Acquisition of hierarchical control in skilled action sequencing

在熟练的动作排序中获得分层控制

• 批准号: 1353360
• 负责人: Matthew Crump
• 金额: $ 35.61万
• 依托单位: CUNY Brooklyn College
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2014
• 资助国家: 美国
• 起止时间: 2014-07-01 至 2018-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1353360&HistoricalAwards=false
• 关键词: Acquisition; hierarchical; control; skilled; action

A critical issue in cognitive science is understanding how behavior unfolds in time. Most human behaviors, from walking and talking, to driving, typing and playing instruments, consist of a series of ordered actions. Little is known concerning how people learn the serial ordering of actions and how this ability changes with expertise. This investigation uses the task of skilled typing to study how high- and low-level cognitive processes interact to allow typists to order their keystrokes rapidly and accurately. Typing is a useful model system for this purpose because experts are plentiful and typing performance can be measured precisely in terms of keystroke timing and errors. Prior studies have shown that typing skill is controlled hierarchically by two processing loops: an "outer loop" turns ideas into words and sentences and an "inner loop" translates words into motor movements for executing individual keystrokes. Three related projects test how these loops control serial ordering across levels of typing skill. Project 1 uses a large-scale web-based approach to quantify how novices and experts become sensitive to the statistics of natural keystrokes and learn to optimize typing performance for predictable letter patterns. Project 2 explores how the inner loop becomes gradually connected to the outer loop, allowing verbal codes used for language production to control serial ordering of fingers on the keyboard. Project 3 tests how the inner loop schedules the timing of individal keystrokes after it has received word-level instructions from the outer loop. Typing, especially on computers and smartphones, has become an indispensable modern skill. This work can provide new statistical signatures of typing skill that can be used to promote rapid skill acquisition. The science of typing can also serve as a model for investigating other domains that rely on hierarchical control loops to accomplish complex serial ordering, such as playing a musical instrument, assembly work, and interactive technologies. The project also provides training opportunities for graduate and undergraduate students at an institution that serves several minority and underrepresented groups.

认知科学的一个关键问题是理解行为如何随时间展开。大多数人类行为，从走路和说话，到驾驶、打字和演奏乐器，都由一系列有序的动作组成。关于人们如何学习动作的连续顺序以及这种能力如何随着专业知识的变化而变化，人们知之甚少。这项调查利用熟练打字的任务来研究高水平和低水平的认知过程如何相互作用，以允许打字员快速准确地排序击键。为此目的，打字是一个有用的模型系统，因为专家很多，并且可以根据击键时间和错误来精确测量打字性能。先前的研究表明，打字技能是由两个处理循环分层控制的：“外循环”将想法转化为单词和句子，“内循环”将单词转化为执行单个击键的运动。三个相关项目测试这些循环如何控制跨打字技能级别的串行排序。项目 1 使用基于网络的大规模方法来量化新手和专家如何对自然击键的统计数据变得敏感，并学习如何优化可预测字母模式的打字性能。项目 2 探索内循环如何逐渐连接到外循环，从而允许用于语言生成的语言代码控制键盘上手指的串行顺序。项目 3 测试内循环在收到来自外循环的字级指令后如何安排各个击键的时间。打字，尤其是在电脑和智能手机上打字，已成为一项不可或缺的现代技能。这项工作可以提供打字技能的新统计特征，可用于促进快速技能习得。打字科学还可以作为研究依赖分层控制循环来完成复杂串行排序的其他领域的模型，例如演奏乐器、组装工作和交互技术。该项目还为一个为多个少数族裔和代表性不足群体提供服务的机构的研究生和本科生提供培训机会。