What is the TNL-2? The Test of Narrative Language explained

The TNL-2, or Test of Narrative Language, Second Edition, is a "norm-referenced" test (one that compares a child's performance with a representative sample of same-age peers) that measures how well children ages 4 through 15 understand and tell stories. Published by PRO-ED in 2017 and written by Ronald B. Gillam and Nils A. Pearson, it is administered one-on-one by a speech-language pathologist or another professional trained in standardized testing, and it takes about 15 to 20 minutes. It is not an IQ test. It measures a specific language skill, "narrative discourse," that intelligence tests do not sample, and clinicians typically use it alongside oral-language batteries and, where relevant, cognitive testing. This article explains what the TNL-2 asks children to do, how it is scored, what its norms look like, and how it relates to IQ testing.

According to PRO-ED, the TNL-2 measures "children's narrative language abilities (i.e., children's ability to understand and tell stories)." The publisher's rationale is that narration is an important part of spoken language that most oral-language tests do not measure directly, and that it provides a foundation for later literacy. Understanding a story requires a child to follow events, track characters and their goals, and draw inferences the storyteller never states. Telling one requires organizing all of that into connected sentences a listener can follow.

The publisher describes the TNL-2 as "a natural complement" to standardized tests that assess pieces of oral language in more contrived formats, such as pointing to pictures or repeating sentences. PRO-ED also says the test is "especially useful for diagnosing language-based learning disabilities," which is one reason it appears in school evaluations and in research on developmental language disorder.

Because the test asks children to speak at length, it is administered individually by a trained examiner, usually a speech-language pathologist, though school psychologists and educational diagnosticians with appropriate training also use it. PRO-ED lists it as a Level B instrument, meaning purchasers must document relevant credentials. Diagnosis of a language disorder remains the job of a qualified clinician who interprets the TNL-2 together with other evidence.

The TNL-2 uses a picture book, an examiner's manual, and a record booklet. The test alternates between comprehension and production tasks within each of three story formats. According to a 2022 study of the test published in Language, Speech, and Hearing Services in Schools by Beula Magimairaj and colleagues at Utah State University and the University of Texas at Austin, "alternating Comprehension and Production subtests provides children with adult narrative models followed by literal and inferential questions, and then story retell or story creation tasks." The three formats are:

• Script: The examiner tells a familiar, everyday story about children going to McDonald's with their mother, supported by a single picture. The child answers comprehension questions, then retells the story. In the second edition, the publisher added a picture to this story so that every task has the same kind of visual support.

• Personal narrative: A sequence of five pictures supports a story about a shipwreck. After answering comprehension questions, the child looks at a different five-picture sequence (a boy who is late for school) and tells a story of their own.

• Fictional narrative: A single scene supports a story about a dragon. The child answers questions about it, then invents a story from a single picture of aliens.

The research team's summary of the test lists comprehension items worth 20, 14, and 13 points across the three formats, while each production task is scored on two dimensions, "story content" (what the child included) and "story complexity" (how the story is structured). One of the practical selling points, according to PRO-ED, is that the scoring system "does not require clinicians to transcribe the stories," which the publisher says saves hours of transcription time compared with traditional language-sample analysis.

The TNL-2 yields two subtest scores, one for narrative comprehension and one for oral narration (production), plus a composite called the "Narrative Language Ability Index." A University of Kansas study registration on ClinicalTrials.gov describes the composite this way: "each child's subtest scaled scores are summed and converted to the narrative language ability index," and the index "ranges from 50 (less than 1st percentile) to 155 (greater than 99th percentile)." Higher scores mean stronger narrative skills.

In other words, the TNL-2 uses the same family of standard scores and "percentiles" (the percentage of same-age children scoring at or below a given point) that readers of IQ reports will recognize, and a child's score describes where their storytelling and story comprehension fall relative to same-age peers.

Researchers have used specific cut points on the index. In a 2023 study of children with developmental language disorder, a team led by Karla McGregor reported that a standard score of 92 "is the cutpoint that maximizes the sensitivity and specificity" of that diagnosis on the TNL-2; in their sample, children with the disorder scored between 61 and 91 on the composite while typically developing children scored between 94 and 127. A 2025 intervention study by Gordon and colleagues likewise used a narrative index of 91 or lower as evidence of a language deficit. These are research conventions, and any clinical decision should rest on the full evaluation rather than a single number.

PRO-ED reports that the TNL-2 norms are based on new standardization data collected between 2013 and 2015, that the sample is larger than the original TNL sample with "many more children at each age," and that its demographic characteristics match those reported by the U.S. Census Bureau. The publisher does not state the sample size on its product page, so readers who need that figure should consult the examiner's manual. For a longer discussion of why a representative norm sample matters, see why a norm sample matters for IQ test accuracy; the same logic applies to a language test.

On reliability, Magimairaj and colleagues report from the manual that "coefficient alpha" (a measure of internal consistency) is .81 for the Comprehension subtest and .87 for the Production subtest in the normative sample. In their own sample of 357 first- through fourth-graders in Utah and Texas at risk for language and literacy difficulties, values ranged from .75 to .90. Their study also found that administering the TNL-2 over video conference, with permission from the publisher during the COVID-19 pandemic, produced psychometric properties that "were not significantly different from in-person testing." That said, standard administration remains in person with the printed picture book.

The second edition changed the original TNL in several ways. According to PRO-ED, the age range widened from 5 through 12 to 4 through 15, the pictures were redrawn in color, more inferential comprehension questions were added along with more items for the youngest and the oldest examinees, and the scoring system for the oral narratives was made consistent across the three story types. The publisher also says the manual now includes a diagnostic accuracy study and an expanded validity chapter. One note for careful readers: PRO-ED's revision notes and the distributor WPS give the range as 4 years 0 months to 15 years 11 months, while a separate line on the PRO-ED page says 5-0 through 15-11; the peer-reviewed literature describes the test as standardized for ages 4 through 15.

The TNL-2 does not produce an IQ score, and it is not designed to estimate general intelligence. Narrative language draws on vocabulary, grammar, memory, and reasoning, so it is correlated with the verbal side of intelligence tests, but it is a narrower construct. A comprehensive oral-language battery such as the TOLD-I:5 or the CELF-5 samples several language components in structured formats; the TNL-2 adds the discourse level, where children have to hold an entire story together.

Cognitive testing often sits alongside the TNL-2 in evaluations. In the Magimairaj study, for example, every participant also took the Kaufman Brief Intelligence Test, Second Edition, and only children with nonverbal IQ standard scores above 60 were included. That pairing illustrates a common clinical logic: a nonverbal ability measure helps an evaluator judge whether a low narrative score reflects a specific language weakness or a more general cognitive difficulty.

For adults, the picture is different. There is no adult version of the TNL-2. Adults who want a measure of their reasoning ability can take the RIOT IQ test, the Reasoning and Intelligence Online Test, developed by RIOT IQ with psychometrician Dr. Russell T. Warne. It is designed for people 18 and older, uses 15 subtests across six cognitive indices (verbal reasoning, fluid reasoning, spatial ability, working memory, processing speed, and reaction time), takes about 52 minutes, and reports scores on the familiar mean-100, standard-deviation-15 scale. It is an online ability measure, not a substitute for an individually administered diagnostic evaluation.

• Verbal output is required: Because the production tasks depend on extended speech, the TNL-2 is not appropriate for children who are nonverbal or whose speech is highly unintelligible, and examiners should consider articulation and fluency when interpreting stories.

• A single session is a snapshot: As with any standardized test, a score carries measurement error. Fatigue, shyness, or unfamiliarity with the examiner can lower a child's storytelling performance on a given day.

• English-language norms: The published norms describe performance in English. Children who are still acquiring English or who speak a different dialect at home may tell stories that are organized differently, and clinicians need to weigh that context.

• Scoring requires training: Although transcription is not required, judging story content and complexity consistently takes practice, and research teams typically check interrater agreement before using the scores.

The TNL-2 is standardized for children from 4 years 0 months through 15 years 11 months, according to the publisher's revision notes, distributor listings, and the peer-reviewed literature. The original TNL covered ages 5 through 12.

PRO-ED lists the testing time as 15 to 20 minutes. It is administered individually.

No. It measures narrative language, meaning the ability to understand and tell stories. It does not produce an IQ score. Evaluators often pair it with an intelligence or nonverbal ability measure to interpret a low score in context.

It gives scaled scores for narrative comprehension and oral narration and a composite Narrative Language Ability Index. The index runs from 50 to 155, with percentile ranks available for comparison with same-age peers.

Professionals with training in standardized assessment, most often speech-language pathologists. PRO-ED sells it as a Level B instrument, which requires documented qualifications to purchase.

According to PRO-ED, the second edition has a wider age range, new color pictures, more inferential comprehension questions, a consistent scoring format across story types, a larger norm sample collected in 2013 to 2015, and an expanded validity chapter that includes a diagnostic accuracy study.

1. PRO-ED. (2017). TNL-2: Test of Narrative Language–Second Edition, Complete Kit [Product page]. proedinc.com

2. Western Psychological Services. (n.d.). (TNL-2) Test of Narrative Language–Second Edition [Distributor page]. wpspublish.com

3. PAR, Inc. (n.d.). Test of Narrative Language–Second Edition (TNL-2) [Distributor page]. parinc.com

4. Magimairaj, B. M., Capin, P., Gillam, S. L., Vaughn, S., Roberts, G., Fall, A.-M., & Gillam, R. B. (2022). Online administration of the Test of Narrative Language–Second Edition: Psychometrics and considerations for remote assessment. Language, Speech, and Hearing Services in Schools, 53(2), 404–416. files.eric.ed.gov

5. McGregor, K. K., Ohlmann, N., Eden, N., Arbisi-Kelm, T., & Young, A. (2023). Abilities and disabilities among children with developmental language disorder. Language, Speech, and Hearing Services in Schools, 54(3), 927–951. pmc.ncbi.nlm.nih.gov

6. Gordon, K. R., Moss, D., Swinburne Romine, R. E., Fleming, K. K., & Storkel, H. L. (2025). Interactive book reading to accelerate word learning by kindergarten children with developmental language disorder: Incorporating retrieval-based practice into training. Language, Speech, and Hearing Services in Schools, 56(4), 1110–1125. pmc.ncbi.nlm.nih.gov

7. University of Kansas. (2025). Kindergarten children acquiring words through storybook reading (ClinicalTrials.gov Identifier NCT03586479). clinicaltrials.gov

Hero photo: Story dice with picture symbols on a wooden table. Photo by joho345, via Wikimedia Commons, CC BY-SA 4.0 (cropped).