Heil M, Wahl K, Herbst M
Department of Psychology, Phillipps University, Marburg, Germany.
Psychol Res. 1999;62(1):48-61. doi: 10.1007/s004260050039.
Two reaction-time experiments using the psychological refractory period paradigm examined whether two prominent tasks, i.e., mental rotation and memory scanning, require access to a single-channel mechanism and must therefore be performed sequentially with other operations requiring the same mechanism. On each trial, subjects made speeded responses to a tone (Exp. 1) or a character (Exp. 2, with symbolic SR-compatibility of the character manipulated) as Task 1 and to a letter (for blocks with mental rotation) or a digit (for blocks with memory scanning) as Task 2. The set-size effect was constant across SOAs, suggesting that memory scanning cannot be performed in parallel with response selection of Task 1. The effect of orientation, however, decreased with decreasing SOA. The decrease was even intensified if Task 1 bottleneck processes were prolonged by symbolic SR-compatibility. The exact pattern of underadditivity, however, was not predicted by current theories of dual-task performance. The results contradict a central bottleneck model but are in line with extensions of the model proposed by Meyer and Kieras.
两项使用心理不应期范式的反应时实验，检验了两项突出任务，即心理旋转和记忆扫描，是否需要使用单一通道机制，因此必须与其他需要相同机制的操作按顺序执行。在每次试验中，受试者对一个音调（实验1）或一个字符（实验2，操纵字符的符号性刺激-反应兼容性）快速做出反应作为任务1，并对一个字母（心理旋转组块）或一个数字（记忆扫描组块）快速做出反应作为任务2。刺激集大小效应在不同的刺激呈现间隔（SOA）下是恒定的，这表明记忆扫描不能与任务1的反应选择并行进行。然而，方向效应随着刺激呈现间隔的减小而降低。如果任务1的瓶颈过程因符号性刺激-反应兼容性而延长，这种降低甚至会加剧。然而，当前的双任务绩效理论并未预测到不完全相加性的确切模式。结果与核心瓶颈模型相矛盾，但与迈耶和基拉斯提出的模型扩展一致。
