Proving Matrix Relationships

Proving matrix relationships means using the algebraic rules of matrix multiplication, identity matrices, and inverses to transform an equation into an equivalent one. The key challenge is that matrix multiplication is generally not commutative, so the side on which you multiply matters and the order of factors must be preserved. A strong proof combines valid algebraic steps, dimension awareness, and explicit use of identities such as $AA^{-1}=I$, $A^{-1}A=I$, and $(AB)^{-1}=B^{-1}A^{-1}$.

• Matrix relationships are equations involving matrices, such as $AB=C$ or $AX+B=0$, where the goal is often to prove an identity or make one matrix the subject. Unlike ordinary algebra, matrix equations must respect both order and dimensions, because multiplication is only defined for compatible sizes.
• Equality of matrices means corresponding entries are equal, so a matrix proof is valid only if every transformation preserves the entire matrix equation. This is why each algebraic step must be justified by a rule such as associativity, existence of an inverse, or multiplication by the identity matrix.
• Identity matrices act like the number 1 in matrix algebra: for a suitable square matrix $A$, both $IA=A$ and $AI=A$. These identities are central to proofs because they let you simplify expressions after creating factors such as $A^{-1}A$ or $BB^{-1}$.
• Inverse matrices undo multiplication by a matrix, provided the matrix is invertible. If $A^{-1}$ exists, then $AA^{-1}=I$ and $A^{-1}A=I$, which is the basis for isolating a matrix on the left or right side of an equation.

• Non-commutativity is the main reason matrix proofs differ from ordinary algebra. In general, $AB\neq BA$, so you cannot move a matrix across an equation or swap factors inside a product just because it seems convenient.
• Associativity allows regrouping without reordering: $(AB)C=A(BC)$. This matters because many proofs depend on turning a product like $A^{-1}(AB)$ into $(A^{-1}A)B$, which then simplifies to $IB=B$.
• Inverse cancellation works only on the correct side. For example, from $AB=C$, pre-multiplying by $A^{-1}$ gives $A^{-1}AB=A^{-1}C$, but post-multiplying by $A^{-1}$ would usually not help because $ABA^{-1}$ does not simplify.
• Dimension compatibility is a hidden condition behind every valid step. A proof can be algebraically elegant but still invalid if the matrices being multiplied do not have matching inner dimensions or if an inverse is used for a non-square or singular matrix.

Making a matrix the subject

• Start by identifying where the target matrix sits in the product. If the unknown is on the right in $AB=C$, remove the left factor by pre-multiplying with its inverse; if the unknown is on the left, remove the right factor by post-multiplying with its inverse.
• Work one factor at a time when several matrices appear. For an equation like $PQR^{-1}=I$, isolate $Q$ by first cancelling the left-side factor with $P^{-1}$, then cancelling the right-side factor with $R$ on the correct side.
• Use associativity to justify each simplification. A line such as $A^{-1}AB=(A^{-1}A)B=IB=B$ is better than jumping straight to $B$, because it shows the logical structure of the proof and makes the cancellation valid.
• State invertibility conditions when needed. If you use $A^{-1}$ or $B^{-1}$, you are assuming the relevant matrix is invertible; without that condition, the step is not guaranteed to be valid.

Core technique: To solve $AB=C$, use $A^{-1}AB=A^{-1}C$ to get $B=A^{-1}C$, and use $ABB^{-1}=CB^{-1}$ to get $A=CB^{-1}$.

Proving an inverse-of-product identity

• To prove $(AB)^{-1}=B^{-1}A^{-1}$, a standard strategy is to show that one expression is both a left inverse and a right inverse of $AB$. This works because inverses are characterized by multiplication giving the identity matrix.
• Compute $(AB)(B^{-1}A^{-1})$ and use associativity: $A(BB^{-1})A^{-1}=AIA^{-1}=AA^{-1}=I$. Then compute $(B^{-1}A^{-1})(AB)=B^{-1}(A^{-1}A)B=B^{-1}IB=B^{-1}B=I$.
• Since multiplication on both sides gives the identity, $B^{-1}A^{-1}$ is the inverse of $AB$. This proof is conceptually stronger than memorization because it explains why the order must reverse.

Key formula: $(AB)^{-1}=B^{-1}A^{-1}$, because the factor nearest to $A$ must cancel $A$ first when multiplying back to form $I$.

Side of multiplication matters

• A matrix can be removed only by multiplying on the side where its inverse meets it directly. This is why $AB=C$ leads to $B=A^{-1}C$ through pre-multiplication, while $A=CB^{-1}$ comes from post-multiplication.

• The phrase "cancel a matrix" can be misleading if taken too literally. Cancellation in matrices is not a free action; it is the result of forming an adjacent identity matrix through a carefully chosen multiplication.

Reversing order versus preserving order

• When proving identities involving inverses of products, the order reverses: $(AB)^{-1}=B^{-1}A^{-1}$. This happens because multiplying back to $I$ requires the most recently applied factor to be undone first.

• Students often confuse associativity with commutativity. Associativity lets you change brackets, but commutativity would let you change order, and that second move is generally false for matrices.

• Write every multiplication step clearly, especially where you pre-multiply or post-multiply. In matrix proofs, markers reward valid algebraic structure, so a fully shown step like $A^{-1}(AB)=A^{-1}C$ is safer than jumping straight to the answer.
• Check dimensions before starting. If a product or inverse is not defined, no amount of symbolic manipulation can rescue the proof, and dimension mistakes are a common source of lost marks.
• Use identity formation as your checkpoint. After multiplying by an inverse, explicitly spot the pair that becomes $I$, then simplify $IA=A$ or $AI=A$; this prevents invalid cancellations and keeps the logic easy to follow.
• Preserve order exactly as written. If your working ever changes $ABC$ into $ACB$ or turns $(AB)^{-1}$ into $A^{-1}B^{-1}$, that is a serious warning sign that the proof has become invalid.

Exam habit: After each line, ask "Did I only regroup, or did I accidentally reorder?" Regrouping is often allowed; reordering usually is not.

• Treating matrices like ordinary numbers is the most common misconception. Students often assume they can divide both sides by a matrix or move a factor across an equation, but matrix algebra requires inverse multiplication on the correct side instead.
• Cancelling non-adjacent factors is invalid. In an expression such as $ABC$, you cannot simplify $A^{-1}BC$ unless the inverse is adjacent to its matrix after using associativity; otherwise the intended identity does not form.
• Forgetting existence conditions for inverses leads to incomplete proofs. A matrix inverse exists only for an invertible square matrix, so using $A^{-1}$ silently assumes more than just notation.
• Confusing left inverse with right inverse can cause incorrect conclusions in more advanced settings. In many school-level questions involving invertible square matrices both sides agree, but your algebra should still respect whether you are multiplying on the left or on the right.

• Solving matrix equations is closely connected to proving matrix relationships. The same techniques used to prove identities are used to rearrange equations in linear algebra, transformation geometry, and systems of linear equations.
• The identity $(AB)^{-1}=B^{-1}A^{-1}$ reflects a broader principle about composing and undoing transformations. If one matrix applies transformation $B$ after $A$, then reversing the effect requires undoing $B$ first and then undoing $A$.
• Proof methods with matrices prepare students for more abstract algebraic reasoning. They develop discipline in handling structure, assumptions, and operation rules, which is essential in topics such as determinants, similarity, eigenvalues, and linear mappings.
• Verification is itself a proof strategy in some contexts. If you suspect a formula for an inverse or rearrangement, multiplying out both sides to recover the identity matrix can provide a clean and rigorous confirmation.