8.2 Diagnostic Practice: Errors, Transfer, and Study Planning

Diagnose the work, not the person

Independent practice can show which steps need attention. It cannot supply an official CogAT percentile, diagnose a disability, or prove a permanent limit in reasoning. A useful review explains the specific error and selects a fresh exercise that tests the repair.

Begin with a small set across all nine formats. Learn each direction before adding time pressure. Keep the format label, attempted count, correct count, and reasons for misses. Do not treat an overall percentage as a complete picture if one format is much less familiar than the others.

Use meaningful denominators

Suppose a twenty-item practice set yields twelve correct, sixteen attempted, and four skipped. Accuracy among attempted items is $12/16=75\%$. Accuracy against the full set is $12/20=60\%$. Both are useful, but they answer different questions.

The first concerns performance on work attempted; the second includes completion. Reporting only seventy-five percent hides the four omissions. Neither percentage is an official percentile rank. An official percentile compares a scored assessment with a specified norm group, which a local practice set does not provide.

Sort errors into actionable categories

Categories can overlap. A wrong number analogy may involve both an unsupported rule and an arithmetic error. Record both rather than assigning every miss to carelessness. The correction should show what happened in the work.

Work a diagnostic example

A student sees $(3,8)$ and $(5,12)$, then predicts twelve for input seven by adding five. The arithmetic $7+5=12$ is correct. The problem is evidence: adding five fails the pair $(5,12)$ because $5+5=10$.

Repair the method by testing both known pairs. Double then add two fits them and predicts sixteen. Now test transfer with different values under a stated compound rule. If the student repeats the correct checks on the fresh example, the method has improved. Merely recalling sixteen from the first problem would not establish transfer.

For a folding miss, the corresponding review might identify a punch outside a partial flap. The repair is to count layers at the actual location, then solve a different partial-fold example. Repeatedly doing full half-folds would not directly address that error.

Make a flexible plan

One optional cycle is: learn a format, solve a few untimed items with explanations, review misses, solve fresh transfer items, then try a short timed set. Repeat for the formats that need work. This is a practice structure, not a validated number of hours required for gifted qualification.

A sample week might include a verbal session, a quantitative session, a spatial session, and a mixed review. Change that distribution according to the error record and your available time. If sentence completion is already accurate but folding is unfamiliar, spending equal time on both may be unnecessary.

Short sessions with clear goals can be easier to review than long sessions containing many unexplained guesses. Stop when fatigue makes the work uninformative and resume later. There is no official preparation schedule in the public task descriptions that guarantees a score increase.

Preserve novelty

Repeated questions become easier partly because their answers are familiar. Keep a separate set of fresh examples for checking transfer. Explain the rule before checking the key, and compare your reasoning with the explanation afterward.

The site's practice inventory and a complete official battery are different. A bank of original questions may not contain the same number, distribution, difficulty, or timing controls as the real assessment. Do not label a mixed local set an official simulation simply because its total happens to match 176.

Track change with context

Record the date, item source, format, time condition, and whether questions were previously seen. A higher percentage on repeated untimed questions is different evidence from better performance on fresh timed questions. Keep the conditions visible so comparisons remain honest.

Small samples fluctuate. If performance changes by one item in a ten-item set, avoid interpreting it as a stable battery-level strength or weakness. Look for consistent reasoning improvements across new tasks. When official scores are available, use the actual report and qualified interpretation rather than trying to convert practice accuracy into a Standard Age Score.

Prepare for the actual session

Confirm the assigned date, format, materials, timing, and accommodations with the school or testing provider. Use the examples or tutorial to understand the response procedure. Follow ordinary routines that help you arrive ready; this guide does not prescribe a medical regimen or a performance supplement.

After each practice cycle, choose one next action supported by the work. “Check both number pairs before calculating the target” or “distinguish inside a polygon from inside its bounding box” is concrete. “Become smarter” or “fix a weak brain area” is neither a valid diagnosis nor a useful study instruction.