The answer lives in this podcast

Answer extracted from the Education: Joining the Dots podcast — listen to the full episode below.

🎧 Listen to the episode on Listenly

Why internally designed tests fail to compare student performance fairly

6 September 2026 Education: Joining the Dots · Becky Sinjin

← See the Education: Joining the Dots fiche

Internally designed tests are inherently inconsistent in difficulty, making them poor tools for comparing performance across different students, time periods, schools, or subjects. Standardized tests solve this problem by converting raw scores onto a scaled distribution, enabling fair comparison against national benchmarks and meaningful tracking over time.

The core problem: no common yardstick

When schools create their own assessments, they have no way to ensure that one test is as challenging as another. A vocabulary test designed in one classroom may be significantly harder or easier than an identical test written by another teacher, not because of teaching quality, but simply due to how questions are phrased, selected, or weighted. This inconsistency makes it impossible to know whether a student's score reflects their actual learning or just the difficulty of the specific test they took.

The real damage happens when educators try to use these internally designed scores for comparison. You cannot reliably compare one student's performance to another's if both took different tests, even if those tests supposedly cover the same material. You cannot track genuine progress over time if assessment difficulty fluctuates from term to term. And you certainly cannot compare results fairly across different schools or subjects, where internal assessment standards vary widely.

How standardized tests create a level playing field

Standardized assessments work differently. They convert raw scores—the number of questions a student gets right—into a scaled score anchored to a standard distribution. This process, refined through rigorous statistical analysis, allows a student's performance to be meaningfully positioned against all other students who took the same test nationally.

This scaling mechanism unlocks genuine insight. A score of 75 on a standardized test tells you not just that a student answered 75% of questions correctly, but exactly where they sit relative to their peers across the entire country. It enables you to measure real growth over time—because the scale remains constant. And it allows fair comparison across different subjects and schools, because all scores are calibrated to the same standard. As Becky Sinjin explains in the episode, this consistency is what transforms raw data into actionable intelligence.

"The standardized test results are there to support the teacher judgment, to contribute to the teacher judgment, not to replace it."

This clarification—that standardized results support rather than replace professional judgment—highlights why the comparison gap matters. Teachers need reliable data to make better decisions faster, whether in selecting intervention students, tracking cohort progress, or identifying learning gaps. A standardized test provides that foundation; internally designed assessments cannot.

The broader implication, discussed in depth across this podcast episode, is that assessment design is not simply an operational detail—it directly shapes what educators can know about their students and, by extension, how effectively they can teach.

See also