Choose range versus IQR ABA data summaries from the question and sample. Range shows the full observed span and responds to every extreme. IQR describes the middle-half spread after a stated quartile convention. Report both when an unusual value matters clinically. Neither summary replaces the ordered data, dates, context, measurement review, or client-level interpretation.

Choose the measure from the question

Use range when the full observed span matters. It answers how far apart the minimum and maximum are and keeps every extreme in the result. Use IQR when the question concerns the spread of the middle half under a declared quartile convention. The two measures describe different features, so the decision note should state the clinical or measurement question before the numbers are calculated.

Ximena's worksheet asks whether most observed rates are clustered while any unusual rate still deserves review. That question calls for both measures: IQR to describe the middle spread and range to preserve the full span. Choosing one after seeing which looks more favorable would conceal information.

Verify the sample and source data

Define the target, unit, phase, setting, observation rule, exposure, eligible dates, and analysis unit. Ximena's sample contains seven compatible rates per hour from one declared phase. Each rate remains linked to its numerator, observation duration, date, and source record. A raw count should not be mixed with a rate, and unequal observation periods require the reviewed rate calculation before spread is summarized.

Reconcile expected, observed, missing, excluded, duplicate, and corrected records. Sort a copy of the values while preserving source lineage. Investigate abrupt unit changes, default zeros, truncated observations, transcription errors, and values copied from a graph. Missing records can narrow both range and IQR, so report them instead of silently analyzing complete cases.

Calculate both measures for Ximena

The ordered rates are 4, 5, 5, 6, 6, 7, and 20 per hour.

For range:

  1. Minimum = 4 per hour.
  2. Maximum = 20 per hour.
  3. Range = 20 minus 4 = 16 per hour.

For IQR, use the median-of-halves convention and exclude the overall median from the two halves because the sample size is odd:

  1. Median = 6 per hour.
  2. Lower half = 4, 5, 5, so Q1 = 5 per hour.
  3. Upper half = 6, 7, 20, so Q3 = 7 per hour.
  4. IQR = 7 minus 5 = 2 per hour.

The arithmetic shows a narrow middle spread and a much wider full span. It does not explain the rate of 20.

Investigate the contrast without deleting it

Trace the 20-per-hour observation to its numerator, exposure, observer, definition, phase, setting, and contextual notes. Check whether it reflects a real event, changed opportunity, recording error, short observation, health or environmental condition, or another factor. NIST's outlier guidance supports investigation before deletion. Keep any correction or exclusion rule explicit and source based.

If the observation is valid, IQR still describes the middle half, while range ensures the unusual value remains visible. A rare event may carry more clinical or safety importance than the typical spread. If the observation is corrected, recalculate both measures and preserve the prior result and reason.

Handle small samples and edge cases

Quartile algorithms can differ, particularly with small or odd samples, so name the method and software version. With fewer observations, IQR may shift sharply when one value is added. With identical values, range and IQR are both zero, which does not prove measurement sensitivity or future stability. With one observation, range computes as zero but IQR is not a useful spread summary.

Predeclare the measure for recurring reports and retain both when the full span has decision value. Avoid switching from range to IQR solely because an extreme increases variability, or switching to range solely because it makes a phase look more changed.

Show both on the graph and report

Keep the time-series graph primary, with rates shown in date order and phase or context changes labeled. A box plot can display the quartiles and endpoints but removes temporal order. NIST's box-plot page supplies the related structure. A moving average, described by NIST, answers a smoothing question and should not replace either spread measure.

Across seven compatible observations, Ximena's rates ranged from 4 to 20 per hour, a range of 16. Using the median-of-halves convention, Q1 was 5, Q3 was 7, and the IQR was 2 per hour. The verified 20-per-hour observation remains visible for contextual and clinical review.

Include sample dates, missingness, unit, phase, method, and source version. Lobo and colleagues and Wolfe, Barton, and Meadan provide context for visual analysis beyond a single statistic.

Review meaning and access with the client

Review the pattern with Ximena through accessible communication. Maintain AAC, interpreters, mobility supports, food, water, bathroom access, prescribed care, rest, relationships, and emergency help during measurement and review. ASHA supports continuous access to AAC tools or devices.

Ask what the rate represents in everyday life, whether the unusual observation affected comfort or risk, and whether the chosen summary answers a useful question. Pair the spread measures with implementation evidence, observation conditions, client feedback, and qualified clinical judgment.

Keep professional sources in scope

The BACB ethics hub and CASP public summary offer professional and clinical context. The BCBA Test Content Outline lists measurement, display, and data-based decisions as examination content. NIST's distribution page distinguishes location, spread, and shape. None supplies a universal rule requiring range or IQR for an ABA decision.

Clinician review checklist

For a range versus IQR ABA data review, confirm the decision question, target and unit, exposure, phase, cohort, source records, missingness, ordered values, quartile convention, range arithmetic, Q1, Q3, IQR arithmetic, endpoint investigation, graph, report language, access, client perspective, implementation evidence, qualified decision owner, and review date. Lock the choice rule before recurring comparisons.

Limits of range and IQR

Neither measure shows time order, trend, measurement validity, treatment integrity, benefit, uncertainty, generalization, or cause. Range is driven by endpoints. IQR can hide clinically important extremes and varies by quartile method. A difference between range and IQR is a prompt to inspect the raw sequence, not proof of an outlier or stable performance. Use both as descriptive tools within a broader clinical review.

Related resources

Sources