Skip to content Skip to navigation

SNAP Report: An Analysis of Existing Science Assessments

An Analysis of Existing Science Assessments and the Implications for Developing Assessments for NGSS

The analysis of the landscape of existing assessments was conducted to to identify existing assessments that showed promise as models for development of assessments for NGSS. SNAP created a collection of hundreds of assessment resources that were designed to engage students in scientific reasoning, not just evaluating rote knowledge. Even among these promising resources, we identified few examples of assessments that met the needs of NGSS.

In our report we outline four major lessons learned from this analysis about ways that development of new assessments must change to support implementation of NGSS and we include numerous sample items to illustrate each lesson.


1. Assessments should integrate multiple dimensions

Most existing assessments evaluate a single dimension at a time. Even items that at first appear to evaluate how students use their knowledge to answer a question or solve a problem often only require one dimension to get to the correct answer. Assessments for NGSS should evaluate how students use their knowledge as they engage in a practice, or even multiple practices.

2. Assessments should focus on the big ideas in science

The NGSS emphasize the big ideas and themes in science; details and terminology are included only as they are central to making sense of phenomena. Most existing assessments evaluate recall of isolated facts and technical vocabulary, which are much easier to evaluate. Many of these facts are not the focus of the new standards. Assessments for NGSS should evaluate conceptual knowledge, not just easily assessed discrete facts.

3. Assessments should evaluate the full range of practices

For some science and engineering practices, the practice is entirely new to standards. For others, only narrow segments of practice have been assessed. For instance, numerous items evaluate students on identifying the variables in a fair test, but few evaluate how students decide what type of data could be used to design a  fair test or how the data should be recorded. Assessment development should target the full range of skills that make up each practice.

4. Assessments should use a variety of task formats

Most existing assessments probe a very small piece of content or a practice, and therefore provide limited insight into how students draw on their knowledge and skills to make sense of real-world phenomena. Well-designed performance tasks are able to probe much more deeply into students’ reasoning and their ability to  investigate science and engineering questions and design solutions. Assessments for NGSS should employ a variety of formats to create a complete picture of students’ progress.


Analysis of existing assessmentsRead the full report