Abstract: Distribution of attention in time information processing is one of the hot areas of research science, and the dual-task paradigm is one of the most common ways to study distribution of attention. It requires an individual to perform two tasks simultaneously, the less the attention allocated to a temporal interval, the shorter it is judged(Brown, 1997). The attention sharing effect is discussed within the framework of the scalar expectancy model of timing. In such paradigm the parallel processing itself may interfere with time perception and lead to unexpected deviations. In order to avoid such interference, the timing interruption paradigm would be a better solution as it uses a blank interruption instead. The researchers consistently found a similar position and interruption effect in both paradigms (Casini & Macar, 1997; Cortin, & Remblai,, 2006; Remblai, & Cortin,, 2003). Furthermore, the results showed both the discontinuity and interference of current information processing were belong to interruption effect, but to varying extents (Cortin, & Masse, 2000; Macar, 2002). However, though the position and interruption effect were similar in the two paradigms, they have not been explored in a same stimuli series. As we know, information exchange with the outside world is not dependent on single sensory channel, but rather the interaction of cross-modal information processing. It would be valuable to explore the position and interruption effect in the context ofcross-modal processing. It would not only help to uncover the cognitive mechanism of time processing, but also have important practical values as it is more similar with daily life. Therefore, the present study was designed to investigate the position and interruption effect in the two paradigms in the cross-modal conditions. To this end, the study consisted of two experiments. In experiment 1, 2500 ms and 4500 ms were set for the target time intervals, using the same stimulus sequence (visual presentation, with aural interruption), participants were allocated to control, break and interference condition respectively. In experiment 2,the target intervals were set to 1500 ms and 2500 ms. Results of experiment 1 showed that the interruption effect is more significant in break condition regardless of target time intervals. Furthermore, under the 2500ms, position effect were found in all three conditions, whereas under the 4500ms condition, the position effect only existed in the break condition. Experiment 2 found that there was position effect consistently, regardless of the interpolation conditions or target time intervals. Besides, the interrupt effect was more significant in the break condition. Such results imply that whether the position effect is the same in the two paradigms depends on the length ofthe target intervals to some extent. Besides, in the timing interruption paradigm, the break effect on timing is more significant than the interference effect in dual-task paradigm.

Key words: dual-task paradigm, timing interruption paradigm, position effect, interruption effect, scalar expectancy model of timing