Building a Defensible Pre/Post Assessment in a Rotation Schedule
Ms. Okafor teaches K-5 general music in two buildings: 620 students, 45-minute classes on a six-day rotation, so she meets each class about 24 times a year. Her principal has asked every special-area teacher to submit a student learning objective with pre- and post-assessment data by the end of the second quarter, which gives six class meetings per section. The target standard reads: Students will respond to music using appropriate music vocabulary. Ms. Okafor's gradebook currently contains only participation marks. Three students in the grade she plans to assess have individualized education programs: one has a significant cognitive disability and communicates with a combination of speech and a picture symbol system, one has ADHD and takes medication that wears off by early afternoon, and one has a hearing loss and uses an FM system paired with the teacher's microphone. Her classes are 45 minutes with no aide. She has one class set of tablets that must be shared with the art teacher.
1. Design a pre/post assessment that fits one 45-minute period, actually measures the standard, and can be defended to the principal.
Read the worked answer
Choose a single grade, say fourth, and design backward from the verb respond. A matching or definition quiz would measure recall of terms, not the ability to respond to heard music, so it is invalid for this standard.
The task, in one 45-minute period:
- Listening response (about 15 minutes). Play a 75-second excerpt with clear contrasts - a march, or a movement with a tempo and dynamic change - twice. Students answer three prompts on a half sheet: name two elements you hear and say where you hear them; describe one thing that changes and what the performers did to make it change; choose a word for the mood and give one thing in the music that made you choose it.
- Performance response (about 15 minutes). Individually or in pairs, students echo a four-beat rhythm and sing a three-note pattern. This takes about 20 seconds each and covers the perform side of respond.
The rest of the period is transitions, directions, and collection - about 15 minutes. Pre-test in meeting one, post-test in meeting six.
Rubric: four points across three criteria - accurate use of vocabulary, evidence drawn from the recording, and accuracy of the rhythmic and tonal response. Same excerpt, same prompts, same rubric both times. Score without names visible.
Defensibility: the assessment asks students to do the thing the standard describes, it is short enough to survive a rotation schedule, the performance component keeps it from being a writing test, and the pre/post design lets her report growth rather than a single snapshot.
2. What accommodations do the three students with IEPs need, and which adjustments would damage the validity of the assessment?
Read the worked answer
The principle is to change access without changing the construct. Each accommodation below preserves the requirement that the student respond to heard music with vocabulary.
Significant cognitive disability: keep the same excerpt and the same three criteria, but reduce the output demand. Use two prompts instead of three, supply a symbol and word bank with pictures for fast/slow, loud/soft, high/low, and allow pointing, signing, or an adult scribe to record the response. The IEP team decides whether the standard itself is modified; if it is, report the score as modified rather than proficient.
ADHD: give the prompts in writing before the excerpt plays, seat the student away from the door and near but not on top of the speaker, allow the response to be given orally to the teacher during the performance station, and permit completion across two sessions if needed. Do not reduce the number of criteria.
Hearing loss with FM: cable the audio source directly into the FM transmitter rather than relying on the microphone, and check the equipment with the teacher of the deaf before both the pre- and post-test. Provide the vocabulary cards visually, seat the student with a clear view of the teacher, and use the same excerpt both times so the listening conditions are comparable. Do not substitute a written-only version - that would remove the listening construct entirely.
Adjustments that would invalidate the results: excusing a student from the listening portion, scoring participation instead of response quality, or having an adult supply the vocabulary during the response. Document exactly which accommodations were used on which date.
3. How should the results be reported to the principal and then used instructionally?
Read the worked answer
Report at the criterion level, not as a single average. For each class, give the percentage of students proficient on vocabulary, on evidence, and on the performance response, pre and post, alongside a two- or three-sentence description of what was taught between the two points. A single composite score hides the useful information: a class that improved in naming elements but not in citing evidence needs a different next unit than one that improved in neither.
State the limits honestly. Six class meetings is a short treatment, there is no comparison group, the assessment is one period long, and the sample is a single grade in one building. The data supports a claim of observed growth in a specific class, not a causal claim about the teacher or the program. Saying so makes the report more credible, not less.
Use it: if vocabulary accuracy grew but evidence did not, build the next cycle around describe what you hear and why, with sentence frames and paired listening. Share an anonymized exemplar that shows strong evidence with the principal and with colleagues.
Logistics: keep the scored work in the district system, remove names from anything shared, and get written family permission before using a student response as a published exemplar. Return feedback to students within a week so the assessment functions as instruction rather than compliance. Finally, resist using the results to rank special-area teachers against one another; class composition and schedule differ too much for that comparison to mean anything.