The Effects Of Different Attention Regulation Mode On The Non-symbolic Magnitude Representation Strategies

Non-symbolic number processing capacity is an essential part of daily life. Through a series of previous research, three kinds of non-symbolic magnitude representation strategies were proposed:subitizing, estimation and counting. Subitizing is the ability to enumerate quickly and accurately for small sets (one to four objects); estimation is a rapid but imprecise strategy for large sets; counting is an accurate but time-consuming strategy for larger sets. Whether the three representation strategies share the same processing mechanism is still under debate. Some researchers introduced attention to study magnitude representation strategy. The results showed that subitizing does not require attentional resources while counting need attention to participate in. Thus, they argued that whether a strategy requires attention resources can be served as a criterion to distinguish the different magnitude representation strategy. However, some researchers have found both subitizing and counting require attentional resources. So far, no previous research studied three strategies at the same time. In addition, previous studies mostly paid attention to the attentional load, while few studies used different approaches to control attention. Therefore, the present study adopts two control modes, attention load and attention distribution, to investigate how different attention control modes influence different magnitude representation strategies.Both study 1 and study2 adopt a dual-task paradigm to control the attention. In study 1, in order to investigate the influence of bottom-up attention regulation on the magnitude representation strategies, we change the difficulty of the target detection task to regulate the attention load. In study2, in order to investigate the influence of top-down attention regulation on the difference magnitude representation strategies, we provide cues to regulate the attention distribution. The results of study 1 showed that all three strategies were influenced by attention load, but the effects are different. Attentional load affected the subitizing time only, the higher the attentional load, the longer time needed. For estimation, attention load not only affected the estimation time, but also affected the estimation accuracy. In the high-load attentional condition, the accuracy reduced and the reaction time increased. For counting, attentional load only affected the counting time. In the high-load attentional condition, the counting time increased significantly. The result of study 2 revealed that attention distribution had an impact on the accuracy and reaction times for all three magnitude representation strategies, i.e., accuracies reduced and reaction times increased as the attentional distribution reduced. Taken together, these results showed that the influence of different attentional modes on magnitude representation strategies was different, i.e., attentional load mainly affected the processing time, while attentional distribution not only affected the processing time but also accuracy. In other words, attentional distribution tends to influence the internal mechanism of the magnitude representation. Different strategies needed different attentional regulation. Subitizing depends a lot on attention distribution, while estimation and counting need sufficient attentional distribution and also require low attentional load. These results challenge the idea of a single enumeration mechanism and support the proposal that small numerosities and large numerosities are enumerated by different mechanisms.