Test of Language Development, Second Edition

Purpose: Designed to assess language development in children.

Population: Ages 4-12.

Score: Five types of scores.

Time: (50-65) minutes.

Authors: Donald D. Hammill and Phyllis L. Newcomer.

Publisher: PRO-ED.

Description: The rationale for the Test of Language Development (TOLD) as reported by the authors serves four purposes: 1) to identify children who are significantly below their peers in language, 2) to determine children's specific strengths and weaknesses, 3) to document children's language progress as a consequence of special intervention programs, and 4) to serve as a measurement device in research involving language behavior.

Scoring: The test is available in two editions: Primary (TOLD-P) and Intermediate (TOLD-I). The TOLD-Pís seven subtests are: Picture Vocabulary, Oral Vocabulary, Grammatic Understanding, Sentence Imitation, Grammatic Completion, Word Discrimination, and Word Articulation. The TOLD-Iís five subtests are: Sentence Combining, Characteristics, Word Ordering, Generals , and Grammatic Comprehension. Both the TOLD-P and TOLD-I yield five different types of scores: raw scores, language ages, percentiles, standard scores, and quotients for composite scores.

Reliability: On the TOLD-P, two studies report split-half coefficients greater than .80 for all subtests except Picture Vocabulary, where coefficients ranged between .61 and .72. Similarly, low coefficients were obtained for two subtests, Oral Vocabulary and Grammatic Understanding, at the five- and eight-year-old groups. For the TOLD-I, internal consistency coefficients were greater than .80. Test-retest reliability: For the TOLD-P, 21 children were tested twice over a period of five days; on the TOLD-I, 30 children were tested after one week. In both cases, Pearson product-moment coefficients were computed on the raw scores, with coefficients being above the .80 level for all measures.

Validity: For criterion-related validity, the authors compared the two TOLD tests to existing criterion measures. The resulting correlation coefficients between the TOLD-I and the Test Of Adolescent Language were adequate. For the TOLD-P several of the coefficients were quite low. For example, the Grammatic Understanding subtest had a correlation coefficient of .13 at the four-year-old level and .47 at the eight-year-old level when compared to the Northwestern Syntax Screening Test's Reception subtest..

Norms: The TOLD-P was standardized on 1,836 children, with 198 subjects being represented at the smallest age level, whereas the TOLD-I was standardized on 871 children, with 166 being the least number per age level. Subjects were selected from all regions of the United States and controlled for sex differences, residence, race, geographic area, an occupation of parents.

Suggested Uses: Recommended uses include language assessment in clinical, educational, and research settings.