Latency recording examples for task initiation and transitions
By the IEP Atlas editorial teamLast updated
Short answer
Latency recording measures the time from the end of a cue to the start of a defined response, in seconds or minutes. Each example below names the cue that starts the timer, the response that stops it and a week of sample data. For example: timer starts when the teacher says "Begin page 12" and stops when the student writes the first answer; Monday 240 seconds.
A latency record has three parts: a cue with a clear end, a response with a clear start, and a stopwatch. Most bad latency data fails on the first two. "Started work" can mean opening a folder, picking up a pencil or writing an answer, and three adults will stop the timer at three different moments.
The 20 examples below each define the start and stop points, show sample data and mark what makes the setup strong. Students and numbers are fictional, not benchmarks. For latency goal wording, see the task initiation goal templates and the guide to task initiation IEP goals. For how latency fits with the other methods, see the IEP data collection guide.
Which latency setup fits the student
| What you see | What to time | Unit | Sessions to sample |
|---|---|---|---|
| Sits for minutes before starting independent work | Direction to first written mark | Minutes and seconds | 1 to 2 work periods a day |
| Slow to leave one activity for the next | Transition signal to arrival at the next spot | Seconds | Every scheduled transition, or 3 chosen ones |
| Slow to follow a spoken direction | End of direction to start of the action | Seconds | 5 to 10 directions per session |
| Slow to answer a question or a greeting | End of the question to the first word or sign | Seconds | 5 planned opportunities |
| Slow to come back after a break | End of the break signal to work restarted | Minutes | Each break |
Use latency when speed of starting is the problem. If the student starts fast but stops soon after, time how long they keep going with duration recording instead. The frequency vs duration guide covers that choice.
How to record latency on the sheet
The latency recording measurement guide gives the method. In short:
- Write the cue word for word at the top of the latency data sheet.
- Start the stopwatch the moment the cue ends, not when it begins.
- Stop it the moment the defined response begins.
- Record seconds for anything under 3 minutes and minutes with seconds above that.
- Note any prompt given while the timer ran, and when.
- Summarize each week with the median, not the mean.
Set a ceiling. If the student has not started after 10 minutes, stop timing, record "10:00+" and prompt. Without a ceiling, one bad morning eats the whole period and skews the week.
Task initiation examples (1 to 5)
| # | Timer starts | Timer stops | Sample week (median) | What makes it strong |
|---|---|---|---|---|
| 1 | Teacher finishes "Start the 8 problems on page 12" | Pencil writes the first digit of problem 1 | 4:10, 6:30, 3:45, 5:20, 2:50 (4:10) | The stop point is a visible mark on paper, not "looks ready" |
| 2 | Written direction card placed on the desk | Student writes their name and the first word of a response | 3:00, 2:15, 4:40, 1:50, 2:30 (2:30) | Same card every day keeps the cue constant |
| 3 | Timer on the board beeps to start a timed write | First word written on the lined page | 6:00, 7:15, 5:30, 8:00, 4:45 (6:00) | Timing writing apart from math shows which task type stalls |
| 4 | Laptop opens to the assigned reading program | First click inside the lesson, not on the menu | 95, 140, 80, 210, 60 s (95 s) | Excludes logging in and browsing from the "start" |
| 5 | Lab partner hands the student the materials | Student completes step 1 of the posted lab steps | 2:20, 1:40, 3:10, 2:00, 1:30 (2:00) | Ties the start to the student's own role, not the group's |
Weak: "Timed how long it took Mason to get going in math." Nobody can repeat that. Strong: "Timer starts when the teacher finishes the direction; stops when Mason writes the first digit of problem 1. Ceiling 10 minutes." Any adult in the room can take the same data.
Transition examples (6 to 10)
| # | Timer starts | Timer stops | Sample week (median) | What makes it strong |
|---|---|---|---|---|
| 6 | Chime for the end of centers | Student is seated at the carpet spot | 85, 120, 60, 150, 70 s (85 s) | A fixed spot gives an exact stop point |
| 7 | Teacher says "Line up for lunch" | Student is standing in line, facing forward | 45, 30, 90, 40, 55 s (45 s) | "In line" is defined as a position, not an attitude |
| 8 | Visual schedule card is flipped to "Work" | Student sits at the work table with the bin open | 3:00, 2:10, 4:30, 1:45, 2:20 (2:20) | Bin open marks readiness, not just arrival |
| 9 | Bell at the end of passing period | Student crosses the classroom doorway | 0, 40, 75, 20, 0 s after the bell (20 s) | Late arrival is timed from the bell, so 0 is on time |
| 10 | Recess whistle | Student joins the class line by the door | 2:30, 1:50, 3:40, 2:00, 1:20 (2:00) | Times outdoor transitions apart from indoor ones |
Transitions often need a second number. If the student arrives in 60 seconds but needs 3 adult prompts on the way, latency alone hides the support. Add a prompt count column, coded with the levels in the prompt hierarchy examples.
Weak: "Transitions within a reasonable time." Reasonable is not a number. Strong: "Timer starts at the end-of-centers chime; stops when Ava sits at her carpet spot. Record prompts given while timing."
Following directions examples (11 to 15)
| # | Timer starts | Timer stops | Sample session of 5 directions | What makes it strong |
|---|---|---|---|---|
| 11 | Adult finishes a one-step direction ("Get your folder") | Student begins moving toward the folder | 4, 12, 7, 20, 5 s (median 7 s) | Short latencies need a stopwatch, not a wall clock |
| 12 | Adult finishes a two-step direction | Student begins step 1 | 9, 15, 6, 30, 11 s (median 11 s) | Times the start only; a separate accuracy note covers both steps done |
| 13 | Coach finishes "Start your warm-up laps" | Student starts running | 25, 40, 18, 60, 22 s (median 25 s) | Same cue in PE as in class lets the team compare settings |
| 14 | Job coach finishes "Fold these 10 towels" | First towel is picked up | 35, 50, 20, 45, 30 s (median 35 s) | Vocational starts are timed the same way as academic ones |
| 15 | Para finishes "Put your phone in the basket" | Phone leaves the student's hand | 15, 70, 25, 120, 40 s (median 40 s) | Shows when a demand is followed, without judging tone |
Directions vary in difficulty, so log which direction each time belongs to. A student who starts "get your folder" in 4 seconds and "start your essay" in 4 minutes has a task problem, not a listening problem.
Weak: "Follows directions promptly 80% of the time." That mixes speed with yes or no. Strong: "Begins the first step of a one-step direction within 10 seconds; record seconds per direction, 5 directions per session."
Communication and return-to-work examples (16 to 20)
| # | Timer starts | Timer stops | Sample data | What makes it strong |
|---|---|---|---|---|
| 16 | Peer finishes "Hi, Leo" | Leo says or signs a greeting back | 6, 4, 10, 3, 8 s (median 6 s) | The peer's greeting is the cue, so the data reflects real social speed |
| 17 | Teacher finishes a yes or no question | Student gives an answer by voice or device | 12, 8, 15, 5, 9 s (median 9 s) | Counts device answers, so the goal measures response time, not speech |
| 18 | SLP finishes a picture-naming prompt | First sound of the target word | 3, 6, 4, 2, 5 s (median 4 s) | Word retrieval speed can show progress before accuracy changes |
| 19 | Break timer ends and the teacher says "Back to work" | Student writes or types again | 4:00, 2:30, 5:10, 3:20, 2:00 (3:20) | Captures a different problem from starting fresh work |
| 20 | Student returns from the calm-down area | Student is seated and working on the task left behind | 1:30, 3:00, 0:50, 2:10, 1:10 (1:30) | Shows whether breaks restore work or end it |
Example 20 matters for behavior plans. A break that leads back to work in 90 seconds is doing its job. A break followed by 15 minutes of not working is a different story, and the data shows it.
Weak: "Responds to peers when greeted." That is yes or no, and it hides a 30-second pause. Strong: "Timer starts when the peer finishes the greeting; stops at Leo's first word or sign back. Record seconds, 5 greetings per day."
Reading the week of data
Take example 1. Mason's times were 4:10, 6:30, 3:45, 5:20 and 2:50. Sorted, that is 2:50, 3:45, 4:10, 5:20, 6:30, so the median is 4:10. The mean is about 4:31, pulled up by the 6:30 day. Report the median and the range.
Graph one point per week. If the goal is "within 2 minutes in 4 of 5 work periods," also count how many periods met 2 minutes. Mason met it in 0 of 5 this week. Both views belong on the goal graph: the median shows gradual change, and the count shows the criterion.
The IEP must say how progress toward each goal will be measured.[1] Latency goals should name the cue, the response and the time limit, so the data sheet and the goal match.
Common latency recording errors
- Timing from the start of the cue. A 20-second direction adds 20 seconds to every trial. Start when the cue ends.
- Stopping at "looks ready". Pick a visible action: first mark, first word, seated at the spot.
- No ceiling. Set 5 or 10 minutes, record it as a ceiling score, then prompt.
- Changing the cue. "Start your work" one day and a written card the next is two different measures.
- Averaging a skewed week. Use the median when one day is far off the rest.
- Prompting without noting it. A prompt at 3 minutes means the latency after it was not independent.
Frequently asked questions
What is the difference between latency and duration recording?
Latency measures how long it takes to start a response after a cue. Duration measures how long the response lasts once it starts. A student who takes 6 minutes to start writing has a latency problem; a student who starts at once but stops after 2 minutes has a duration problem.
Should I use seconds or minutes for latency?
Use seconds for anything under about 3 minutes, such as following directions or transitions. Use minutes and seconds for longer starts, such as independent work. Keep the same unit across the whole goal so the graph is readable.
What do I record if the student never starts?
Set a ceiling before you begin, such as 10 minutes. If the student has not started by then, record the ceiling with a plus sign, prompt the student, and note the prompt. Do not leave the box blank.
How many latency trials do I need each week?
Aim for 3 to 5 timed opportunities a week for slow-changing skills like starting work. For directions or greetings, time 5 to 10 in one session. You need enough points to take a median that one odd day cannot swing.
Use it in IEP Atlas
- Task initiation goals (executive function)
- Latency recording measurement guide
- Latency data sheet
- Latency recording in the IEP glossary
- School transitions goals
Related guides
- IEP data collection: methods, sheets and schedules that hold up
- Frequency vs duration recording: which one fits the behavior
- How to choose a data collection method for an IEP goal
- How to take trial-by-trial data during instruction
- Task analysis examples for life skills goals
- Task analysis for washing hands (printable steps)
- Task analysis for brushing teeth (printable steps)
Sources
- 34 CFR §300.320: Definition of individualized education program, Legal Information Institute, Cornell Law School
- Progress Monitoring: Data & Goal Setting, National Center on Intensive Intervention
Written and checked by the IEP Atlas editorial team under our editorial policy. This guide describes general practice and federal rules. It is not legal advice; your state may add requirements, and the IEP team makes the decisions for each student.