Speech sample count calculator

How to use this tool

Use this speech sample calculator to practice summarizing separately scored consonant accuracy and understood-word counts.

• Describe the sample context, language, scoring rule and denominator.
• Enter clinician-scored counts, keeping consonants and understood words in their own columns.
• Select a compatible series and review numerator, denominator and missing measures before exporting.

Fictional example and how to read the result

Fictional example: 18 correct consonants out of 20 scored consonants is 90%. This is a different measure from 18 understood words out of 20 words.

These arithmetic summaries contain no norms or diagnostic conclusions. Scoring and interpretation require a qualified speech-language professional.

Score first, then review the arithmetic

Enter counts scored by a qualified clinician. Consonant accuracy and intelligibility answer different questions and remain separate measures. This tool does not transcribe speech, apply developmental norms or assign severity.

A compatible group has the same language, dialect, sample type, listener condition, setting and declared scoring rule. Pooling adds numerators and denominators within that group; it does not average unequal percentages. Zero denominators remain not calculable.

• Define which consonants or words are eligible before collecting the sample.
• Enter both counts for each measure used; leave both blank for a measure not collected.
• Keep changed conditions in separate series. The graph uses equally spaced observations, not elapsed calendar time.
• Use anonymous sample labels; exported files remain on your device after clearing this page.

What to enter in each field

• Measure to graph: Select the percentage to graph: correctly scored consonants or words understood. These use different denominators and are not interchangeable.
• Chart series label (optional): Optional: enter an existing series label exactly to select its first compatible context for the graph. Leave blank to use the first context. All other contexts remain separate in the report.
• Observation date: Use the date of this observation. Imported dates must be real YYYY-MM-DD dates, for example 2026-10-01. Dates order the graph; points remain equally spaced.
• Series label: Use an anonymous label for repeated compatible speech samples, such as conversation A. A changed language, listener or scoring rule creates a separate context.
• Language: Name the language of the scored sample. Keep different languages in separate compatible groups.
• Dialect / language variety: State the dialect or language variety considered when scoring. Differences in variety are not automatically treated as errors.
• Sample type: Describe the elicitation type, such as conversation or a word sample. Only the same sample type is pooled.
• Listener condition: Describe the listener condition, such as familiar or unfamiliar listener, without names. It affects comparability of understood-word counts.
• Setting / context: State the sample setting, such as a quiet room. Changed settings stay in separate groups.
• Scoring rule / version: Record the clinician-defined eligibility and scoring version. This tool accepts already scored counts and does not transcribe or provide norms.
• Consonants correct: Enter the whole number of consonants scored correct by the clinician, no larger than consonants scored. Leave both consonant fields blank if that measure was not collected; 0 means none correct.
• Consonants scored: Enter the whole number of eligible consonants scored, including errors. A zero denominator is not calculable; leave both consonant fields blank when not collected.
• Words understood: Enter the whole number of words understood under the declared listener rule, no larger than words scored. Leave both word fields blank if uncollected; 0 is a recorded zero.
• Words scored: Enter the whole number of words scored under that rule. Words understood ÷ words scored gives the percentage; zero has no calculable percentage.

Original fictional example and data fields

Use the editable blank CSV or local import to collect your own de-identified entries. The example below is fictional.

Professional sources and permissions

Original worksheets and arithmetic support recording and planning. Follow the official publisher routes for standardized instruments.

Your browser, your entries

Entries are processed in this browser. These pages do not run marketing analytics or send entries to TherapyCRM. Download your work before leaving. Any optional device storage must be selected explicitly.