Throughout my childhood (from age four to age nine), I was repeatedly given the Peabody Picture Vocabulary Test Revised. Each time the test was administered, I scored very low. Why? Being a quadriplegic with severe Cerebral Palsy grossly impaired my physically abilities to point with a finger and to speak. At a result, I could not communicate my answers to the testers. Instead of reporting that this test failed to measure my verbal skills, all the testers concluded that I had low intelligence. From personal experience, many severely physically disabled children, who received low scores on the Peabody Picture Vocabulary Test Revised from not being able to communicate their answers, were denied a public education. If my mother did not believe that I had above intelligence and advocate for my educational rights, I would never have received a high school diploma and Bachelors.
Psychological Test
During the 1950s and 1960s, Lloyd Dunn and Leota Dunn wrote the Peabody Picture Vocabulary Test, Peabody Individual Achievement Test, Peabody Language Development Kits, and Peabody Early Experience Kits (Gainer, 2000). Due to working with mentally retarded children, Lloyd Dunn became passionate in understanding and preventing developmental and learning problems (Gainer, 2000). Miller and Lee (1993) reported that the Peabody Picture Vocabulary Test was composed in 1959 and revised in 1981.
Function of Test
The Peabody Picture Vocabulary Test accessed children’s verbal vocabulary skills (Miller & Lee, 1993). Dunn and Dunn chose 3885 words from Webster’s New Collegiate Dictionary and his black-and-white drawings illustrated the words’ definitions.
From an initial pool of 3,885 words, 300 stimulus words and decoys were selected to form 150 plates; the same set of 150 plates was used to create both forms of the test, with each form comprising 150 stimulus words ordered in increasing difficulty (Miller & Lee, p. 438, 1993).
The revised edition of the Peabody Picture Vocabulary Test measured language ability (Hohnen & Stevenson, 1999) The object of the Peabody Picture Vocabulary Test Revised was “[to] measure receptive vocabulary and … used as a screener for verbal ability if the background of examiee is Standard American English” (Gainer, p. 2, 2000). This test was a nonverbal, multiple choice format, which was either written in Form M or Form L (Miller & Lee, 1993). The test only contained 144 instead of 300 words due to identifying research concerning item bias, the percentage curves (Miller & Lee, 1993). These words were taken from the Oxford English Dictionary, Webster’s Third New International Dictionary of the English Language Unabridged, and Merriam Webster Ninth Collegiate Dictionary. As a result, Dunn and Dunn made certain that children understood all test items. Miller and Lee (1993) stated that the revised edition was developed due to utilizing the Rasch-Wright Model of Latent Trait Analysis. The model was essential for making the curve, regarding verbal vocabulary skills, more accurate and for precisely calibrating test items based on level of difficulty (Miller & Lee, 1993). This design allowed children to be tested at any age.
Construction of Test
Miller and Lee (1993) postulated that there was very little evidence concerning validity claims. For example, content validity depended on the following assumption: Conducting a dictionary search offered a representative sample of the standard English vocabulary. “Evidence for criterion-related validity is cited in the concurrent validation of the PPVT-R against the PPVT, with coefficients ranging from .53 to .87, and … to predict scholastic aptitude” (Miller & Lee, p. 439, 1993). Overall, the Peabody Picture Vocabulary Test Revised was correlated with the McCarthy Scales of Children’s Abilities General Cognitive Index, Wechsler Full Scale Test, and Stanford-Binet Test; the coefficients ranged from .40 to .60 (Miller & Lee, 1993).
Also, hardly any evidence supported construct validity.
Claims for content validity are based on the illusive corollary that ensure[d] construct validity and on data demonstrating that scores on the PPVT-R ... increase with increasing age (Miller & Lee, p. 439, 1993).
Concurrent validity referred to intelligence and achievement measures (Hohnen & Stevenson, 1999). Even though the Rasch-Model was beneficial in calculating fairly exact measures, it contradicted the construct validity (Miller & Lee, 1993). This model was viewed as an interval scale of measurement. Without knowing the Nomological Theory, construct validity statements could not be declared. Although being aware of the importance of construct validity and test validation was a continuous process in which a specific test’s data provided a dependable source of information, the Peabody Picture Vocabulary Test Revised’s data supporting construct validation was used to access verbal vocabulary (Miller & Lee, 1993).
If a theory explaining why some words are learned before others can account for the ranking ordering of items in the PPVT-R, then PPVT-R scores can be assumed to reflect true developmental levels of vocabulary (Miller & Lee, p. 439, 1993).
Miller and Lee (1993) hypothesized that frequency of word usage, word length, and polysemy were great predictors concerning the placement order of words. This also represented a reliable source of data proved that construct validation was needed. Word length referred to an estimated number of vowels, syllables, and letters.
These variables [predictors] were transformed to natural logarithms. As some words had a frequency of zero in one or more of the frequency of the frequency norms, an arbitrary increment of 0.1 was added to frequency of each word to allow transformation (Miller & Lee, p. 439, 1993).
Once all the data and word order were recognized as single indicators, these indicators were set at 1.0 and the measurement was fixed at zero (Miller & Lee, 1993). This indicator / reference indicator was a latent variable scale, which was established by fixing a loading of 1.0 for each indicator variable. The number of syllables, the frequency count, and the number of meanings found in the Oxford English Dictionary were reference latent variables. “The correlations among the measured variables ... was used to obtain maximum likelihood of the parameters specified in the model using the corresponding covariance matrix” (Miller & Lee, p. 439-440, 1993).
Standardization
Peabody Picture Vocabulary Test Revised had adequate psychometric properties such as retest reliability, measures of internal consistency, and alternate-form reliability.
... measures of internal consistency range[d] from .88, and alternate-form from .71 to .91 based on the standardization sample. Retest reliability, as calculated on a subsample of the standardization sample, range[d] from .52 to .90 (Miller & Lee, p. 438, 1993).
Administration
According to Hohnen and Stevenson (1999), the Peabody Picture Vocabulary Test included different directions for children under age eight, children over age eight, adolescents, and adults. 828 adults and 4200 juveniles was the representative national sample (Hohnen & Stevenson, 1999). Completing this test took fifteen to twenty minutes. The protocol tells you where you should begin. The basal is the highest 8 consecutive responses, and the ceiling is the lowest consecutive responses containing 6 errors. Symbols beside each answer tell the examiner every eighth response (Hohnen & Stevenson, p. 591, 1999).
Barch and Berenbaum (1996) gave thirty-nine children, with Schizophrenia, the Peabody Picture Vocabulary Test Revised to obtain verbal intelligence measures. Only thirty-eight participants completed the test because one respondent was unable to take it. The Peabody Picture Vocabulary Test Revised’s mean score was 84.8, standard deviation score was 20.1, and range score was 59-133 (Barch & Berenbaum, 1996). In order to examine that the associations between thought disorder and language production resembled a relationship between verbal intelligence and thought disorder, the correlations between thought disorder and language production, and verbal scores were studied. The results indicated that the associations between thought disorder and language production did not resemble a relationship between verbal intelligence and thought disorder.
Senechal, Thomas, and Monker (1993) hypothesized that young juveniles had small or large vocabularies from examining their Peabody Picture Vocabulary Test Revised’s median split scores. Thirty-two children took Form L of the Peabody Picture Vocabulary Test Revised during a twenty-five to thirty minute session. Results suggested that reading at home significantly improved vocabulary knowledge. Lange, MacKinnon, and Nida (1989) administered the Peabody Picture Vocabulary Test Revised to ninety-three children ages three and four. Results illustrated that verbal object recall depended on engaging in mastery-related behaviors.
Hohnen and Stevenson (1999) administered the Peabody Picture Vocabulary Test Revised to 126 children between ages of six and seven. The results from the multivariate analysis support the view that there is a single underlying dimension of individual difference that is genetically mediated and is specific to verbal skills. This ability is independent of general intelligence and influences of general language ability, phonological awareness, and literacy (Hohnen & Stevenson, p. 601, 1999).
Wright, Huston, Reitz, and Piemyat (1994) also gave the Peabody Picture Vocabulary Test to children between ages of five and seven. It cognitively assessed participants’ television reality perceptions. Low perceived reality scores were linked with high testing scores. This meant that responded were expected to comprehend that that television fiction was not real. In addition, high testing scores indicated that subjects did not believe adult shows were unscripted and adult fiction, television characters had the same jobs in real life. Nevertheless, thinking educational television shows occurred in real life was associated with high testing scores.
Both age and PPVT-R were positively to each score, but PPVT-R was significant only for news segments. That is, children with high PPVT-R scores tended to perceive people in both real and fictional segments as similar to people in their experience, but that was especially true for people in live news broadcasts (Wright, Huston, Reitz, & Piemyat, p. 233, 1994).
Test Reviewers
Piotrowski and Keller (1989) stated that the Peabody Picture Vocabulary Test and its’ revised edition were frequency used, ability, child assessments. Wright, Huston, Reitz, and Piemyat (1994) all agreed that the Peabody Picture Vocabulary Test Revised was valid. Vocabulary scores associated with age changes was the key to children’s understanding of television programming. Older and more intellectually youngsters better understood television programming than younger less and intellectually juveniles. Nevertheless, more cognitively advanced juveniles could not compare and contrast televised characters to their personal experiences.
Senechal, Thomas, and Monker (1993) claimed that the Peabody Picture Vocabulary Test Revised proved that reading at home dramatically improved children’s vocabulary skills. Furthermore, all other teaching methods, regarding expanding vocabulary knowledge, should be used as secondary options (Senechal, Thomas, & Monker, 1993). Even though Lange, MacKinnon, and Nida (1989) admitted that their study was poor due to several limitations, they believed that the Peabody Picture Vocabulary Test Revised was reliable and valid in providing accurate data.
Conclusion
In 1981, Dunn and Dunn revised the Peabody Picture Vocabulary Test. These test developers accessed children’s and adults’ verbal skills by having them point to different icons. Participants completed either Form L or Form M of this multiple choice, nonverbal test between fifteen to twenty minutes. The Rasch-Wright Model of Latent Trait Analysis was essential in making the Peabody Picture Vocabulary Test Revised valid, reliable, and standardized. Based on the literature review, the researchers successfully measured young children’s verbal intelligence.
Personally, I was disappointed in not finding any literature concerning how the Peabody Picture Vocabulary Test Revised affects physically disabled children’s verbal knowledge! Throughout my childhood, I have witnessed so many disabled juveniles, including myself, being mislabeled as having low intelligence due to not being able to point to the test’s pictures. Instead of ignoring this problem, researchers should thoroughly investigate it. There is a high demand for computerized Peabody Picture Vocabulary Tests among severely physically disabled children and adults. Without being able to speak and physically point with fingers, it is completely ridiculous and unjust to mislabel these people as having low intelligence or mentally retarded when they cannot take the general, psychological assessment. If they can successfully communicate on computers, they must be allowed to independently demonstrate their knowledge on computers.
Barch, D. M., & Berenbaum, H. (1996). Language production and thought disorder in Schizophrenia. Journal of Abnormal Psychology, 105(1). 81-88.
Gainer, J. (2000). Lloyd Dunn and Dunn and the Peabody Picture Vocabulary Test. John F. Kennedy Center for Research on Human Development. (On-line) Available E-mail: wysiwyg//44/http://www.vanderbilt.edu/kennedy/about/Dunn and Dunn.html Worldwide Web: 1-2.
Hohnen, B., & Stevenson, J. (1999). The structure of genetic influences on general cognitive, language, phonlogical, and reading abilities. Developmental Psychology, 33(2). 590-603.
Lange, G., MacKinnon, C. E., & Nida, R. E. (1989). Knowledge, strategy, and motivational contributions to preschool children’s object recall. Developmental Psychology, 35(5). 772-779.
Miller, L. T., & Lee, C. J. (1993). Construct validation of the Peabody Picture Vocabulary Test revised: A structural equation model of the acquisition order of words. Psychological Assessment, 5(4). 438-441.
Piotrowski, C., & Keller, J. W. (1989). Psychological testing in outpatient mental health facilities: A national study. Professional Psychology: Research and Practice, 20(6). 423-425.
Senechal, M., Thomas, E., & Monker, J. (1993). Individual differences in 4-year-old children’s acquisition of vocabulary storybook reading. Journal of Educational Psychology, 87(2). 218-239.
Wright, J. C., Huston, A. C., Reitz, A. L., & Piemyat, S. (1994). Young children’s perceptions of television reality determinants and developmental differences. Developmental Psychology, 30(2). 229-239.