How I use assessment and evaluation

This started as a course assignment, but it’s really just a direct account of how I actually work. Assessment and evaluation aren’t a theory I hold, they’re tools I use on every project, and this is what that looks like in practice.

Assessment happens first. Evaluation happens later.

I treat every piece of assessment data as evaluation material

Assessment is collecting evidence about learning. Evaluation is interpreting that evidence to decide what needs to change. I don’t wait for a separate “evaluation phase” to ask what the data means, I read assessment results as they come in, formative or summative, and use them to judge the course as much as the learner. If evidence turns up, I use it.

Only graded assessments matter.

I pay attention to formative results, even ungraded ones

Formative checks often get treated as a formality before the assessment that counts. I use them the opposite way: as an early-warning system. Where are learners consistently getting stuck? Which concepts aren’t landing before we even reach the summative stage? If the same pattern shows up in both, I treat that as a signal about the instruction, not just about the learners who missed it.

A good assessment has the right format.

I judge a question by what it proves, then build to match

Format doesn’t tell you if an assessment is good, the thinking it demands does. These three questions test the same concept and share the same correct answer, but I wouldn’t use them interchangeably:

Which part of the mind acts as the realistic mediator between impulse and consequence?

A hungry student wants to grab food from a coworker’s desk but knows it’s against office rules, so they walk to the vending machine instead. Which part of the mind mediated that?

A patient says, “Whenever I want something, I just take it. I don’t think about rules or consequences.” Which part of their mind is failing to exert control?

The first measures recall, you can memorize the definition without understanding it. The second and third require applying the concept to a new situation. That’s the test I run on my own work, and it’s also why I stopped being against multiple-choice questions on principle. The problem was never the format, it was poorly designed multiple-choice questions. Once I know what a question needs to prove, I pick the format that proves it: projects and applied scenarios when I need to see judgment and revision over time, and smaller tools, matching exercises, scenario questions, structured discussion, performance-based checks, when they fit better than a bigger production would.

Evaluation ends when training ends.

I evaluate past the course boundary

A course can succeed on every internal measure and still not change performance, because passing an assessment isn’t the same as being able to use the skill afterward. Resources, culture, tooling, and leadership support all determine whether someone can act on what they learned. So I don’t stop evaluating at course completion. I look for whether the environment on the other side actually lets the learning stick.

A living statement

Started: January 12, 2026 · Revision 3 · Most recent revision: July 26, 2026

See it applied

This is what it looks like in a real project.

Designing for Competence walks through how this plays out in a staged, defensible assessment for a regulated skills program.

Designing for Competence Back to Approach