Response magnitude

Magnitude describes the size of a response

Count records how many responses occur. Duration records how long they last. Magnitude records a physical extent or an anchored level. A response can occur once with high magnitude or many times with low magnitude, so frequency and magnitude answer different questions.

Select magnitude when response size is important to safety, access, function, comfort, or progress. Keep other relevant dimensions alongside it rather than forcing every change into one number.

Prefer a direct physical measure when feasible

Examples include sound pressure in decibels, applied force in newtons, distance in centimeters, mass in grams, or electrical output from an appropriate instrument. The chosen unit should match the response and the practical decision.

Document the device, model, settings, placement, sampling rate, calibration method, and valid range. Sound measured one meter away cannot be compared casually with sound measured across a room. A force sensor can saturate and display the same maximum for several different events.

Use equipment within its intended purpose. A behavior team should obtain qualified help for clinical interpretation of physiological, medical, biomechanical, or safety data outside its scope.

Anchored scales need observable boundaries

When direct measurement is unavailable, a rating scale can define levels through observable features. “Mild, moderate, severe” invites subjective judgment. A stronger scale might use distance moved, visible damage, assistance required, or a calibrated device range.

Each level needs inclusion examples and close nonexamples. Train observers using the same kinds of settings, angles, noise, clothing, and materials found in actual use. Sample agreement at boundary cases, since those are where drift is likely.

A practice-focused measurement article discusses physical dimensions and rating scales as ways to capture intensity. It emphasizes matching the measure to the question. Its framework remains a proposal rather than a universal protocol.

Define the opportunity and scoring window

State what counts as one episode, when it begins and ends, and whether the highest, average, total, or first value represents it. If several readings occur during one response, using each as a separate episode inflates the denominator.

Define eligible opportunities before collection. Report device failures, blocked views, and events outside the calibrated range separately. Missing values do not become zeros.

Threshold measures need both sides of the fraction. “Three high-magnitude events” means little without the number of measured episodes and the threshold source. A threshold chosen for equipment protection differs from one selected for a clinical question.

A fictional community example

Tessa is a fictional teenager who wants a quieter community recreation room. The team measures naturally occurring room sound during 12 scheduled transitions with a calibrated meter in a fixed location. It defines three environmental bands before observation: below 65 dB, 65 through 74 dB, and 75 dB or higher.

Four transitions fall below 65 dB, five fall in the middle band, and three reach 75 dB or higher. The high-band proportion is 3 of 12, or 25%. These are room-sound observations, not measures of Tessa's behavior.

Tessa uses an agreed rating after each transition and identifies all three high-band transitions as uncomfortable. The team changes room setup and schedule. It never recreates a loud event for testing. Later results can be compared only with the same device placement, bands, and transition definition.

Protect safety, consent, and communication

Do not provoke impact, loud sound, self-injury, aggression, pain, feeding risk, or another hazardous event to obtain a measurement. Immediate safety and medical response take priority over completing a score.

Keep communication and AAC, food, water, bathroom access, mobility, prescribed care, breaks, and emergency help available. Explain the measurement in an accessible way, obtain required consent and assent when applicable, and respond to withdrawal or distress under the governing plan.

When equipment records audio, video, location, biometrics, or health information, verify privacy, security, retention, access, and recording requirements before use.

Report magnitude without overstating meaning

Provide the unit, numerator, denominator, range, median or mean when appropriate, threshold counts, device, setting, and time window. Show frequency and duration when they matter. A lower average can hide one extreme event, while a maximum can hide improvement in most episodes.

Keep the raw measurement and any category derived from it. If the threshold changes, analysts can recalculate the category without inventing a new observation. Version the rule and state which records were reclassified.

The BACB BCBA Test Content Outline, Sixth Edition identifies response magnitude among measurement dimensions and covers reliability, validity, and representative procedure selection. It does not set a universal scale or safe threshold.