Thinking about semantic concepts in schizophrenia: the more familiar the less deviation.
Thinking about semantic concepts in schizophrenia: the more familiar the less deviation.
复制标题
对精神分裂症语义概念的思考：越熟悉越少偏差。
DOI：
10.1016/j.schres.2009.11.013
复制
发表时间：
2010
影响因子：
4.5
通讯作者：
Elvevåg,Brita
中科院分区：
文献类型：
--
作者：
Storms,Gert;Elvevåg,Brita
许多任务已被用来检查精神分裂症中假定的语义缺陷。在语义句子验证任务中，患者犯了更多的错误，并且建立真实性的速度较慢（“Desks wear clothes”-Tamlyn et al.，1992年），而当对生活和人造物品的知识进行探测时，在大多数测试中表现更差（MacKay等人，1996年）。其他研究表明语义知识是等价表示的（Cohen et al.，2005年）。此外，一些技术（多维缩放和聚类技术来分析类别流畅性和三重比较数据）用于表明增加的变异性可归因于语义缺陷，但不符合关于参与者之间一致和可靠变异性的统计要求（Elvevåg & Storms，2003）。这使得许多技术不适合于临床人群，其中存在关于更多方差的隐含假设（Prescott等人，2006; Storms等人，2003年）。虽然这些发现并没有反驳精神分裂症语义缺陷的假设，但它们强调了更适合患者内部和患者之间存在差异的测试的必要性。因此，我们收集了类别的典型性评级（例如，牛奶在类别食物中的典型性），以确定患者和对照组是否同意一个项目在其类别中的代表性（Elvevåg et al.，2005年）。典型性预测与类别相关的任务的性能，例如分类速度（Smith & Medin，1981）和归纳推理（Osherson et al.，1990年）。我们对典型性评级的一致性和可靠性（即，患者和对照组之间以及对照组和患者组之间的个体间一致性）的检查显示，各组之间和各组内的相关性相似。与流畅性和三联任务相比，患者间的一致性较高，精神分裂症患者的语义表征与健康受试者相似。
Numerous tasks have been employed to examine putative semantic deficits in schizophrenia. In a semantic sentence verification task, patients made more errors and were slower establishing veridicality (“Desks wear clothes”-Tamlyn et al., 1992), and when knowledge of living and human-made items was probed performed worse on most tests (MacKay et al., 1996). Other studies suggest semantic knowledge is represented equivalently (Cohen et al., 2005). Furthermore, some techniques (multidimensional scaling and clustering techniques to analyze category fluency and triad comparison data) employed to suggest increased variability is attributable to semantic deficits fail statistical requirements concerning consistent and reliable variability across participants (Elvevåg & Storms, 2003). This renders many techniques illsuited for clinical populations where there is an implicit assumption about more variance (Prescott et al., 2006; Storms et al., 2003). Although such findings do not refute the hypothesis of semantic deficits in schizophrenia, they highlight the need for tests better suited where variability within and between patients exists.Therefore we collected typicality ratings of categories (eg, how typical is milk of the category food) to establish whether patients and controls agree on how representative an item is of its category (Elvevåg et al., 2005). Typicality predicts performance on category-related tasks, such as speed of categorization (Smith & Medin, 1981) and inductive inference (Osherson et al., 1990). Our examination of consistency and reliability of typicality ratings (ie, interindividual consistency within patients and controls and between control and patient groups) revealed similar correlations both between and within each group. The inter-patient consistency derived was high compared to that from fluency and triad tasks, and semantic representations in schizophrenia paralleled those in healthy participants.