What is the TWF-3? The Test of Word Finding explained

The TWF-3, short for the Test of Word Finding, Third Edition, is an individually administered test of how accurately and how quickly a child can retrieve words the child already knows. A speech-language pathologist or similarly qualified examiner gives it one-on-one to children aged 4 years 6 months through 12 years 11 months, and it yields a single standard score called the Word Finding Index. Written by Diane J. German and published by PRO-ED in 2015, it measures one narrow expressive-language skill. It is not an IQ test, and it is not a vocabulary test either. A child can know a word perfectly well and still fail to produce it on demand, and that gap is what the TWF-3 is built to detect. This article explains what "word finding" means, how the test works, and how it differs from tests of vocabulary and general intelligence.

"Word finding" is the act of pulling a specific word out of memory at the moment you need it. Everyone has occasional failures, the tip-of-the-tongue moment when a name refuses to surface. For some children the failures are frequent enough to interfere with speaking, answering questions in class, and reading aloud. Researchers Julie Dockrell, David Messer, and colleagues, in a 1998 survey of language services in South-East England, defined a word-finding difficulty as being present when a child can pick out the right picture from a set of choices but has trouble producing the word when shown the picture or when talking. They reported that 23 percent of children receiving language support had word-finding difficulties.

The same survey described the everyday signs: long pauses inside phrases, heavy use of fillers such as "uh" and "um," and "circumlocution," which means talking around a word ("the thing you cut paper with") instead of saying it. In a 2006 review, Messer and Dockrell summarized the leading explanations, which point to the meaning system, the sound system, and processing speed as the places where retrieval can break down. A widely cited model of speech production by Willem Levelt and colleagues describes word production as a chain of steps from concept to word to sound form, and a word-finding error can occur at any link in that chain.

According to PRO-ED, the TWF-3 is "a norm-referenced, single-word expressive language test expressly designed to assess children's word-finding ability." "Norm-referenced" means a child's performance is compared with that of a large sample of same-age children rather than against a fixed pass mark.

The test comes in three forms. PRO-ED describes a Preprimary form for preschool and kindergarten students, a Primary form for Grades 1 and 2, and an Intermediate form for Grades 3 through 6. Each form contains the same four naming sections, and the results are combined into the overall Word Finding Index.

• Picture Naming: Nouns: the child names pictured objects. PRO-ED describes the targets as compound words and words of one to four syllables.

• Sentence Completion Naming: the examiner reads a sentence aloud with the final word missing and the child supplies it, so the cue is meaning and grammar rather than a picture.

• Picture Naming: Verbs: the child names pictured actions, including present-tense and past-tense forms of regular and irregular verbs, so the child must retrieve both the right word and the right form.

• Picture Naming: Categories: the child names pictured objects and then the category those objects belong to, which taps retrieval at two levels of specificity.

Two design choices set the TWF-3 apart from a simple naming task. The first is the "Comprehension Check." After the naming sections, the examiner returns to each item the child missed and checks whether the child understands the word. The User's Guide states the logic plainly: a naming error cannot be counted as a word-finding error unless the examiner is reasonably sure the child knows the target word. The second is a set of five informal analyses: the Phonemic Cueing Procedure, the Imitation Procedure, and the Substitution Analysis examine what kinds of errors the child makes, while the Delayed Response Procedure and the Secondary Characteristics Tally help interpret the index score.

The TWF-3 is given individually, and PRO-ED lists a testing time of 20 to 30 minutes. It is sold at Test Level B, which means the examiner needs training in standardized test administration and interpretation, and it is used mainly by speech-language pathologists.

The distinctive feature of scoring is that it counts speed as well as accuracy. According to the User's Guide, the Word Finding Index is "a standard score based on a student's accuracy and speed in retrieving single words." On the record form, a response is scored as accurate when the correct word is produced within four seconds. A correct answer that arrives after four seconds or more is recorded as a delayed response.

Raw scores from the four sections are combined and converted to the Word Finding Index with a percentile rank. The index uses the familiar scale on which 90 to 109 is labeled Average, 80 to 89 Below Average, 70 to 79 Weak, and below 70 Very Weak, with matching labels on the high side. The publisher reports a cut score of 90 for identifying word-finding difficulty. The User's Guide adds a safeguard: the index should be interpreted only when the child has demonstrated understanding of most of the words, because a low naming score on unknown words is a vocabulary result rather than a word-finding result.

PRO-ED also sells a computer-administered version, the TWF-3: CA, which selects the form by age, records accuracy, delay, comprehension, and phonemic-cueing scores, and calculates the standard score and percentile rank. The examiner still sits with the child and judges each response; the software handles the timing and the arithmetic.

PRO-ED reports that new normative data were collected between 2009 and 2012. According to the User's Guide, the standardization sample consisted of 1,283 children in 23 states, matched to 2011 U.S. Census figures for region, gender, race, ethnicity, exceptionality, and parent education. The publisher notes that children with diagnosed word-finding difficulties were included, which matters because a test meant to identify a condition should be normed on a population that contains it.

Reliability, the consistency of scores, is reported at a level appropriate for a diagnostic tool. The User's Guide gives internal-consistency coefficients of .85 to .90, test-retest coefficients of .74 to .86 with a combined value of .81, and interscorer agreement of .99. In plain terms, two examiners scoring the same child agree almost perfectly, and a child tested twice will usually land in a similar place, though the test-retest figures leave room for some movement in individual scores.

For validity, PRO-ED reports "large or very large" correlations between the Word Finding Index and other expressive language measures. The publisher also reports diagnostic accuracy statistics: a sensitivity of .86, a specificity of .78, and an area under the ROC curve of .88 at the cut score of 90. "Sensitivity" is the proportion of children with a known word-finding difficulty whom the test flags; "specificity" is the proportion of children without the difficulty whom the test correctly clears. The test therefore catches most true cases while missing some, and clears most typical children while flagging roughly one in five who do not have the difficulty, which is why clinicians combine the index with the informal analyses and other evidence before drawing conclusions. PRO-ED also reports item-bias studies supporting the absence of racial and gender bias.

A vocabulary test asks how many words a child knows. Receptive vocabulary tests have the child point to a picture that matches a spoken word, and expressive tests such as the EVT-3 ask the child to name pictures, usually without a time limit. Because word knowledge grows with age and reading, vocabulary tests are among the strongest single indicators of verbal ability. The TWF-3 turns that logic around: it confirms that the child knows the words and then scores how efficiently the child retrieves them. A child with an average vocabulary and a low Word Finding Index has a retrieval problem rather than a knowledge problem, and the two call for different kinds of help.

An IQ test measures something broader: the ability to reason, solve novel problems, hold and manipulate information, and process it quickly across verbal and nonverbal content. The TWF-3 contains no reasoning items and should never be read as an estimate of general intelligence. In a 2011 study, Messer and Dockrell found that children with word-finding difficulties showed very slow naming even though their nonverbal abilities were often within the typical range.

The two kinds of tests work as companions. A comprehensive evaluation often pairs a cognitive measure, to establish general ability, with a language measure such as the TWF-3 to describe the specific skill in question. When a child's Word Finding Index sits well below the child's reasoning scores, that pattern points toward a specific retrieval weakness rather than a global one. For adults, a similar contrast is drawn with confrontation naming tasks such as the Boston Naming Test. PRO-ED publishes a separate adolescent and adult version, the TAWF-2, for examinees beyond the TWF-3 age range.

Naming speed also has a documented relationship with reading. Elizabeth Norton and Maryanne Wolf, reviewing the rapid automatized naming literature in 2012, describe how slow naming of familiar items predicts reading fluency problems. The TWF-3 measures single-item naming rather than the serial naming used in those tasks, and Messer and Dockrell found the two only loosely related, so a TWF-3 score is not a substitute for a reading evaluation.

The TWF-3 answers one specific question well: does this child struggle to retrieve words the child already knows, and if so, how? No diagnostic label should be attached to a child on the strength of one score; diagnosis of a language disorder belongs to a qualified clinician who integrates test results with history, observation, and other measures.

For adults who want a broad, norm-referenced measure of cognitive ability, the RIOT IQ test, the Reasoning and Intelligence Online Test, was developed by RIOT IQ with psychometrician Dr. Russell T. Warne for people aged 18 and older. It includes 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 same mean-100, standard-deviation-15 scale used by the major clinical batteries. It does not replace an individually administered diagnostic evaluation of the kind described here. You can learn more or take the test at riotiq.com.

It measures word-finding ability, meaning how accurately and how quickly a child can produce words the child already knows, across four naming tasks. The result is a single standard score called the Word Finding Index.

PRO-ED lists the age range as 4 years 6 months through 12 years 11 months, with Preprimary, Primary, and Intermediate forms for different grade levels. Older examinees are covered by the separate TAWF-2.

No. It contains no reasoning items and measures one expressive-language skill. Research shows that children with word-finding difficulties often have nonverbal abilities in the typical range.

A vocabulary test measures how many words a child knows. The TWF-3 first confirms the child knows the target words through its Comprehension Check and then scores how efficiently the child retrieves them, including whether a correct answer arrived within four seconds.

PRO-ED lists 20 to 30 minutes for the standardized portion. The Comprehension Check and optional informal analyses add time depending on how many items the child missed.

A trained examiner, most often a speech-language pathologist, gives it one-on-one. PRO-ED sells it at Test Level B, which requires training in standardized assessment.

1. PRO-ED. (n.d.). TWF-3: Test of Word Finding, Third Edition, Complete Kit. proedinc.com

2. German, D. J. (2015). Test of Word Finding, Third Edition: Computer Administered. User's guide. PRO-ED. proedinc.com

3. Dockrell, J. E., Messer, D., George, R., & Wilson, G. (1998). Children with word-finding difficulties: Prevalence, presentation and naming problems. International Journal of Language and Communication Disorders, 33(4), 445-454. pubmed.ncbi.nlm.nih.gov

4. Messer, D., & Dockrell, J. E. (2006). Children's naming and word-finding difficulties: Descriptions and explanations. Journal of Speech, Language, and Hearing Research, 49(2), 309-324. pubmed.ncbi.nlm.nih.gov

5. Messer, D., & Dockrell, J. (2011). Lexical access and literacy in children with word-finding difficulties. International Journal of Language and Communication Disorders, 46(4), 473-480. pubmed.ncbi.nlm.nih.gov

6. Levelt, W. J. M., Roelofs, A., & Meyer, A. S. (1999). A theory of lexical access in speech production. Behavioral and Brain Sciences, 22(1), 1-38. doi.org

7. Norton, E. S., & Wolf, M. (2012). Rapid automatized naming (RAN) and reading fluency: Implications for understanding and treatment of reading disabilities. Annual Review of Psychology, 63, 427-452. doi.org

Hero photo: Picture-naming cards laid out on a table during a language session. Photo by ABDEM, via Wikimedia Commons, CC BY-SA 4.0 (cropped).