Duration

Define when the timer starts and stops

An operational definition should identify observable onset and offset. For reading, onset might be the first interaction with the chosen text and offset might be 20 continuous seconds away from reading. For a help conversation, onset could be the recognizable request and offset the partner’s completion of the agreed response.

Define how brief gaps, pauses, overlapping responses, session boundaries, and lost observation affect the timer. Two observers need the same rule.

Total duration and episode duration answer different questions

Total duration adds all episode lengths during the observation. Duration per occurrence preserves the length of each episode. Mean or median episode duration summarizes those values.

Suppose three episodes last 2, 4, and 12 minutes. Total duration is 18 minutes. Mean episode duration is 6 minutes. Median episode duration is 4 minutes. The individual values reveal the long episode hidden by either summary.

Compare time with an exposure denominator

When observation periods differ, report the portion of observed time:

Percentage of observed time = total duration divided by valid observed time times 100.

Twelve minutes during a 15-minute observation is 80%. Twenty minutes during a 60-minute observation is 33.3%. The larger total occupies a smaller share of its observation.

Keep the total minutes and observed minutes beside the percentage. Missing observation stays outside the denominator and receives its own coverage measure.

A fictional everyday-life example

Amara is a fictional 17-year-old who wants more uninterrupted time for digital illustration and an easy way to pause when her hands hurt. With Amara, the team defines drawing duration from the first stylus mark to a chosen break, task completion, or 30 continuous seconds away from the canvas. A pain or pause message through speech or AAC ends the episode and remains a successful self-advocacy response.

Across one 50-minute session, drawing episodes last 9, 14, and 11 minutes. Total duration is 34 minutes. Mean duration is 11.3 minutes, and median duration is 11 minutes. Amara chooses two breaks and rates the session comfortable.

Five additional minutes are unobserved because the tablet logging tool disconnects. Observation coverage is 50 of 55 planned minutes, or 90.9%. The pattern describes timing and provides no evidence that a support caused change.

Longer or shorter depends on the goal

Longer duration may fit a person-selected leisure, participation, or independence goal. Shorter duration may matter for pain, delayed support, an unwanted system interruption, or another outcome the person wants reduced.

Neither direction is automatically better. A shorter conversation could reflect efficient communication or a partner ending it too soon. A longer activity could reflect enjoyment or difficulty accessing a break. Review choice, comfort, effort, quality, access, and partner response.

Count still matters

The same total duration can come from one long episode or many short episodes. Report episode count when clustering affects interpretation. For health or safety concerns, retain individual episode lengths and any urgent action rather than relying on an average.

If responses occur too rapidly or boundaries overlap, duration recording may become unreliable. Consider product measures, event recording, or carefully chosen sampling when they fit the decision better.

Treat pauses and missingness consistently

Decide whether ordinary pauses remain inside an episode and how long a gap closes it. Mark observer interruptions, blocked views, device failures, and canceled opportunities separately. A missing timer record is not zero duration.

If an episode is already underway when observation begins, label it left-censored or use the team’s plain-language equivalent. The observed portion is a partial duration, rather than the full episode length.

Protect access and consent

Keep AAC, interpreters, mobility and sensory supports, breaks, food, water, bathroom access, prescribed care, relationships, and emergency help available. Do not delay a chosen break to extend a duration score.

For video, audio, telehealth, or device logs, follow applicable consent, assent, privacy, access, security, retention, and deletion rules. Explain what timing data are collected and who reviews them.

Review duration quality

Useful checks include valid observed minutes divided by planned minutes, episodes with complete onset and offset, partial episodes, timer failures, exclusions by reason, and sampled observer agreement on boundaries.

The Association for Behavior Analysis International basic-principles page places measurement within higher-education content. The BACB BCBA Test Content Outline, Sixth Edition identifies duration as a temporal dimension and covers reliability, validity, and representative procedure selection. Neither source sets a client’s duration target.