Directed self-placement (DSP) asks students to choose their own course level after engaging with structured information about what each level requires. Developed by Royer and Gilles at Grand Valley State University in the late 1990s, DSP has genuine research support — but also real limitations that advocates sometimes underemphasize.
When given good information, students often make reasonable placement decisions. DSP eliminates the equity problems of standardized testing, respects student agency, and has essentially no per-student cost. Studies at several institutions found self-placed students performed comparably to test-placed students.
DSP works best at selective institutions with relatively homogeneous incoming populations. At open-enrollment community colleges, students from underrepresented backgrounds tend to underplace themselves — reproducing inequity in a different form. DSP outcomes depend entirely on the quality of orientation materials. And DSP gives instructors no data about individual student writing ability.
Standardized tests like Accuplacer predict grammar recognition, not writing ability. Research consistently shows they misplace roughly one in four students, with errors skewing toward over-placement in developmental courses — disproportionately affecting students of color, multilingual students, and first-generation students.
Both DSP and standardized testing are indirect measures. Direct writing assessment — having students write and evaluating what they produce — is the most valid method. AI-assisted scoring makes it practical at scale, with AI handling the initial pass and faculty reviewing borderline cases. The best placement systems combine methods: GPA for students with strong records, direct writing assessment where placement is genuinely uncertain.
See our guides on Accuplacer alternatives for community colleges and setting up a writing placement test.