Aachen Test for German Sign Language Competence

This German Sign Language test (ATG) has been developed by an interdisciplinary team at the Rheinisch-Westfälische Technische Hochschule (technical university) Aachen (Fehrmann et al., 1995a, 1995b).

The goal of this test is to measure the basic competence in German Sign Language (DGS). The basic competence of DGS is defined as the language competence that an adult deaf native signer would consider as the minimum fluency/knowledge that is required to be considered a fluent DGS user. The basic competence in DGS of a tested subject is determined by the linguistic judgment of a native signer.

The ATG can be applied for the following areas, e.g. (1) diagnosis of the language development in children, (2) monitoring sign language development in school, (3) linguistic self-assessment of deaf adults, (4) GSL assessment for hearing parents with deaf children, (5) assessment for hearing professionals working in the deaf field, e.g. interpreters, teachers for the deaf, speech therapists etc.

The ATG provides an in-depth investigation of specific linguistic structures of DGS and is no screening instrument. The test can be applied for both children (from age six onward) and adults, both deaf and hearing. The first part of a subtest is to be used for children (until puberty) as well as for adults. The second part is for adults only.

The nine measures of the ATG assesse expressive, receptive, and expressive-receptive language skills. It focuses on different linguistic units, such as signs, phrases, and text.

Task 1: This task focuses on expressive language skills, in particular to measure the use of spontaneous signing. Task 1 is divided into 2 sub-tasks. In the first part, the test- takers are asked questions about certain topics, e.g. their biography and leisure time activities. In the second part, the test-takers are asked to explain certain terms, e.g. computer, airplane, or soccer in German Sign Language. They are videotaped during the testing for later analysis.

Task 2:Task 2 focuses on receptive skills of DGS. The reception of lexical and sub-lexical components should be tested. The subjects were shown a video with 60 (30 for children) pictures (static) containing objects, living creatures, and situations. For each stimulus item, five to eight DGS signs were shown on the video (moving). The subjects are asked to judge on a four point scale ranging from yes to no if a sign represents the meaning according to the picture. From the five to eight provided answers, one sign represents only part of the meaning of the pictures and one the entire meaning of the pictures.

Task 3:Comprehension of phrases. This task focuses on the acting-out of receptive language use on the lexical, morphological, and syntactic levels. The candidates were shown a stimulus video consisting of 36 signed phrases: 15 for the children’s and 21 for the adult’s version of the ATG were used. They were given dolls to mimic the signed phrases. The subjects were required to differentiate between the grammatical and the topological use of the signing space.

Task 4 & Task 5:Describing and naming of picture cards. The subjects were given 60 picture cards (30 for the children) with drawings. This was similar to task 2, but in a different order. At first the subjects should describe the content and then name it. This task tests the subjects’ use of grammar and word order. If the description is insufficient, the subjects are asked to describe the picture again. When the subject is not given a sign for the picture or starts to describe it again, he/she will be asked to name the picture

Task 6:Understanding texts. The acting-out method was used. Phrases that were already used in task 3 are now shown to the subjects as DGS texts, which the subjects have to mimic with dolls. Each text consists of eleven or eight (children’s version) DGS phrases. The stimulus video is shown to the subjects twice. The difference with task 3 is that not all nouns are mentioned explicitly. Some of them occur as pronouns. In order that the subjects can mimic the texts, they have to have linguistic knowledge about the principle of simultaneity in sign language and the use of space. When the subjects have this knowledge, they can decode the morphological, lexical, and syntactic devices used.

Task 7 & Task 8:Imitation of single sign utterances and phrases. This task measures the receptive-expressive language skills on the lexeme and phrase level.

Task 7 consists of 36 single sign utterances and task 8 consists of 24 signed phrases. The 36 single sign utterances of task 7 were divided into three groups: (1) single/lexical signs, (2) combined signs (compounds) functioning as nouns, and (3) predicate signs.

Task 8 consists of four groups of items, each consisting of three phrases: (1) main clause with spatial markers, (2) main clause with tense markers, (3) complex sentences with incorporated conjunction, and (4) complex sentences with explicit conjunction. The subjects should not give an identical reproduction of the stimulus sign, but rather a reproduction of the content. The total number of items are 60 for the adult’s and 30 for the children’s version of the ATG.

Task 9:Reproduction of stories. The subjects are shown six (three for the children) text stimuli on videos, with each shown twice. The subjects are asked to repeat the signed stimuli from the video. The stimuli were more complex than in Task 6. The content of the signed texts were everyday situations containing unexpected events. The unexpected events are expressed by idioms in German Sign Language. The selected idioms have either an idiomatic or a literal meaning in the context of the story. The assumption for this task is, that single signs are repeated completely, while the syntactic and lexical structure will not be repeated completely.

The linguistic structures, which are tested by the ATG, are mostly drawn from the linguistic research of ASL and other signed languages. The ATG does not measure a certain developmental stage in language development. It tests the percentage of the available score a test-taker can achieve. The assumption is that a native signer should achieve at least 90% of the total score one can reach on the predetermined (language) scale. The tasks include a high number of items, which should guarantee a useful psychometric, criterion-references scale. The design and the chosen items, which follow strict linguistic criteria, expect a high construct validity. For each task a rating sheet has been developed (checklist), e.g. for task 1 five criteria were set up for scoring. These criteria were chosen with regard to the communicative competence in DGS and the quantitative distribution and correctness of DGS devices in the signed utterances. The subjects were rated on a scale ranging from 0-4, with 0 being the lowest and 4 the highest score. At the end all individually given scores were added up to create the total score.

The test instruction is performed live by the tester and is adequately adapted to the language level of the subject (deaf vs. hearing). Hearing subjects who only have basic knowledge in DGS will get additional, written standardized test instructions on a cardboard.

The ATG is a criterion-referenced instrument. The scores for the tested subjects are not compared with a normative group. The subjects perform according to a predetermined language scale, which reveals a relative language mastery.

Approximately 100 people have been tested either with the entire ATG or with only several subtasks. Among the group of these 100 test-takers were 15 to 17 hearing people, while the rest were mostly deaf and a few hard-of-hearing people. Most of them underwent only tasks 1, 2, 6, and 9. A few took the entire ATG. So far, the rating sheets for the analysis of the test materials have been improved based on the results of the 100 tested subjects. There are plans to make changes on task 2 based on the results of the test. The results for reliability and validity are not yet available (I. Werth, personal communication, December 21, 1999).

The total testing time including all nine subtests is two hours for children and four hours for adults. This is a rather long test, but subjects can also be tested on certain tasks, or the test can be split over a two-days period. No clear information on the time spent regarding the analysis of the results is provided. It is also not clear if a knowledgeable assessor is required in order to conduct this test.

The ATG is not available. It is still under development, since the people who developed the test are trying to test more people in order to improve the instrument.

Among the strengths of the ATG are that (1) it assesses both the communicative competence and a large range of occurring linguistic devices and linguistic units, (2) it has a broad applicability, deaf and hearing people, children and adults, (3) it tests both language production and comprehension.

Among the weaknesses of the ATG are that (1) no psychometric testing is available in the published literature, (2) it is too long to be used as an assessment tool in an educational setting serving deaf children, (3) the tested linguistic devices of DGS are mostly based on evidence from other sign languages’ research, and (4) it is not yet available.


From: Tobias Haug: “Review of Sign Language Assessment Instruments”, an earlier version of that paper 2005.

For more information regarding this test, please contact  Isa Werth at the RWTH-Aachen, Germany.