The Gradient That Nobody Got Right
You are teaching a Form 2 geography class of 46 learners. The unit on map interpretation ended with a short test on a 1:50,000 topographical extract. Twenty-nine learners calculated the gradient between two points incorrectly. The scripts show three patterns: dividing horizontal distance in centimetres by the vertical interval and getting answers such as 0.04; forgetting to convert map distance to ground distance; and one learner writing that a river 'flows uphill from the 1,200 m contour to the 900 m contour'. The contour interval on the sheet is 20 m, yet several learners used 50 m, apparently carried over from an earlier exercise. You have 46 copies of the extract, but the photocopier will produce only 20 more before the toner runs out. The next unit, settlement patterns, is timetabled to start next week, and your head of department expects the scheme of work to be followed. You have one 40-minute lesson before that start.
1. What is the most likely root cause of the gradient errors, and what will you do about the scheduled settlement topic? Justify the decision.
Read the worked answer
Diagnosis. The errors are procedural rather than conceptual. Learners can locate the two points and read contours, but they collapse the ratio by mixing units. Getting 0.04 means they divided 2 cm by 50 m without converting, so the idea of gradient is intact while unit conversion is not. The wrong contour interval is a reading habit: learners did not check the legend before starting.
Decision. Do not restart the unit and do not push settlement patterns back by a week. The evidence supports one targeted 20-minute re-teach followed by distributed practice. A 63% failure rate on a single procedure is best fixed by short, focused instruction plus retrieval, not by repeating the whole topic. A full reteach would cost two lessons and still would not address the specific error.
Sequence. Open with a two-minute 'same units' demonstration: 4 cm on a 1:50,000 map is 2,000 m, so a 200 m rise gives 200 ÷ 2,000 = 0.1, or 1 in 10. Require learners to state the contour interval aloud from the legend before every calculation. Then begin settlement patterns as planned, but start each of the next three lessons with a five-item gradient starter. This keeps the scheme of work on track while giving the procedure the spaced practice it needs.
2. Design a 15-minute activity that works with only 20 photocopies for 46 learners.
Read the worked answer
Grouping and resources. Pair learners so that each pair shares one extract; 20 copies therefore serve 40 learners, with the remaining six joining pairs as checkers. Write the two points to be used on the board so no time is lost choosing them.
Activity sequence (15 minutes).
- (3 min) Each pair writes the contour interval and the map scale at the top of the sheet, then converts 1 cm to ground distance.
- (5 min) Pairs calculate the gradient for one assigned pair of points. Assign different point pairs across the room so answers can be compared.
- (4 min) Pairs exchange sheets and mark against a worked solution you write on the board. Each checker reports the most common error seen.
- (3 min) Exit ticket: one gradient calculation at a different scale, 1:25,000, collected at the door.
Why this works. The activity targets the two identified errors, unit conversion and legend-checking, and uses peer marking to give feedback at scale when you cannot read 46 scripts in a week. Assigning different points prevents copying and produces a spread of answers that makes the conversion step visible. The exit ticket uses a new scale, so it tests transfer rather than memory of the worked example.
3. How will you judge whether the intervention worked, and what evidence would make you reteach fully?
Read the worked answer
What to measure. Assess the two sub-skills separately: converting map distance to ground distance, and applying the gradient formula. A single combined item cannot tell you which part is still broken.
Instrument. A four-item exit ticket: one pure conversion item, one contour-interval reading, one full gradient at 1:50,000, and one at a different scale. Mark each item 0 or 1 and record item-level scores, not just totals.
Decision rule set in advance. If at least 80% of learners get the conversion item right and at least 70% get both full calculations right, proceed with settlement patterns and keep a weekly starter. If fewer than 60% get the conversion item, the problem is deeper than a reminder: reteach with a concrete model, using a marked string laid on the floor so learners walk the horizontal distance while holding a card for the vertical rise. If scores split by scale, reteach ratio as a general idea instead of drilling 1:50,000.
Why a rule in advance. Choosing the threshold after seeing the scripts invites adjusting the standard to protect the timetable. Fixing it now keeps the decision about learning rather than coverage.