Twelve Weeks, Five Children Moving
Ms. Patel teaches 22 four-year-olds. For 12 weeks she has photographed every child's art and coded the photos for three indicators: uses three or more colors, adds detail beyond a shape, and tells a story about the work. At the 12-week review, only 5 of 22 children show change on any indicator. She lays the photos out and notices that about 80 percent show the same die-cut product: a tissue-paper umbrella, a paper-plate snowman, a handprint turkey, each assembled to a teacher-made model. Parents praise the bulletin board and one parent emails asking why there are no letter worksheets. Her principal wants "measurable growth" documented against a rubric before the spring conference. A literacy coach visits in three weeks and has offered to look at her documentation.
1. What is the most likely explanation for the flat data, and what are the limits of her measure?
Read the worked answer
The most likely explanation is that the measure cannot detect learning because the task removed it. If 80 percent of the work is a teacher-designed die-cut product assembled to a model, the three indicators are almost pre-determined: everyone uses the same colors, everyone's detail is the same, and "tells a story" rarely happens because the child did not invent the image. The flat data are a finding about the curriculum, not about the children.
Limits of the measure:
- Ceiling and floor effects. A fixed product produces a fixed score.
- Construct validity. Color count and "adds detail" are weak proxies for what she actually values - intentional representation, choice, narrative, persistence.
- Sampling. Photographs of finished products miss the process: the planning, the revising, the child's explanation.
- No baseline. Without a comparable starting sample or a repeated task, change is hard to claim.
The honest statement is: "This instrument measured compliance with a model. It was never able to show growth."
2. What would you change in the next three weeks, and what evidence would you bring to the coach?
Read the worked answer
Change the task first, then the measurement.
For three weeks:
- Replace one product-based activity per week with an open studio session: a small set of materials, a prompt ("make something about your family," "draw something you did outside"), and no model.
- Keep one repeated task constant: every other week, "draw a story from your weekend," same paper, same medium, same prompt. That gives a matched pre/post comparison against her existing samples.
- Collect the child's dictation verbatim and a 30-second video of the child talking about the work. Narrative growth shows up there.
Evidence for the coach: two matched drawings per child (before and after), a four-point descriptor scale on four dimensions - intentionality of marks, detail, narrative, material handling - plus the child's own words. Ask the coach to score five samples independently so she can check her own reliability.
What she should not do is add more indicators. The problem was never too few measures.
3. How should she respond to the parent who wants worksheets and to the principal who wants measurable growth?
Read the worked answer
To the parent: lead with the child's work, not with philosophy. Show three dated pieces and the child's words across eight weeks - "This is my grandma's house. The red is the door." Explain that worksheets mostly practice following directions, while open drawing builds the fine motor control, planning, and storytelling that later support writing. Offer a concrete invitation: a family studio night, or a one-page sheet of questions to ask at home ("Tell me about your picture" rather than "What is it?").
To the principal: do not argue against measurement; offer a better instrument. Bring a one-page descriptor scale with four dimensions and two anchored examples each, a matched pre/post sample, and a note that the scale takes about three minutes per child twice a term. Use the language of the early learning standards the school already reports against. Say plainly what the old data did and did not show.
Keep the two conversations aligned by using the same artifact - one child's sequence over eight weeks. It is harder to argue with a child's own explanation than with a rubric.