Prater R J, Swift R W
J Commun Disord. 1982 Sep;15(5):395-410. doi: 10.1016/0021-9924(82)90007-7.
Sixty subjects between the ages of 21 and 48 months were placed into groups based on mean length of utterance (MLU) and chronological age (CA) to test Stampe's hypothesis about the development of phonological processes. The subjects were given the Phonological Process Analysis (Weiner, 1979), and a proportion score for each of sixteen phonological processes was derived for each subject. The use of the various phonological processes across MLU groups and age groups was analyzed by multiple regression correlation, three-way ANOVA, and Tukey's HSD tests. MLU was found to be the best classification for the subjects for describing the phonological processes. The results and discussions present MLU-referenced guidelines regarding the emergence, duration, elimination, and universality of the sixteen phonological processes.
60名年龄在21至48个月之间的受试者根据平均语句长度（MLU）和实际年龄（CA）进行分组，以检验斯坦普关于音系过程发展的假设。这些受试者接受了音系过程分析（韦纳，1979年），并为每个受试者得出了16个音系过程中每个过程的比例分数。通过多元回归相关性、三因素方差分析和图基的HSD检验，分析了不同MLU组和年龄组中各种音系过程的使用情况。结果发现，MLU是描述音系过程的受试者的最佳分类方式。结果和讨论提出了关于16个音系过程的出现、持续时间、消除和普遍性的以MLU为参照的指导方针。
