Introduction

LLAMA_B is one of a suite of programs dealing with language aptitude. It deals specifically with vocabulary acquisition – learning names for things. The programs were developed as part of a research training programme for Masters students at Swansea University. The main motivation behind the programs was not in fact to develop a language aptitude test, but rather to provide a research environment in which students could naturally develop a critical attitude towards the data collection tools that they were using. In our experience, this critical attitude towards tools was rather rare in the students. Most of them would happily download a program from the web, and use it uncritically in their own dissertation work, even when it was obvious to us that the tool was of dubious origins, not at all reliable, or just completely inappropriate. So we gave the students the LLAMA tests, and instructed them to find out if they were any good. There were four tests in all, somewhat unimaginatively called LLAMA_B, LLAMA_D, LLAMA_E and LLAMA_F.

Each program was loosely based on one of the parts of Carroll and Sapon's Modern Language Aptitude Test (Carroll & Sapon, 1959): LLAMA_D was a sound recognition task that owed something to the work of Elizabeth Service (e.g. Service, 1992; Service & Kohonen, 1995); LLAMA_E was a sound–symbol correspondence task that tested test takers' ability to associate familiar sounds with unfamiliar symbols; LLAMA_F was an innovative test that exposed users to an invented language, and asked them to work out what its grammatical features were; LLAMA_B covered vocabulary acquisition skills. This feature of language aptitude was considered to be very important by Carroll and Sapon, contributing a significant amount of variance to overall language-learning performance. LLAMA_B was similar in content to the vocabulary learning section of MLAT, but it used a more modern interface. Carroll and Sapon had asked their testees to learn a small set of unfamiliar L2 words, but their format merely presented the test takers with a list of English and Kurdish word pairs.