如今，科技的突飞猛进使得社会环境愈发复杂多变。在这样的时代背景下，创造力的重要性日益凸显。创造力不仅是推动社会进步的重要动力，更是个人发展和社会竞争中的关键因素。如何评价创造力、提升创造力一直以来都深受学界关注。创造性写作是一种强调创意与个性化的写作形式，其意义在于能够全面、深入地挖掘和评估个体的创造力水平。通过创造性写作，研究者不仅能够观察到个体在思维发散、想象创新方面的表现，还能够了解其在问题解决、观点表达等方面的能力。

然而，由于作答文本具有高度自由化、想象性和个性化等特点，目前创造性写作的评价主要依靠人工评分，存在评分者主观效应和评价时间、人力成本过高的局限性。同时，在创造力领域中，凡是涉及创造性写作的绝大多数研究中仅报告创造性总分，缺少对于创造性写作的评价维度和评分流程的阐述，使该任务测量的信效度存疑。目前关于创造性写作任务的测量结构和评分方法的研究还有进一步深入的空间。

本研究旨在提出创造性写作任务的一般性评价模型，建立一套标准化的评分框架，辅助研究者使用创造性写作任务进行更为全面、准确的创造力评估。研究一在对现有创造性写作理论、课程、评分准则的文献调研基础上，邀请共20名来自文学和影视领域从业多年的专家进行半结构化访谈，分别针对一般性写作和创造性写作任务中的创造性特点展开论述，完成录音文本整理、开放式编码、观点抽取和总结，形成创造性写作的初版评价维度。研究二为了验证模型，采用六选三和十二选三写作任务通过问卷星进行试测验，并以创造力测评的准确性和操作性为目标，从专家中选取五位进行二次访谈，筛选并归纳创造性写作的重要评分维度及其定义和评分方法。确定创造性写作评价模型，该模型共有五个维度：新颖性、反定势思维、概念组合、想象力、情节惊奇。为验证创造性写作评价模型的信效度，研究三在467名大学生中采用关键词写作和PISA2021写作任务进行正式测验，并招募5名评分员进行评分。评分者一致性良好，验证性因素分析的结果表明创造性写作评价模型具有较好的一致性。

本研究的创造性写作评价模型极适用于学术研究，帮助研究者全方位了解写作中的创造性表现，为创造力领域进一步研究与选拔提供了有效的工具支持。未来研究方向将进一步验证创造性写作评价模型的外部效度和重测信度问题，并探索一套针对创造性写作维度的自动化评分程序。

Nowadays, the rapid development of technology has made the social environment increasingly complex and changeable. Against such a backdrop, the importance of creativity has become increasingly prominent. Creativity is not only an important driving force for social progress, but also a key factor in personal development and social competition. The evaluation and enhancement of creativity have always attracted the attention of academic circles. Creative writing is a form of writing that emphasizes creativity and individuality, and its significance lies in the comprehensive and deep excavation and evaluation of individual creativity. Through creative writing, researchers can not only observe individuals' performance in divergent thinking and imaginative innovation, but also understand their abilities in problem-solving and opinion expression.

However, due to the highly free, imaginative, and personalized characteristics of writing responses, the evaluation of creative writing currently relies mainly on manual scoring, which has limitations such as subjective effects of the scorer, excessive evaluation time, and high labor costs. Meanwhile, in the field of creativity, most studies involving creative writing only report the overall creative score, lacking explanations of the evaluation dimensions and scoring processes for creative writing, which casts doubt on the reliability and validity of this task measurement. There is still room for further research on the measurement structure and scoring methods of creative writing tasks.

This study aims to propose a general evaluation model for creative writing tasks and establish a standardized scoring framework to assist researchers in conducting more comprehensive and accurate assessments of creativity using creative writing tasks. In the first study, based on a literature review of existing creative writing theories, courses, and scoring criteria, 20 experts from the fields of literature and film and television with years of experience were invited for semi-structured interviews. They discussed the creative characteristics of general writing and creative writing tasks, completed the transcription and organization of audio texts, conducted open coding, extracted viewpoints, and summarized them to form the initial evaluation dimensions of creative writing. In the second study, to validate the model, six-out-of-three and twelve-out-of-three writing tasks were used for pilot testing through Questionnaire Star. With the accuracy and operability of creativity assessment as the goal, five experts were selected for a second interview to screen and summarize the important scoring dimensions, definitions, and scoring methods of creative writing. The creative writing evaluation model was determined, consisting of five dimensions: novelty, anti-stereotypical thinking, conceptual combination, imagination, and plot surprise. To verify the reliability and validity of the creative writing evaluation model, the third study conducted a formal test among 467 college students using keyword writing and PISA 2021 writing tasks, and recruited five scorers for scoring. The scorers showed good consistency, and the results of confirmatory factor analysis indicated that the creative writing evaluation model had good consistency.

The creative writing evaluation model in this study is highly suitable for academic research, helping researchers to fully understand the creative performance in writing and providing effective tool support for further research and selection in the field of creativity. Future research directions will further validate the external validity and test-retest reliability of the creative writing evaluation model, and explore an automated scoring program for the dimensions of creative writing.