[in Japanese]

View full abstract

We developed a method for efficiently quantifying communicative orientations in accordance to the language teaching observation scheme COLT part B (Spada & Fröhlich, 1995) by means of integrating transcripts and features in tabular format. The drawback of COLT part B is that reviewing a lesson in order to plan for the next is impractical because recordings of lessons must be transcribed. Spada and Fröhlich's time sampling procedure reduces time at the expense of skipping some transcripts and ignoring some schematic features. Our method maximizes transcript coverage. Tab and line-feed characters inserted within the transcripts allow spreadsheet applications to tabulate utterances and turns along with their speaker tags. Numerically coding of verbal interaction features provides frequency counts of verbal interaction features by speakers or utterances. All episodes in every turn are reflected in the analyses. Evaluation experiments of our method suggest that, compared to literal coding representations, transcribing lessons in our format covers more phenomena in less time.

View full abstract

[in Japanese]

View full abstract