Choosing a student assessment instrument you will keep using

Co-curricular assessment usually fails from ambition rather than neglect. A division picks the most comprehensive instrument available, administers it once with real enthusiasm, gets a disappointing response rate the second year, and quietly stops. Two administrations, no trend, nothing to report.

Change over time is the only thing anyone actually wants to know, which means consistency beats coverage. A short instrument you will still be running in four years is worth more than a thorough one you abandon after two.

Here is how to pick that instrument.

Name the decision before you pick anything

Write down the decision the data will inform. Whether to expand the program to a second cohort. Where to place a staff line. What to put in front of a provost in March. What to change about the program itself.

If nothing changes based on the answer, you are collecting data for its own sake, and that is the project that gets dropped when the term gets busy. Naming the decision also shortens the instrument automatically, because most items on a comprehensive survey do not inform the decision you actually face.

Check what your campus already administers

Before adding an instrument, ask institutional research what is already in the field. Many campuses run a national engagement or satisfaction survey on a cycle, and it very often already contains items on belonging, involvement, and interaction with staff and faculty.

Reusing it has three advantages you cannot manufacture: it costs you nothing to administer, it arrives with comparison data across peer institutions, and it carries credibility with academic colleagues that a locally written survey will not. Ask whether your program’s participants can be flagged in the existing sample: that one request sometimes replaces an entire assessment project.

Use validated instruments unmodified

Where a validated instrument exists for what you want to measure, use it, and use it as written.

Rewriting three items to fit local vocabulary is the most common way a division destroys the thing it was buying. Comparability with every other administration of that instrument, including your own next year, is the whole reason to choose a validated tool. Edit the items and you have a homegrown survey with someone else’s name on it.

If you need local questions, and you usually do, add them as a clearly separate block after the validated items. Report them separately too.

Take the baseline before launch, or lose it permanently

This is the single most common unrecoverable error in co-curricular assessment: the program launches, it goes well, someone asks for evidence, and the pre-measure that would have made the answer easy was never collected.

A short pre-survey at orientation costs almost nothing in week one and is impossible in year two. Take it even if you are unsure the program will continue, especially then, because a program fighting for renewal is exactly the situation where a baseline is worth the most.

If you have already missed it, the honest options are narrower but real. You can begin a baseline with the next cohort, and you can report reach and implementation now while you wait. What you cannot do is reconstruct a pre-measure after the fact.

Design for the response rate, not the ideal sample

A beautiful instrument with a 15% response rate from a self-selected slice of a self-selected program is not a defensible result.

Administer where students already are: inside a session, at a required check-in, at a program milestone. Build it into the sequence so completing it is part of participating rather than a favor you are asking. Then set a response-rate floor in advance and report the actual rate every time, because a reader who is not told the rate will assume the worst and be right often enough.

Keep it short enough to finish in the time you actually have. Ten well-chosen items completed by most participants outperforms forty completed by a third of them.

Measure the thing you can act on

Two categories of measure are worth more than they cost.

The gap between availability and use. Our State of Higher Ed 2026 research found 84.3% of undergraduates know their school offers career services while only 46% of those have used them: 38.8% of all undergraduates. That is a measure that changes a decision, and almost nobody keeps it. The equivalent for your program is the ratio of eligible students to participating students, disaggregated.

Movement on a small number of defined competencies. Pre-to-post on three or four skills you actually teach, reported as movement among participants and described as exactly that. Where a program you bring in includes its own assessment (the NSLS provides annual program assessments, for instance), treat the result as an input you interpret rather than a finding you forward under your own name, and say whose instrument it was.

Give the number somewhere to sit

An instrument produces a number, and a number needs a reference point before it means anything. For retention comparisons, use your sector rather than the national figure: median full-time retention is 71% across all US degree-granting institutions but 77% at private nonprofit four-year and 63% at public two-year institutions (IPEDS, Fall 2023). A 68% rate is a problem in one setting and an achievement in the other.

For everything else, your own prior year is usually the most honest comparison available, which is another argument for choosing the instrument you will keep.

Do this next

Write the decision on one line, then find the shortest instrument that informs it, then email institutional research to ask what is already in the field. Do those three in that order and you will spend less and report more. Our guide to measuring student outcomes covers what to do with the results, and reporting co-curricular outcomes upward covers how to write them for the audience that will read them.