Although large-scale assessments (LSA) of school achievement claim to measure domain-specific achievement, they have been criticized for primarily measuring domain-general abilities. Numerous studies provide evidence that LSA of mathematical achievement as well as verbal achievement cover both general cognitive abilities (GCA) and domain-specific achievement dimensions. We extend previous research by analyzing a standards-oriented and literacy-oriented LSA in the domain of science to determine the relation of these two assessment types with domain-general abilities. While literacy-oriented assessments focus on the knowledge and skills students need to meet the demands of modern societies, standards-oriented assessments focus on national educational standards and curricula. A sample of 1722 students worked on three assessments: (a) the PISA scientific literacy assessment; (b) a standards-oriented assessment based on the German National Educational Standards in biology, chemistry, and physics developed by the Institute for Educational Quality Improvement (IQB); and (c) a GCA test. Comparisons of competing structural models showed that models differentiating between domain-specific achievement and GCA best represented the structure of the assessments. Furthermore, standards-oriented and literacy-oriented LSAs in science shared common variance with GCA but also comprised specific variance. In addition to a factor representing students' GCA, we identified a science literacy-oriented and two standards-oriented factors. Relations with school grades in various STEM and non-STEM subjects were mixed and only partly provided evidence for the specificity of science LSAs. Our findings are important for understanding and interpreting results of LSAs in the contexts of GCA and science. We discuss our outcomes with respect to educational monitoring practices.

尽管对学校成绩的大规模评估（LSA）声称可以衡量特定领域的成就，但由于主要衡量领域通用能力而受到批评。大量研究表明，数学成就和语言成就的LSA涵盖了一般认知能力（GCA）和特定领域的成就维度。我们通过分析科学领域的面向标准和面向读写能力的LSA来扩展先前的研究，以确定这两种评估类型与领域通用能力之间的关系。面向识字的评估侧重于学生满足现代社会需求的知识和技能，而面向标准的评估则侧重于国家教育标准和课程。来自1722名学生的样本进行了三项评估：（a）PISA科学素养评估；（b）由德国教育质量改进研究所（IQB）制定的基于德国生物学，化学和物理国家教育标准的，面向标准的评估；（c）GCA测试。竞争性结构模型的比较表明，区分特定领域的成就和GCA的模型最能代表评估的结构。此外，科学中面向标准和面向读写能力的LSA与GCA具有共同的差异，但也包含特定的差异。除了代表学生GCA的因素外，我们还确定了面向科学素养和两个面向标准的因素。在不同的STEM和非STEM科目中，与学校成绩的关系是混杂的，并且仅部分提供了科学LSA特殊性的证据。我们的发现对于在GCA和科学的背景下理解和解释LSA的结果非常重要。我们讨论有关教育监控实践的结果。