Every district tracks science scores. Fewer track whether the materials students use all year are built to produce the kind of performance the test actually measures. On the California Science Test, that is where points quietly go missing.
The CAST is not a recall test. It measures the full range of the California Next Generation Science Standards, which are three-dimensional by design: disciplinary core ideas, science and engineering practices, and crosscutting concepts, and each test item is aligned to at least two of those three dimensions. It also includes performance tasks, extended activities that measure a student's ability to integrate knowledge and skills across multiple standards rather than pick a correct answer. In plain terms, a student can know the content and still lose points if they have never practiced doing the science the way the test asks.
The dimension most materials underbuild
Of the three dimensions, the science and engineering practices are the ones traditional materials tend to underbuild. Practice #2 is Developing and Using Models. Not labeling a diagram, not watching a simulation run, but constructing a representation of a system, using it to make a prediction, and revising it when the evidence disagrees.
That is a doing skill, and doing skills are built through repetition, not coverage. A worksheet can deliver a core idea. It cannot, on its own, give a student reps at building and revising a model. When a performance task asks for exactly that, the students who have done it before hold an advantage that has nothing to do with how much they memorized.
What a student-built model leaves behind
Here is the part curriculum directors tend to find most useful. A pre-built simulation leaves behind a worksheet answer. A model the student constructed leaves behind the student's actual reasoning: the components they chose, the relationships they defined, the prediction they made, and the change they made when the model behaved in a way they did not expect. That is observable, gradeable evidence of Practice #2, generated during normal instruction, before anyone sits for a summative test.
That matters for two reasons. First, it gives teachers a formative signal during the unit, while there is still time to act on it, the same role the CAST interim assessments are designed to play, with item-level data available to teachers in the California Educator Reporting System. Second, it means standards alignment stops being a claim on a pacing guide and becomes something visible in student work.
"NGSS-aligned" and actually building the practice are different things
Most materials marketed to districts carry an NGSS-aligned label. On paper, that usually means the core ideas are covered. It rarely means students get sustained practice with the science and engineering practices, which is the dimension a three-dimensional test weights and the one that does not improve from coverage alone.
So the question worth asking any science material is not whether it is labeled aligned. It is this: when a student finishes, what evidence of a science and engineering practice did they leave behind? If the answer is a multiple-choice score, you have evidence of recall. If the answer is a model the student built and revised, you have evidence of the practice the test is built to sample.
What this looks like in practice
On ModelIt!, a student studying gene regulation or the immune system builds the network themselves: chooses the components, defines what activates or inhibits what, runs it, and explains why knocking out one node changes the downstream behavior. No coding, no advanced math. What the teacher collects at the end is not a completion checkmark. It is a record of a student developing and using a model, the exact practice the standard describes, with assessment items embedded in the lessons.
Science scores are a lagging measure of whether students got to practice the science. The materials that move them are the ones that turn a standard into something a student does, and leaves evidence of having done.