in regular lessons with advancedlevel learners. This experience had illustrated that two
listenings to each segment was sufficient for learners to provide more than enough data for
analysis in this study, despite the challenging nature of the input. It entailed the learners
looking at the screen and listening as one of a series of segments of an authentic BBC news
videotext was played. Each segment was shown twice. Learners could not write while the
segment was played to ensure they had the chance to utilize information from the audio and
visual channels to achieve comprehension, as in a ‘reallife’ context. Once a segment had
finished playing the first time, learners made notes within a specified time. This delayed
note taking ensured test takers were not distracted from listening, nor had their
comprehension impaired by a concurrent response task. The segment was then shown again
to allow for the effect of shortterm memory decay on the information learners were able to
recall. In terms of the analysis, it was also of interest to see whether decoding errors
persisted across more than one listening (for lessadvanced learners the number of listenings
can, of course, be increased). Following this, learners used their notes to write extended
responses reflecting their comprehension of the segment contents. The learners were told
they were not required to recall the exact words in each segment, but should focus on
recording their propositional understanding. Clearly, however, this would still require the
recall of key phrases and vocabulary to construct written responses. These extended
responses provided the raw data for the subsequent data analysis. As such, unlike other
studies of L2 decoding, no attempt was made to manipulate the input and present
participants with particular words, parts of words, or nonwords containing the feature to be
examined. Two native speakers completed the listening stage to ensure that across two
listenings it was possible for a listener to comprehend each segment in full, which both
native speakers were able to achieve.