Perlesko: Vocabulary test for German Sign Language

The Perlesko (German: Prüfverfahren zur Erfassung lexikalisch-semantischer Kompetenz) is a receptive vocabulary test for deaf and hard-of-hearing elementary school-aged children (3rd to 5th grade. The purpose of the Perlesko is to assess the vocabulary development of German Sign Language (DGS), and spoken and written German (Bizer & Karl, 2002). Its primary goal is to provide information to practitioners to monitor language development in deaf elementary school children (Bizer & Karl, no year). It is the first standardized receptive vocabulary test for sign language in Germany.

The Perlesko is primarily a screening tool for young deaf children from 3rd to 5th grade (Bizer & Karl, 2002). It can be used to assess individuals’ comprehension skills in three language modalities, i.e. DGS, spoken, and written German. The test format for these modalities is the same: it has three subtests each of which focuses on a different class of words: nouns, verbs, and adjectives. The item for these subtests were seleced based on a frequency list which was used in a number of vocabulary development studies of German (Bizer & Karl, no year). This list contained the most frequently used nouns, verbs, and adjectives of the targeted age group for the standardization of the Perlesko. Based on that list, the authors selected 54 words for item development, 18 items for each word class (18 nouns, verbs, and adjectives), in order to obtain enough information on the vocabulary development of children, and also to meet the criteria for effective test administration (i.e., length of test). The items of the test are ordered according to their level of difficulty. Each subtest begins with two practice items. The test uses a format follows a multiple-choice format based on which each correct response is presented along with three distractors. The distractors are chosen differently for each language subtest, i.e. distractors with specific sign, mouthing, or written German pattern.

The Perlesko was developed as a Ph.D. project at Hamburg University, Germany. It originated from research done on deaf children who were part of a bilingual classroom program trial at a local school of the Deaf (Günther, 1999). Two test versions were developed; a pilot followed by a standardized version. The following information is based on the standardized test version.

Bizer and Karl (2002) state that their norming sample (of the standardization study) consisted of 112 deaf and hard-of-hearing children from 3rd to 5th grade from seven schools in Germany and one school in Switzerland. The age of their sample ranged from 7;11 to 13;3 (M = 10;5) for the norming of the DGS and written German section of the Perlesko, while the age range for the spoken German section of the Perlesko at the time of the norming study was 8;0-13;4 (M = 10;6). Since the Perlesko is aimed for use in schools at these grade levels, the norming sample needs to be representative of this population. According to the authors, about 25% of this specific population is covered with their sample of 112 children from 3rd to 5th grade, thus the norming sample is deemed representative. For the standardization of the study, the results of the 3rd, 4th, and 5th graders were taken together and not analyzed separately. The primary reason for this procedure was that the goal of the Perlesko is not to make any inference between 3rd and 5th grade deaf children, but rather to have a tool at hand for practitioners to see if deaf children’s vocabulary knowledge is comparable to their hearing peers (at the same age).

The reported psychometric properties of the Perlesko are good. The Perlesko shows evidence for item discrimination and facility index (Bizer & Karl, no year). The reliability of the test was established by using the split-half method (using Spearman-Brown and Kristof) and the reliability coefficient of internal consistency (using the Horst and Kuder-Richardson 8 formula). The results generated from the split-half method ranged between .79 and .87 (total: .85) on each subtest (i.e. nouns, verbs, adjectives) of the Spearman-Brown and Kristof formula on each of the three language sections. The results of the internal consistency ranged from .67 to .91 (total: .89 and .80) on each subtest, applying the Horst and the Kuder-Richardson formula, respectively. In sum, the calculated reliability coefficient (using different methods) shows strong evidence of reliability across the entire test as well as for each subtest on each of the three language sections.

Evidence for content validity of the Perlesko is provided since the item representation is derived from a vocabulary list which represents the vocabulary knowledge of the (hearing) age target group (Bizer & Karl, no year). Construct validity was established by correlating intra-individual factors with the test results. These factors were, for example, (1) grade attended in school, (2) educational policy of the school (oral, sign, or bilingual), (3) hearing status of the parents, (4) gender, and (5) age. The results show that the construct “knowledge of receptive vocabulary” is represented by the Perlesko.

In order to establish concurrent validity, a rating procedure was applied which used an external variable due to the absence of a standardized DGS measure. More specifically, teachers rated their pupils’ children’s vocabulary knowledge independently for each of the three language sections. This typf of information on the rating was provided for 62 of the 112 participants. The ratings were correlated with the children’s test results. The correlation of the DGS score was .54, for spoken German .65 and for written German .76, and .70 across all three modalities. All results are highly significant and thus show strong correlation between the test and the tested criteria.

The Perlesko consists of three booklets and scoring sheets. The three booklets are (1) the manual, (2) stimuli for written German, and (3) stimuli for DGS and spoken German together. The stimuli of the three subtests consist of child-friendly, simple black and white drawings. It takes about 10 minutes to complete each language section; each child is tested individually. The tester and the child are sitting opposite of each other on a table with the test booklet facing the child, so s/he can turn the pages him-/herself. The tester takes notes of the children’s responses on the scoring sheet of the subtest.

Example of the Perlesko for DGS and spoken German (© Bizer & Karl, no year; target: shoe)
Example of the Perlesko for DGS and spoken German (© Bizer & Karl, no year; target: shoe)

For the DGS and spoken German sections, the tester presents the target item in spoken or sign language. The child is encouraged to point to the right answer in choosing on of the four pictures provided in the booklet (see picture above). In the written German test section, the child sees a picture and is asked to find the corresponding German word out of four words that are listed on the right side of the picture.

The analysis of the results follows a mixed measures approach. For the quantitative approach, the achieved raw scores for each language section have to be calculated. These raw scores provide information on the percentiles ranks. The percentile ranks of each child show how many percent of the norming sample performed better or worse than the tested child. In a next step, a qualitative approach for analyzing the error pattern of the tested children is recommended in order to obtain a more complete overview of vocabulary development in DGS, spoken, and written German of every child. The analysis of the error pattern provides information on the distractors a child selected over the correct answer. 

The Perlesko has been used recently (after it has been published) at the school for the Deaf in Hamburg (Wildemann, 2008). However, the school used only the DGS and the written German sections of the test, primarily with children from 3rd to 5th grade. They also included children in 1st and 2nd grade. The results revealed that the DGS section did not differentiate between the younger (1st and 2nd grade) and the older children (3rd to 5th grade). The differentiation worked well with the written German section of the Perlesko. The authors of the Perlesko stated that they plan to standardize the Perlesko in the future for deaf children in 1st and 2nd grade (Bizer & Karl, no year).

The Perlesko is one of the few normed and available sign language (and spoken and written German) tests worldwide and the first one for DGS. It can be downloaded for free here.

Among the strengths of the Perlesko are (1) that it is standardized and commercially available on the market, (2) that it shows sound psychometric properties, and (3) that it is easy to administer, score, and to analyze.

Among the weaknesses of the Perlesko are (1) that no detailed information is provided on the issue of variations in DGS (dialects) and (2) the reported problems that DGS section does not differentiate sufficiently between younger and older children (Wildemann, 2008).

Summarized by Tobias Haug (2009).

For further information on the Perlesko, please contact  Sibylle Bizer in Hamburg directly.


The Perlesko can be downloaded for free here.