中国已经进入人口老龄化社会，如何积极面对老龄化问题并保证老人的生活质量成为当下以及未来30年中不可忽视的议题。认知老化的执行衰退假说提出，老年人在执行功能方面的衰退是造成日常生活中行为表现下降的主要原因。
       任务切换作为执行功能的子成分，反映了个体的认知灵活性。衡量个体任务切换能力的指标是切换代价，即执行不同任务与执行相同任务间的反应时和正确率的差异。在以往的研究中，研究者们对切换代价的来源提出了三种假设，即任务设置重构理论、任务设置惯性理论和联结竞争理论：任务设置重构理论强调切换代价受个体在任务前的主动准备过程影响，是一种内源控制过程；任务设置惯性理论强调个体执行前一任务的认知惯性会延迟到下一个任务，从而对下一个任务的表现产生影响，是一种外源控制过程；而联结竞争理论强调切换代价是由任务间高度重合的刺激-反应联结导致的，具体情境会使某个联结在竞争中胜出，处理竞争的过程造成了反应时和正确率的损耗。然而最近的研究表明，切换代价不是由单一认知过程决定的，而是受到以上三种理论的综合影响。
       大量研究表明老人的执行功能存在老化现象。然而当以切换代价为考察指标时，虽然部分学者在老年人中发现了相较于年轻人更大的切换代价，但大部分文献更支持切换代价没有年龄差异的结论。本文在三种理论框架下，采用任务线索范式结合数字判断任务，分别考察了预先准备时间（CSI）、线索提取难度、抑制与去抑制和联结数量在不同年龄组间产生的切换代价，得到以下结论:
       第一，在任务设置重构观点下，CSI和线索提取难度在两组被试间产生了一致的影响， 短CSI和颜色线索带来了更大的切换代价，但未发现由不同预先准备条件带来的切换代价的年龄差异，与以往大多数研究结果一致。
       第二，在任务设置惯性观点下，老年组比青年组产生了更大的滞2重复代价，提示老年人的抑制能力和去抑制能力均低于年轻人。
       第三，在联结竞争观点下，在 4 个任务间切换没有比在2个任务间切换产生更大的切换代价，也未发现更多联结数量对切换代价造成年龄差异。
       本文最后讨论了各具体因素对切换代价年龄差异的影响，以及实验中的不足和未来进一步研究的方向。

    China has entered the aging society. How to positively face the aging problem and ensure the quality of elderly’s life become an issue that cannot be ignored at present and in the next 30 years. The executive control hypothesis of cognitive aging proposes that the decline executive function in the elderly is the main cause of the decline in behavioral performance in daily life. 
    As a sub-component of executive function, task switching reflects individual cognitive flexibility. The index of individual task switching ability is switching cost, which is the difference of reaction time and correct rate in different tasks and the same task. In previous studies, researchers proposed three hypotheses about the source of switching cost, namely, task-set reconfiguration theory, task-set inertia theory and associative competition theory. Task-set reconfiguration theory emphasizes that switching cost is affected by the individual's active preparation process before the task and is an endogenous control process. Task-set inertia theory emphasizes that the cognitive inertia of an individual performing the previous task will be delayed to the next task, thus affecting the performance of the next one, which is an exogenous control process. The associative competition theory emphasizes that the switching cost is caused by highly-overlapping stimulus-response connections between tasks. The specific situation will make a certain connection win from the competition, and the process of dealing with the competition will lead to the cost of reaction time and accuracy. However, recent studies have shown that switching costs are not determined by a single cognitive process, but by the combined influence of the above three theories.
   A large number of studies have shown that the executive function of the elderly is aging. However, when switching cost is taken as the dependent variable, although some scholars found that the switching cost in the elderly is greater than that in the young, most literature supports the conclusion that there is no age difference in switching cost. Under the above-mentioned three theoretical frameworks, the thesis uses the task-cue paradigm combined with the digital judgment task to investigate the switching costs of advance preparation time (CSI), cue extraction difficulty, inhibition and deinhibition, and number of connections among different age groups, respectively. The and following conclusions have been gained:
    First, under the perspective of task-set reconfiguration view, CSI and cue extraction difficulty have consistent effects between the two groups of subjects. Short CSI and color cues bring greater switching cost, but no age difference in switching cost caused by different advance preparation conditions is found, which is consistent with the results of most previous studies. 
    Second, under the view of task-set inertia, the elderly group has a greater lag-2 repetition cost than the young group, suggesting that the inhibition and deinhibition ability of the elderly are worse than that of the young.
    Third, under the associative competition view, switching among 4 tasks does not produce a greater switching cost than switching between 2 tasks, and no age difference is found in different switching cost due to the greater number of connections.
    At the end of the thesis, the discussions about the influence of various factors on the age differences of switching cost is given, as well as the shortcomings in three experiments and the direction of further research.