diagram, the percentage of events that carry each tag. | An event with two tags counts under each, so percentages can add to more than 100%. | | Pathways | Each antecedent and consequence pairing, with the number of whole events that contain it, such as "5 of 9". Events missing either tag are left out, and the card says how many. | Which sequences repeat. One event can appear in several pathways. | | Function | The Function column in ABC Entries, read from consequence tags: Escape, Attention, Tangible, or Automatic. The footer totals each. | What followed, not a finding of function. An event can count toward more than one. Review events under "Uncategorized tags" or "No consequence tags" first. | A common consequence is not, by itself, evidence that it maintained the behavior. ## Measures by activity When you select an activity, Summary recalculates for that activity: * Frequency divides by the activity's minutes. Under a minute, Summary shows the count with "Too short to rate". * Intervals use the intervals recorded during the activity. * Duration counts only the part of each episode inside the activity, shown as, for example, "23% of Reading time". * Latency trials belong to the activity running at the cue. * ABC events belong to the activity recorded with them. Check which activity is selected before you copy a number into a report. For how to word these measures, see [Describe results in a write-up](https://docs.sightlinebehavior.com/writing-up-observation-results.md). ## Related * [Review an observation in Results](https://docs.sightlinebehavior.com/results.md) * [Compare observations](https://docs.sightlinebehavior.com/comparison.md) * [Record a peer comparison](https://docs.sightlinebehavior.com/peer-comparison.md) * [Fix a mistake](https://docs.sightlinebehavior.com/fix-a-mistake.md) This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.