Previous research indicates that readers’ prior beliefs bias comprehension of conflicting sources and the ensuing representations developed. Against this background, this study investigated how participants’ pre-existing beliefs affect their written representations based on conflicting texts about a well-established controversy. More specifically, adopting a 2 × 2 mixed GLM design and using a series of statistical procedures, the study investigated the propositional content and perspectives that L2 reader-writers adopted in their written representations based on controversial sources across summary vs. argumentation task instruction conditions. The study further investigated the participants’ emotional reactions to the conflicting texts. Results showed that perspectives that participants adopted in their written representations and the propositional content therein were biased towards their prior beliefs. Additionally, the results showed an interaction effect for task instruction and propositional content in the representations. More specifically, the argumentation task showed less of a balance in positively-biased and negatively-biased propositional content than the summary task, although attenuated by the absence of significant cross-condition differences. Furthermore, significant differences were found in the participants’ emotional reactions to the conflicting texts. Moderating effects were also found for curiosity and confusion experienced in relation to the pro-stance texts and the propositional content of the integrated representations.

先前的研究表明，读者先前的信念会影响对相互冲突的来源和随后形成的表述的理解。在此背景下，本研究调查了参与者预先存在的信念如何影响他们基于关于既定争议的相互冲突的文本的书面陈述。更具体地说，该研究采用 2 × 2 混合 GLM 设计并使用一系列统计程序，调查了 L2 读者-作者在基于总结与论证任务教学条件下有争议的来源的书面陈述中采用的命题内容和观点。该研究进一步调查了参与者对相互冲突的文本的情绪反应。结果显示，参与者在书面陈述中采用的观点和其中的命题内容偏向于他们之前的信念。此外，结果显示了表征中任务指令和命题内容的交互作用。更具体地说，论证任务在积极偏见和消极偏见的命题内容方面表现出比摘要任务更少的平衡，尽管由于不存在显着的跨条件差异而减弱了这种平衡。此外，参与者对相互冲突的文本的情绪反应也存在显着差异。还发现了与立场文本和综合表征的命题内容相关的好奇心和困惑的调节作用。