Riefer David M, Chien Yuchin, Reimer Jason F
Department of Psychology, California State University, San Bernardino, CA 92407, USA.
Q J Exp Psychol (Hove). 2007 Oct;60(10):1389-405. doi: 10.1080/17470210601025646.
Research is mixed as to whether self-generation improves memory for the source of information. We propose the hypothesis that positive generation effects (better source memory for self-generated information) occur in reality-monitoring paradigms, while negative generation effects (better source memory for externally presented information) tend to occur in external source-monitoring paradigms. This hypothesis was tested in an experiment in which participants read or generated words, followed by a memory test for the source of each word (read or generated) and the word's colour. Meiser and Bröder's (2002) multinomial model for crossed source dimensions was used to analyse the data, showing that source memory for generation (reality monitoring) was superior for the generated words, while source memory for word colour (external source monitoring) was superior for the read words. The model also revealed the influence of strong response biases in the data, demonstrating the usefulness of formal modelling when examining generation effects in source monitoring.
关于自我生成是否能提高对信息来源的记忆，研究结果不一。我们提出一个假设，即在现实监控范式中会出现积极生成效应（对自我生成信息有更好的来源记忆），而在外部来源监控范式中则倾向于出现消极生成效应（对外部呈现信息有更好的来源记忆）。在一项实验中对这一假设进行了检验，在该实验中，参与者阅读或生成单词，随后对每个单词的来源（阅读或生成）以及单词的颜色进行记忆测试。采用了迈泽尔和布勒德（2002年）针对交叉来源维度的多项式模型来分析数据，结果表明，对于生成的单词，生成（现实监控）的来源记忆更好，而对于阅读的单词，单词颜色（外部来源监控）的来源记忆更好。该模型还揭示了数据中强烈反应偏差的影响，证明了在研究来源监控中的生成效应时形式建模的有用性。
