Landscape Analysis of Early Science Assessments

Overview

To launch the work of CAESART, we are conducting a comprehensive landscape analysis that includes a systematic review of existing preK to grade 5 science assessments for use by researchers. This analysis will document assessments and tools available to capture elementary students’ science learning and identify key features of these assessments, such as the age and grade ranges they target, psychometric details, attention to fairness and accessibility, and resources for administration.

This landscape analysis is a foundational study in CAESART’s mission to improve science education for early elementary students. By systematically reviewing existing assessments, we aim to identify strengths, gaps, and opportunities for innovation in science measurement tools. The creation of a publicly available database of these tools will directly support the center’s goal of advancing high-quality science instruction and assessment, informing the development of evidence-based strategies and tools that can be used by educators, researchers, and policymakers to enhance students’ science learning.

Purpose and Research Questions

Purpose:

To systematically identify and analyze existing early science assessments, highlighting patterns and gaps in current measurement instruments.

Research Questions

RQ1: What assessments measure science learning of students in the early grades?

RQ2: What domains and dimensions (e.g., disciplinary core ideas, science and engineering practices and crosscutting concepts) do the assessments measure? a. What assessments measure more than one dimension of science learning?

RQ3: Which assessments have evidence they are suitable for students in a variety of settings? How are measures designed and validated for classrooms with students from non-dominant cultural and language backgrounds and that may include students with disabilities?

Background

Existing early science assessments vary in quality and focus, and many do not align with new science education frameworks. A clear understanding of what assessments exist is necessary  to guide the development of better tools for measuring science learning in young children.

Study Design and Methods

The landscape analysis begins with a scoping review to establish collection methods and define key features for documenting existing science assessments. We apply refined inclusion criteria to identify and categorize each assessment, documenting their essential features, such as administration mode, psychometric properties, and sample characteristics.

Assessments are selected based on a carefully designed framework considering key factors such as content focus, intended purpose, unit of analysis, age range, development process, language, geography, and publication year.

To identify relevant assessments, we employ a strategic, multi-layered search approach, including:

Google Search and Google Scholar: Leveraging advanced search functions to identify assessments related to early science learning.

Snowballing techniques: both backward and forward snowballing are used to expand the search, reviewing citations from key meta-analyses and scoping reviews and using “Cited by” references to identify newer assessments and validation studies.

Expert outreach: We ensure that no relevant assessments are missed by consulting experts in the field to gather additional insights.

Once assessments are identified, they are systematically reviewed and categorized based on key features. The resulting database will provide direct access to assessment materials, offering a powerful resource to help researchers and policymakers select tools that best fit their specific needs.

Outcomes

The study will result in the first publicly available, searchable database of early elementary science assessments in the U.S. The robust framework used to develop the database will provide a solid, evidence-based foundation for the next generation of science assessments for young learners.

Timeline

We anticipate the database will be completed in fall 2025.

Research Team