Planning, as a task condition, is hypothesized to influence L2 test performance and thus test scores（Skehan, 1998）. This study investigates the effects of lengths of strategic planning time on L2 paired oral test performance, moderated by L2 proficiency. It aims to determine whether differences in performance result from 0, 1, 2 or 3 minutes of planning time, and whether planning time and proficiency interactively affect performance. The participants were 72 Chinese EFL learners, divided into four groups, each performing the same dialogic task with 1 or 2 or 3 minute planning time or none. All speech recordings were rated by two trained raters, and the transcripts of the speech samples subjected to a discourse analysis, measuring fluency features of Rate A and Rate B, complexity features of syntactic complexity, syntactic variety and lexical variety, and accuracy features of error-free clauses and correct verb forms. Findings show a stable accuracy, higher fluency in the planned condition, and greater syntactic complexity when learners are given 3 minutes＇ planning time. No interaction is found between planning time and proficiency.
