Publicly funded PreK programs are under enormous pressure to prove what children know. Funders want outcome data. States want compliance reports. Kindergarten teachers want transition information. And in the middle of all of it, your teaching staff is doing their best—clipboard in hand, trying to document meaningful moments from a room full of four-year-olds who won’t slow down long enough to be observed.
Here’s the hard truth: for many programs, the data coming out of that process isn’t reliable enough to act on.
That’s not an indictment of teachers. It’s an indictment of a system that asks them to infer skill mastery from scattered moments, then submit those inferences as evidence.
The Observational Assessment Problem Nobody Wants to Say Out Loud
Observational systems were built on a sound premise: young children show what they know through play and daily interaction, not through sit-down tests. That premise is correct. The execution, at scale, is where things break down.
Here’s what actually happens when teachers use observational systems in real classrooms:
- Evidence collection is uneven. A child who gravitates toward the teacher gets documented. A quieter child may complete the same skill with less fanfare—and less documentation.
- Checkpoints become scramble weeks. The evidence that should be collected continuously tends to get caught up on at the end of each marking period. “Best-guess” scoring is more common than programs would like to admit.
- Rater variability is real. Research on observational systems consistently shows that scores can vary based on who’s doing the observing, not just what the child actually knows.
- The data is hard to aggregate. When every teacher is documenting differently, comparing outcomes across classrooms, sites, or cohorts becomes an exercise in comparing apples to interpretations.
None of this means you should abandon your observational system, particularly if it’s required for compliance. It means you should stop expecting it to do a job it wasn’t built to do.
What Observational Assessments Can’t Tell You
Observational tools are genuinely good at capturing the whole child: social-emotional development, approaches to learning, engagement patterns, and nuanced behavioral growth that resists quantification. That value is real.
What they struggle to deliver is consistent, comparable, skill-level data across classrooms and time points; the kind of data that answers: Can this child identify letter sounds? Do they understand one-to-one correspondence? Where are they relative to kindergarten readiness benchmarks?
For those questions, you need children to demonstrate skills directly, not be observed while they might.
What a High-Quality Direct Assessment Actually Looks Like
The reason direct assessments have had limited traction in PreK isn’t because the concept is flawed. It’s because most direct assessment tools weren’t designed for this population.
Sit-down assessments built for older students are developmentally inappropriate for three- and four-year-olds. Attention spans vary. Anxiety spikes. Children who would otherwise demonstrate mastery in a playful context shut down when the setting feels like a test.
The design challenge is real: how do you get a reliable signal from a child who isn’t yet a reliable test-taker?
Khan Academy PreK Assessments were built to answer exactly that question. Instead of putting children through a formal testing experience, the assessments use brief, playful interactions—the same kinds of activities children are already comfortable with in the Khan Academy Kids app. Children complete assessments in roughly five minutes, in English or Spanish, across math, literacy, receptive language, and executive function. Because the experience is engaging and age-appropriate, completion rates are high and results are representative—the data reflects what children actually know, not how well they tolerate assessment conditions.
Importantly, the results flow directly into two kinds of reporting: teacher-level views designed to inform instruction and small-group planning, and administrator-level reports built for program monitoring, compliance support, and cross-site comparisons. The same data. Two lenses. No duplication of effort.
This Is Not a Rip-and-Replace Argument
If your program uses an observational system, especially if it’s required by your funder or authorizer, you don’t have to choose between compliance and quality. You can have both.
The case for adding a direct assessment layer isn’t that observational systems are worthless. It’s that direct assessment produces a different kind of evidence—more consistent, less rater-dependent, easier to aggregate—that makes your observational data more defensible, not less.
Programs using Khan Academy PreK Assessments during the current pilot have found that brief, play-based direct checks complement their existing documentation workflows. Teachers describe having clearer signals for small-group instruction. Administrators describe finally having data they can compare across classrooms with confidence.
One pilot educator put it plainly: “The Khan tool gives us the exact data we need on where children are—without us having to go through other assessment methods.”
The Window to Pilot Is Now
Khan Academy PreK Assessments are currently available as a two-year free pilot for publicly funded PreK programs in SY 2026–27 and 2027-2028. If your program serves three- and four-year-olds and you’ve been frustrated by the limitations of observation-only assessment—or if you’re simply looking for a way to reduce teacher documentation burden while improving the quality of your readiness data—this is a timely opportunity.
Join the pilot interest list →
Your teachers are already doing the hard work. They deserve data that’s as rigorous as their effort.
Khan Academy PreK Assessments are currently in pilot, validated in partnership with MDRC across 1,000+ students in five states. Free beta access is available for publicly funded PreK programs for SY 2026–27.
