Word-Problem Difficulty with Mixed Regrouping
A third-grade teacher reviews a subtraction unit. On a 12-item quiz, most students solve items like 84 - 27 and 300 - 156. However, on the two word problems (e.g., 'Maya has 203 stickers. She gives 78 to a friend. How many does she have left?'), 14 of 22 students write the wrong operation or subtract the smaller digit from the larger in each column, arriving at answers such as 275. When the teacher reads the same problems aloud and asks students to act them out with base-ten blocks, most of those 14 produce correct results. The teacher wants to close the gap without abandoning the standard algorithm and has three 40-minute math blocks before the unit test.
1. Diagnose the likely source of the errors and explain why the read-aloud/block version succeeded.
Read the worked answer
The pattern is not a computation deficit. Students can execute the algorithm on bare numbers, so place-value and regrouping procedures are largely intact. The failure appears when they must (a) decide the operation from text and (b) hold the quantities in mind while setting up columns. The 'smaller from larger' errors (275 instead of 125) indicate they are not tracking regrouping/decomposition as a meaning, only as a rule applied digit-by-digit.
Success with read-aloud plus blocks points to two scaffolds doing the work: language comprehension of the problem structure and a concrete representation that makes 'borrow one hundred as ten tens' visible. In other words, the difficulty is at the translation stage (text โ quantity โ operation), not the procedural stage. The fix should target translation and connect it explicitly back to the written algorithm, not reteach subtraction facts.
2. Design the sequence for the three 40-minute blocks, naming what happens in each and what you will look for as evidence.
Read the worked answer
Block 1 - Text to quantity. Use 3-4 problems; students underline the quantity and the change, then build the starting number with base-ten blocks before writing any symbols. Model one problem aloud using think-aloud ('I have 203, so I build 2 hundreds, 0 tens, 3 ones'). Evidence: every student's block model matches the stated quantity.
Block 2 - Blocks to columns. Students build, then physically trade one hundred for ten tens to subtract, and record each trade as a written notation above the algorithm. Pair the concrete trade with the written regroup mark so the symbol means something. Evidence: written regroup marks correspond to actual trades, and answers match the model.
Block 3 - Fade the scaffold. Same problem types, blocks available but optional; students choose and justify. Include one problem where regrouping across a zero (e.g., 300 - 156) is required. Evidence: accuracy on transfer items and students' verbal explanation of why they regrouped.
Keep the algorithm throughout; the goal is connecting it to meaning, not replacing it.
3. One student still answers 275 on every word problem even with blocks and read-aloud. What is your next step and what will you avoid?
Read the worked answer
This is a signal for individual diagnosis, not more whole-class repetition. Sit with the student and ask them to read the problem and tell you what is happening in the story before computing. Check whether the issue is vocabulary/comprehension (misreading 'gives away' as combining), a habit of subtracting smaller-from-larger regardless of meaning, or difficulty holding the quantity while writing. Use a short concrete check: have them build 203, take away 78 with trades, and compare to their written work.
Avoid: giving more of the same worksheets (practice won't fix a translation error), labeling it a math disability before ruling out language and instructional causes, and removing the blocks prematurely. Provide a targeted scaffold (a sentence frame or a quantity/change/result organizer) and monitor with a few parallel items. If the pattern persists across representations and contexts, consult the intervention/specialist team with the evidence you collected.