Doctoral Dissertation Research: Event-Related Potentials Tracking Reading Experience, Lexical Quality, and Text Comprehension in an Artificial Orthography

在美国国家科学基金会的支持下，哈特女士和她的顾问佩尔费蒂博士将对阅读理解问题是由低质量的词汇表征引起的这一假设进行为期一年的调查。 许多研究致力于寻找阅读理解问题的原因。 例如，众所周知，大学时代的读者在理解他们所读的内容方面有困难，他们的词汇量较小，在推理、找到主要思想和保持动机方面有更多的困难。虽然所有这些因素都与阅读理解相关，但它们似乎是一个更深层次来源的表现--低质量的词汇表征。 心理词汇中的单个词由语音、拼写和语义信息组成。 为了支持良好的阅读理解，这些表征必须是可靠的（一个输入连接到一个词条）和连贯的（例如，同一词条的语音和拼写代码）。 当可靠性和连贯性崩溃时，理解就会受到影响。 在本研究中，哈特和佩尔费蒂旨在通过将熟练和不熟练的阅读者在阅读经验上等同起来，来描述理解技能的词汇基础。受试者在实验者控制受试者对语言的体验以及诸如词频和同音异义等属性的条件下学习人工语言。 在掌握这种语言的早期和晚期，受试者接受语言测试，并在阅读人工语言时测量头皮电位（使用ERP方法）。 被测试的假设预测，当在经验不足的情况下进行测试时，所有受试者都会表现得像其他实验中技能较低的人一样，而在丰富的经验之后，所有受试者都会表现得像技能较高的人一样。本研究的发现有助于澄清阅读理解问题的本质。 在词汇水平上解释理解困难对旨在提高阅读理解的干预措施具有启示意义。

With National Science Foundation support, Ms. Hart and her advisor Dr. Perfetti will conduct a year long investigation of the hypothesis that reading comprehension problems are caused by poor quality lexical representations. Much research has been undertaken to find the causes of reading comprehension problems. For example, it is well known that college age readers who have trouble comprehending what they read have smaller vocabularies, more trouble making inferences, finding main ideas, and maintaining motivation. While all of these factors are correlated with reading comprehension, it is plausible that they are manifestations of a deeper source - that of poor quality lexical representations. A single word in the mental lexicon is composed of phonological, orthographic, and semantic information. In order to support good reading comprehension, these representations must be reliable (one input connects to one lexical entry) and coherent (e.g. phonology and orthography code for the same lexical entry). When reliability and coherence break down, comprehension suffers. In the present project, Hart and Perfetti aim to describe the lexical basis of comprehension skill by equating skilled and less skilled comprehenders on reading experience. Subjects learn an artificial language under conditions where experimenters have control over subjects' experience with the language and over properties such as word frequency and homophony. At early and late stages of mastering this language, subjects receive language tests and have the electrical potentials at the scalp measured (using ERP methods) while reading the artificial language. The hypothesis being tested predicts that when tested after little experience, all subjects will behave like less-skilled comprehenders do in other experiments, and after extensive experience, all subjects will behave like skilled comprehenders. The findings from this study are expected to clarify the nature of reading comprehension problems. Explaining comprehension difficulties at a lexical level has implications for interventions designed to improve reading comprehension.