American Sign Language Discrimination Test

The American Sign Language Discrimination Test (ASL-DT) is intended to measure the language proficiency of adults learning ASL as a second language (L2).  The test uses a paired-comparison discrimination task to evaluate learners’ ability to discern phonological and morphophonological contrasts in ASL (i.e., linguistic similarity and dissimilarity within the context of minimal pairs).  The contrasts occur within linguistic categories including movement, handshape, orientation, location and complex morphology.

Each test item consists of two pairs of ASL utterances, each of which contains a standard sentence followed by a comparison sentence.  The respondent must decide if the sentences in each pair are the same or different from each other.  In effect, each sentence pair represents one trial and, for each item, the test-taker must respond to two trials.  Any combination of "same" and "different" trials is possible, allowing for four potential response outcomes: S-D, D-S, D-D, and S-S.  The sentences in each pair may contain one contrasting element.  The contrasting elements in each item generally form a minimal pair.  A sample item is included below for illustrative purposes with the slot containing the contrasting element (minimal pair) underlined. 


Trial 1

           YOUR APPOINTMENT NEED CHANGE.  (standard sentence)

           YOUR HABIT NEED CHANGE.                   (comparison sentence)

Trial 2

           YOUR APPOINTMENT NEED CHANGE.  (standard sentence)

           YOUR APPOINTMENT NEED CHANGE.  (comparison sentence)


The ASL-DT is currently under development as a computer-based adaptive test.  The test content and methodology closely resemble a newly-developed measure of speech recognition, the NTID Speech Recognition Test (NSRT:  Bochner, Garrison & Doherty, 2015; Garrison & Bochner, 2015).  Previous research on hearing L2 learners’ ability to discern linguistic contrasts in ASL showed that movement contrasts were the most difficult and location contrasts the easiest, with the other categories of stimuli being of intermediate difficulty.  In addition, performance was found to be associated with learners’ level of ASL proficiency (Bochner, Christie, Hauser & Searls, 2011). 

More recently, the validity and reliability of a preliminary version of the ASL-DT was investigated, along with the fit of test data to the Rasch model of person measurement.  The results of this study indicated that the measurement procedure: (1) separated learners into groups reflecting beginning, intermediate and advanced levels of ASL skill; (2) provided valid and reliable data concerning learners’ ASL ability; (3) resulted in scores which can be interpreted as a proxy for ASL proficiency; and (4) provided test data which fit the Rasch model of person measurement.  Since the data fit the Rasch model, the stimulus materials and discrimination task can serve as the basis for development of a computer-based adaptive test of ASL ability.  Using an up-down method of item selection, a software application will administer the test, efficiently and accurately determining individuals’ level of ability (Bochner, Samar, Hauser, Garrison, Searls & Sanders, 2015). 

The final version of the ASL-DT will consist of approximately 350 items.  Each test session will require about 35 items and 10 minutes of administration time.  The test is planned for release early in 2017, and will be disseminated as an internet-based computer application.   

The ASL-DT has many strengths, among them is the fact that its validity and reliability are supported by empirical evidence.  Additional evidence of validity and reliability is needed, however, and this is a weakness which should be addressed in future work.  In particular, it will be important to compare ASL-DT scores to concurrent measures of various sorts when they become available (e.g., measures of ASL proficiency, as well as measures of ASL reception and production).




Summarized by Joseph Bochner (2015).


For more information regarding this test, please contact  Joseph Bochner at the NTID, Rochester Institute of Technology.