Topic Brief: Differentiating Negative and Fractional Indices

Provenance: compiled 2026-09-29. Source is research/veridian/WMA11-verified-facts.md §4.4 (per-topic examiner-report material) and its dated 2026-09-14 lesson-audit entry for differentiating-negative-and-fractional-indices.ts, cross-checked against that lesson file's own header and against signal-frequency-and-boundaries.md and codex-formula-and-method-bank.md in this same folder so this brief does not restate either. No new primary-source extraction was done for this document — every claim below traces to material the facts bank already verified against the real question paper, mark scheme and examiner report PDFs. Not affiliated with or endorsed by Pearson Edexcel.

#SIGNAL Topic Master Brief — Differentiating Negative and Fractional Indices

#What this brief covers

One narrow slice of the WMA11 differentiation strand: questions that require rewriting a root or reciprocal term as a negative or fractional power of x, differentiating it with the ordinary power rule, and then (usually) using that derivative for a further step, most often an exact gradient. This is one of several distinct differentiation sub-topics on this paper — it is not "hidden higher-degree equation from a differentiation set-up" (Oct 2021 Q2, Jan 2025 Q5(c)) and not "tangent/normal at a stationary point" (Oct 2021 Q6(c)); both are covered elsewhere in this technique hub and are not repeated here.

#Series that have tested this exact topic

Seen in 2 of the 5 fully-read WMA11 series this course's own frequency table already uses as its denominator (Jan 2019, Oct 2020, Oct 2021, Jan 2023, Jan 2025 — the only five series read question-by-question rather than spot-checked for one question):

• January 2023, Q1 (5 marks, split 3+2) — a self-contained question: differentiate y = 2 + 10x^(1/2) − 2x^(3/2), then find the exact gradient at x = 2.
• January 2025, Q5(a) (2 of that question's 8 marks) — the differentiation entry point into y = 4x³ + 2/x + 9, whose later parts (b)–(c) move into a hidden-quadratic and tangent-point problem already covered by other lessons in this course.

Jan 2019, Oct 2020 and Oct 2021 each test other differentiation sub-topics instead of this one (see above); none of the three carries a confirmed finding on this specific shape. Two series is a real but small base — the patterns below are "confirmed twice," not an established rule.

#Mark-tariff pattern

Both sightings sit at the low end of this paper's tariffs: 5 marks when the sub-topic is the whole question (Jan 2023), or as few as 2 marks when it is only the opening part of a larger, differently-themed question (Jan 2025). In the Jan 2023 mark-weight reconstruction already run in signal-frequency-and-boundaries.md §1, that 5-mark total was the smallest full-question mark count confirmed on that paper. One structural mark-scheme fact holds across both: the step where a constant term differentiates to 0 rather than "+c" has never carried its own mark in either real scheme — it is folded into a shared M1 or A1 alongside a neighbouring term, never issued as an independent B-mark.

#Recurring context and angle

Both confirmed instances share a shape: a short expression of no more than three power-of-x terms, differentiated term-by-term, with the result then used rather than left standing on its own. Jan 2023 asks directly for an exact numeric gradient at a stated x-value; Jan 2025 folds the derivative into a later tangent-point calculation in the same question. In neither case does the question stop at "find dy/dx" — the derivative is a means to a further result, not the end point.

The two recurring, quotable error types both examiner reports name are procedural rather than conceptual, not calculus errors:

• An unneeded "+ c" tacked onto a correct derivative. Jan 2023: "a significant number of candidates... added + c to their expression." Jan 2025: "inclusion of '+ c' or + 9 as a third term."
• The negative- or fractional-power term mishandled while a neighbouring positive-power term is done correctly. Jan 2025, on differentiating 2/x: "it was more common to fail to obtain 2x⁻²" — some candidates instead reached "2x or just 2."

Jan 2023 separately records candidates unable to "manipulate the indices to surds" once the question asked for an exact rather than decimal gradient.

#If this topic appears, here's what to expect

A short opening part worth 2–5 marks, inside a question that is either entirely built around this skill (Jan 2023's shape) or uses it as the first step toward something else — a gradient, a tangent, a point on the curve (Jan 2025's shape). Expect at least one term needing rewriting as a negative or fractional power before it can be differentiated, and expect the question to ask for something exact, not just the derivative itself. The two live traps both confirmed series actually cost marks on: do not add "+ c" after differentiating, and check the sign and power specifically on whichever term is not a simple positive-integer power — that is where marks were lost both times, not on the positive-power terms sitting next to it.