Total duration versus duration per occurrence separates all observed time in a response from the length of each matched episode. Sum episode durations for total duration. Divide total duration by episode count for the mean, and preserve every episode value. Similar totals can arise from one long event or several shorter events with different clinical implications.

Choose the duration question before collecting

Total duration answers how much observed time was occupied by all episodes. Duration per occurrence answers how long a typical or individual episode lasted. The measures can move differently because total duration depends on episode count as well as episode length. State the decision question, response definition, onset, offset, observation window, unit, and episode-separation rule before timing begins.

Nori's worksheet treats each matched onset and offset as one episode. It records continuous timestamps in minutes and keeps episode count, individual durations, total observed exposure, and missing states. The same rules apply across the compared samples.

Build an episode-level source table

Use one row per episode with episode ID, onset, offset, calculated duration, source record, observer, setting, phase, access or health event, and any correction. Preserve the observation start and end and mark periods without valid visibility. A session summary should link back to every episode row.

Define how to handle brief pauses, overlapping response forms, episodes already occurring when observation starts, and episodes still occurring when observation ends. Boundary-truncated episodes may require a separate censored state because their complete durations are unknown.

Calculate Nori's total duration

Nori has three complete episodes lasting 2, 6, and 10 minutes.

Total duration = 2 + 6 + 10 = 18 minutes

The numerator is 18 observed minutes across three episodes. Report the observation exposure when comparing total duration across sessions. Eighteen minutes in a 30-minute observation and 18 minutes in a three-hour observation have different proportional contexts even though their totals match.

If the question asks for percent of observed time, divide total duration by valid observed duration under a predeclared rule. Do not calculate that percentage for Nori here because the example does not supply the full observation duration.

Calculate duration per occurrence

Mean duration per occurrence uses total duration divided by the number of complete episodes:

18 minutes / 3 episodes = 6 minutes per episode

The ordered episode durations are 2, 6, and 10 minutes. Their median is 6 minutes, minimum is 2, maximum is 10, and range is 8 minutes. The mean and median happen to match, yet the raw episodes show substantial spread. Report the individual values or a distribution summary when episode length matters.

Interpret total and per-occurrence values together

Two samples can share 18 minutes of total duration through one 18-minute episode, three 6-minute episodes, or nine 2-minute episodes. Those patterns can imply different interruption, recovery, opportunity, or burden questions. Likewise, a shorter mean episode can coexist with a larger total if episode count rises.

Keep episode count, total duration, mean or median duration, range, and exposure together. Avoid describing a lower mean as improvement without checking total time, frequency, target direction, client experience, risk, and the time-series pattern.

Handle missingness and boundary episodes

Report planned exposure, valid observed exposure, complete episodes, boundary-censored episodes, invalid records, and missing time. If an offset is lost, do not assign the observation end as the true endpoint unless the definition explicitly treats it that way. If observation begins during an ongoing episode, the elapsed time since onset is unknown.

With zero complete episodes, total observed duration may be zero under a valid full observation, but mean duration per occurrence is undefined because the divisor is zero. With one complete episode, the mean equals that episode and supplies no information about between-episode variability. Repeated missing offsets can make the duration summary unusable.

Graph episodes and summaries

Show onsets and offsets in time order using a duration plot, event bars, or a source table linked to the main graph. A companion summary can display exposure, count, total duration, mean, median, range, and missing states. Keep phase and context changes visible.

Nori had three complete episodes lasting 2, 6, and 10 minutes. Total duration was 18 minutes. Mean and median duration per occurrence were both 6 minutes, and episode durations ranged from 2 to 10 minutes. The observation exposure should accompany any comparison with another session.

Lobo and colleagues provide single-case measurement and analysis context. Pritchard and colleagues and Wirth and colleagues show why procedure accuracy and parameters matter when continuous and sampled measures are compared.

Review meaning and access with Nori

Use accessible communication to ask Nori whether the measured episodes, duration, and selected direction reflect a meaningful concern or goal. Maintain AAC, interpreters, mobility supports, food, water, bathroom access, prescribed care, rest, relationships, and emergency help. ASHA supports continuous AAC access.

Record whether support, health, setting, opportunity, observation burden, or treatment implementation changed. Qualified clinical interpretation should use the raw graph, episode pattern, client perspective, and other relevant evidence.

Keep professional sources in scope

The BACB ethics hub and CASP public summary provide professional context. The BCBA Test Content Outline includes temporal dimensions, validity, reliability, and representative measurement as examination content. These sources do not set universal duration boundaries or a clinical cutoff.

Clinician review checklist

When reviewing total duration versus duration per occurrence, confirm the question, response definition, onset and offset, separation rule, unit, planned and valid exposure, episode rows, complete and censored states, missingness, episode count, total-duration arithmetic, mean denominator, median and range when useful, graph, report wording, access, client perspective, implementation evidence, qualified owner, and review date.

Limits of duration summaries

Total duration and duration per occurrence do not establish measurement validity, treatment integrity, clinical importance, client benefit, future performance, or cause. Total duration is exposure dependent. A mean can hide an uneven episode distribution and is undefined when there are no complete episodes. Boundary censoring and missed onsets or offsets can bias both measures. Preserve raw timestamps and episode count, and avoid using either summary alone as a treatment rule.

Related resources

Sources