JC1 H2 Maths Bukit Timah | Vector Projection

Did you know? The number describing how far a vector points in a chosen direction can be negative, even though the length of its projection cannot be negative. That small distinction helps explain why the dot product of two vectors can be positive, zero or negative—and why “negative length” is not a valid answer.

For families seeking a JC1 H2 Mathematics Tutor Bukit Timah, scalar products are a useful test of whether students understand what a vector represents. A vector has magnitude and direction. A numerical component along a chosen axis is directional information, whereas a length records only how large something is. Singapore’s 2027 H2 Mathematics 9758 syllabus includes scalar products, vector products, angles between vectors and the geometric meaning of a·n̂ when n̂ is a unit vector.

This guide explains one practical issue thoroughly: What is the difference between a signed scalar component, the projected vector, the dot product and the length of the projection? We start with the answer a student can use immediately, then show 2D and 3D worked examples, error traps, checks, five small proof ideas, practice problems, teaching steps, an FAQ and useful syllabus-aware reading.

Quick answer: a sign describes direction, not negative distance

Let a be a non-zero vector and let n̂ be a unit vector pointing in the reference direction. The scalar a·n̂ is the signed component of a along that direction.

If a points partly with n̂, this number is positive. If a points partly opposite n̂, it is negative. If a is perpendicular to n̂, it is zero. The signed component records orientation relative to the chosen direction.

The vector projection of a onto that unit direction is (a·n̂)n̂. This is a vector, with both magnitude and direction.

The length of the vector projection is |a·n̂|, an ordinary non-negative real number. The absolute value removes the directional sign when we want only a length.

Dot product a·b: a scalar, equal to |a||b|cosθ for non-zero vectors.
Signed component of a along b: (a·b)/|b| when b is non-zero.
Vector projection of a onto b: [(a·b)/|b|²]b for b ≠ 0.
Length of that projection: |a·b|/|b|, which is never negative.
Perpendicular remainder: a minus its vector projection onto b; it has dot product zero with b.

One of the biggest shortcuts to avoid is writing “component = a·b” when b is not a unit vector. The dot product includes b’s magnitude. To obtain a signed component along b, divide by |b|.

Start with the simplest picture: east versus west

Imagine a straight east–west line. Let the chosen positive direction point east. A displacement of three metres east has signed eastward component +3 m. A displacement of three metres west has signed eastward component −3 m.

The displacement length is three metres in both cases. A negative signed component simply tells us that the vector points in the opposite direction from the chosen positive reference.

The mathematical statement is the same in two or three dimensions. We choose a unit direction vector n̂. Taking a·n̂ measures the signed amount of a in that direction. Taking |a·n̂| measures its size without indicating which way it points.

A useful tutoring question is: “If I reverse my reference direction, does the length of the displacement change?” No. The signed coordinate changes sign, but the physical or geometrical length is unaffected.

How the dot product creates the component

For non-zero vectors a and b, let θ be the angle between them, conventionally between 0° and 180°. The scalar product satisfies

a·b = |a||b|cosθ.

If we turn b into a unit vector b̂ = b/|b|, then

a·b̂ = |a|cosθ.

This number tells us how much of a lies along the oriented direction b̂. The cosine is positive for an acute angle, zero for a right angle and negative for an obtuse angle. That is the origin of the sign.

At θ = 0°, the vectors point in the same direction and the signed component is +|a|. At θ = 90°, a has no component along b̂. At θ = 180°, the vectors point opposite ways and the signed component is −|a|.

The magnitude of the component is |a||cosθ|. It cannot be negative, since both factors are non-negative. Cosine carries the direction sign; absolute value removes it when we want length only.

Worked example 1: a negative component, but positive length

Let a = (−2, 1, 0) and b = (3, 0, 0). The vector b points in the positive x-direction.

Its magnitude is |b| = 3, and its unit direction is b̂ = (1, 0, 0). The scalar component of a along b̂ is

a·b̂ = (−2)(1) + 1(0) + 0(0) = −2.

The negative sign says a points partly against the positive x-direction. It does not mean that the projected piece has a negative length.

The vector projection is (a·b̂)b̂ = (−2)(1, 0, 0) = (−2, 0, 0). This vector points in the negative x-direction, as expected.

The length of the projection is √[(-2)² + 0² + 0²] = 2. It is positive. The distinction is now visible: scalar component −2, projected vector (−2, 0, 0), projected length two.

What does the ordinary dot product give here?

Directly, a·b = (−2)(3) + 1(0) + 0(0) = −6. That is not the signed component −2; it is three times as large in magnitude because b has length three.

To recover the component, divide by |b| = 3: (a·b)/|b| = −6/3 = −2. Both routes agree.

The angle also agrees. Since |a| = √5 and |b| = 3, cosθ = (−6)/(3√5) = −2/√5. This is negative, so θ is obtuse. The vector a points partly in a direction opposite to b.

A common incorrect answer is “the projection length is −6” or even “−2.” Both confuse a signed algebraic result with a non-negative geometric length. A correct solution states what quantity is being calculated.

Worked example 2: a non-unit vector and its projection

Let a = (3, 4) and b = (2, 0). First compute their dot product: a·b = 3(2) + 4(0) = 6.

Since |b| = 2, the signed scalar component of a in b’s direction is 6/2 = 3. The direction unit vector is b̂ = (1, 0).

The projected vector is 3(1, 0) = (3, 0). Equivalently, use the direct formula: [(a·b)/|b|²]b = (6/4)(2, 0) = (3, 0). The two computations are identical.

The length of the projection is three. The remaining perpendicular vector is a − (3, 0) = (0, 4), which is perpendicular to b.

Check by dot product: (0, 4)·(2, 0) = 0. This single check confirms the decomposition into a part parallel to b and a part perpendicular to b.

Notice that the dot product was six, whereas the scalar component was three. The two become equal only when the direction vector used in the dot product has unit magnitude.

Why this matters in H2 Mathematics

The syllabus gives special emphasis to the geometric meaning of a·n̂ with n̂ a unit vector. Students sometimes remember the algebraic formula a·b = a₁b₁ + a₂b₂ + a₃b₃ but forget to ask why the unit length matters.

A unit vector is a reference arrow of length one. Multiplying the signed component by it reconstructs the actual parallel vector. If the reference vector had length five and we multiplied by it without correction, we would artificially make the projected piece five times too long.

Normalisation avoids that scale confusion. If the given direction b is not a unit vector, divide by its magnitude to obtain b̂. The signs then have a clear interpretation, and all the resulting lengths are consistent.

This is one reason why understanding unit vectors is a prerequisite for robust scalar-product work.

Worked example 3: a full three-dimensional projection

Take a = (2, −1, 2) and b = (1, 2, 2). These vectors are non-zero.

First find |b|² = 1² + 2² + 2² = 9, so |b| = 3.

Next find the dot product: a·b = 2(1) + (−1)(2) + 2(2) = 2 − 2 + 4 = 4.

The signed scalar component of a in b’s direction is (a·b)/|b| = 4/3. Because the value is positive, a has a component in the same general direction as b.

The vector projection is [(a·b)/|b|²]b = (4/9)(1, 2, 2) = (4/9, 8/9, 8/9).

The projection length is |a·b|/|b| = 4/3. This is consistent with the magnitude of that projected vector: its squared length is (16 + 64 + 64)/81 = 144/81 = 16/9, whose square root is 4/3.

This is a useful reasonableness check. A projection’s length must not exceed the length of the original vector. Here |a| = √(4 + 1 + 4) = 3, and 4/3 is smaller than three, as expected.

Verify the perpendicular remainder

Subtract the projected vector from a:

a − proj_b(a) = (2, −1, 2) − (4/9, 8/9, 8/9) = (14/9, −17/9, 10/9).

Check its dot product with b:

(14/9)(1) + (−17/9)(2) + (10/9)(2) = (14 − 34 + 20)/9 = 0.

So the leftover vector is perpendicular to b. That verifies that we have separated a into a parallel component and a perpendicular component correctly.

We can also check the Pythagorean relationship: |a|² = 9, the projection length squared is 16/9, and the residual length squared is 9 − 16/9 = 65/9. The two perpendicular squared lengths add to 9.

These extra checks are not decorative. They make a multi-step vector answer auditable when the coordinates involve fractions.

What does zero dot product really tell us?

For two non-zero vectors, a·b = 0 means cosθ = 0, so the angle between them is 90°. Their directions are perpendicular.

For example let a = (1, 2, 2) and b = (2, 1, −2). Then a·b = 1(2) + 2(1) + 2(−2) = 2 + 2 − 4 = 0.

Both vectors are non-zero, so they are perpendicular. A projection of a onto b has signed component zero and is the zero vector. Every part of a is perpendicular to the direction b.

There is an important exception: if one vector is zero, the dot product is also zero, but the ordinary angle involving a zero vector is undefined. Do not claim a 90° angle without checking that the vectors are non-zero.

And if the vectors are direction vectors of two different lines in 3D, perpendicular directions do not prove that the geometric lines meet. They may be skew. This is covered separately in how to recognise skew lines in 3D.

A negative dot product does not mean something failed

A learner sometimes sees a negative dot product and suspects an algebra error, especially after several exercises that use acute angles. But a negative answer can be exactly right.

The sign encodes an obtuse angle between the two non-zero vectors. If the angle is greater than 90° but less than or equal to 180°, cosθ is negative, so a·b is negative.

For instance, a = (−1, 0) and b = (1, 0) point in exactly opposite directions. Their dot product is −1. Their magnitudes are both one, so cosθ = −1 and θ = 180°. Their projected scalar component is −1, while the length of that projection is one.

The mathematical response is not to erase the negative sign. It is to interpret it correctly. Negative is a directional fact, not automatically an incorrect magnitude.

Reversing the reference direction: what changes?

Let b̂ be a unit reference direction. If we reverse it to −b̂, the signed component of a changes from a·b̂ to a·(−b̂) = −(a·b̂).

But the vector projection onto the unoriented line does not change. Using −b̂, the projected vector is [a·(−b̂)](−b̂). The two negative factors cancel, leaving (a·b̂)b̂.

This is subtle but powerful. The signed component depends on which direction along the line we call positive, whereas the projected vector onto that same line does not. The geometry of the line has not changed merely because we reversed its orientation label.

For parents helping a student, a physical arrow drawn on paper can make this visible. Reverse the arrow that labels “positive” and ask which reported quantities change sign and which stay the same size.

Projecting a vector is not the same as projecting a point onto an arbitrary line

Our vector projection formula describes a component along a direction. If we want the perpendicular foot from a point X to a line passing through another point P with direction b, we must first form the displacement X − P.

Then the foot H on that line can be written as H = P + proj_b(X − P), using position-vector notation and b ≠ 0.

Why do we subtract P first? Because a line that does not pass through the origin has a different base position. Projecting the point’s position vector directly onto a line through the origin would use the wrong line.

This connection is useful for later H2 three-dimensional geometry, where the syllabus includes perpendicular feet and distances from a point to a line or plane. However, the first skill to master is still the scalar component along a unit direction.

A perpendicular-distance example without extra machinery

Suppose X = (5, 4) and the line passes through P = (2, 0) parallel to b = (1, 0). This is simply the horizontal line y = 0 through P.

Compute X − P = (3, 4). Its projection onto b is (3, 0), so the foot is H = P + (3, 0) = (5, 0).

The perpendicular difference is X − H = (0, 4), whose length is four. Therefore the perpendicular distance from X to the line is 4 units.

Nothing complicated occurred. We formed the displacement relative to the line, retained the parallel part, then measured the leftover perpendicular part.

A student who instead projects X = (5, 4) directly would obtain (5, 0), which happens to be the same foot in this special horizontal-line example. That coincidence is misleading because the line is also y = 0 and passes through the origin. In general, for a shifted line not containing the origin, direct projection of X is wrong.

To see the general danger, take the vertical line x = 2 through P = (2, 0), with direction b = (0, 1). The perpendicular foot of X = (5, 4) is (2, 4). Projecting X onto the line through the origin parallel to b would give (0, 4), which is a different point.

A quick checklist before solving a projection problem

Look carefully at the wording: are you finding a dot product, a signed scalar component, a vector projection, a projected length, or a distance to a line? These terms are related but not identical.

Then check the reference vector. Is b already a unit vector? If not, choose the appropriate normalisation. Do not divide by |b| twice or forget the squared magnitude in the vector formula.

Finally inspect the sign and units. A dot product can be negative; a signed component can be negative; a projected vector can point backward; a length or distance cannot be negative. A zero result may mean perpendicularity when the direction vectors are non-zero.

Dot product: compute a·b.
Signed component along b: divide a·b by |b|.
Projected vector along b: multiply b by (a·b)/|b|².
Length: take absolute value of the signed component.
Perpendicular remainder: subtract the projected vector and check the dot product is zero.

This sequence avoids over-calculating when the question requests only one quantity.

Worked example 4: downward component along a vertical reference

Let a = (4, −3, 0) and b = (0, 5, 0). The chosen b direction is upward along the y-axis.

The dot product is a·b = 4(0) + (−3)(5) + 0(0) = −15. The magnitude of b is five, so the signed component of a along b is −15/5 = −3.

The projection vector is [−15/25](0, 5, 0) = (0, −3, 0). It points downward, opposite to the chosen positive direction. Its length is 3, not −3.

The perpendicular remainder is (4, −3, 0) − (0, −3, 0) = (4, 0, 0). Its dot product with b is zero.

This example makes the interpretation particularly transparent: four units horizontal, three units downward, and a 3–4–5 resultant. The signed y-component is negative while the downward projected length is positive three.

Worked example 5: unit vector first, or direct formula?

Let a = (1, 2, 2) and b = (1, 0, 0). Since b has magnitude one, it is already a unit vector.

The signed component of a along b is a·b = 1. The vector projection is 1(1, 0, 0) = (1, 0, 0).

The perpendicular remainder is (0, 2, 2). Its length is √8 = 2√2. The original vector has length three, and the Pythagorean identity 1² + (2√2)² = 1 + 8 = 9 agrees with |a|².

Now suppose b is replaced by (4, 0, 0). It points in exactly the same direction but has magnitude four. The ordinary dot product becomes four, not one. Yet the scalar component is still four divided by four, equal to one. The vector projection remains (1, 0, 0).

This illustrates why the result should depend on the direction of the line, not on how long we arbitrarily drew the direction arrow.

Ten short practice questions with answers

1. a = (3, 4), b = (2, 0). Find a·b. Answer: 6.

2. For that same pair, find the signed component of a along b. Answer: 6/2 = 3.

3. Find the projected vector of a onto b. Answer: (3, 0).

4. What is its length? Answer: Three.

5. a = (−2, 1, 0), b = (1, 0, 0). What is the signed component? Answer: −2.

6. What is the projected length in question five? Answer: |−2| = 2.

7. If non-zero vectors a and b satisfy a·b = 0, what is their angle? Answer: 90°.

8. Can a negative dot product correspond to an obtuse angle? Answer: Yes. For non-zero vectors, negative a·b means cosθ < 0, so θ is obtuse.

9. What goes wrong if the reference vector b = 0? Answer: Its unit direction b/|b| is undefined. The usual projection-along-b formula is invalid.

10. a = (4, −3, 0), b = (0, 5, 0). Find the projected vector and its length. Answer: (0, −3, 0), length three.

Try asking a learner to justify questions six, eight and nine aloud. They require the concept of signed direction, the geometric meaning of cosine and the importance of a non-zero reference vector.

How to check a three-dimensional answer independently

For non-zero b, the vector projection of a onto b must be parallel to b. This is clear because the formula produces a scalar multiple of b.

The remainder a − proj_b(a) must be perpendicular to b. A dot product check confirms that property, and it catches many sign errors in long 3D calculations.

The Pythagorean relationship is another check. Because the two components are perpendicular, their squared lengths add to |a|². If the claimed projection is longer than a, or the squared lengths fail to add up, revisit the calculations.

A final common-sense check is that reversing b must not change the projected vector onto the same geometric line. If a solution changes the projection vector merely because b was replaced by −b, the orientation sign has likely been mishandled.

A calm three-stage JC Mathematics tutorial plan

Session one: components and reference directions

Begin with east–west and north–south arrows in 2D. Ask the learner to give a signed component and the length of that component separately. Reverse the positive reference arrow and compare what changes.

Session two: calculate from the scalar product

Introduce non-unit vectors, compute a·b, then divide by |b| to obtain the signed component. Compare with the projected vector formula. Use examples with positive, zero and negative dot products.

Session three: transfer to 3D and justify a projection

Work a full 3D example such as a = (2, −1, 2), b = (1, 2, 2). Require a projected vector, a perpendicular remainder, a dot-product check and a statement explaining the sign. The learner then solves a new pair independently.

The aim is not to memorise three almost-identical formulas without knowing which quantity each one represents. The aim is to read the language of a vector question and produce the correct object with a meaningful check.

What should Bukit Timah parents listen for?

A student who says “negative dot product is impossible because lengths cannot be negative” is mixing two valid facts in the wrong way. Lengths are indeed non-negative, but dot products and signed components may be negative.

Ask the student what happens when a vector points west while positive is east. A correct response distinguishes a signed component of −3 from a displacement length of three.

Then move to a 2D vector and ask why a·b must be divided by |b| to obtain a signed component along b if b is not unit length. If that explanation is clear, the student has a better chance of handling an unfamiliar exam question than if they only remember a formula.

If the learner cannot yet explain this, the next lesson should repair the idea of direction, unit vectors or component, depending on the first misunderstanding. Assigning a large stack of cross-product problems would not target the cause.

Frequently asked questions

Can the dot product be negative?

Yes. For two non-zero vectors, a negative dot product means the angle between their directions is obtuse.

Is a signed projection allowed to be negative?

Yes. A signed scalar component records direction along a chosen axis. It is negative if the vector points partly opposite that reference direction.

Can the length of the projection be negative?

No. Length is the magnitude of a vector, so it is always non-negative. Use the absolute value of the signed component.

Why is a·b not always the scalar component of a along b?

Because a·b also includes the magnitude of b. The signed scalar component equals (a·b)/|b|. If b is a unit vector, |b| = 1 and the values coincide.

Does a dot product of zero always mean perpendicularity?

For non-zero vectors, yes. If either vector is zero, the usual angle between vectors is not defined, so the conclusion needs that condition.

If I reverse the direction vector, will the projected vector change?

No, when projecting onto the same geometric line. The signed scalar component reverses, but the unit reference direction also reverses, so the projected vector remains the same.

Is the topic relevant to 2027 H2 Mathematics?

Yes. The 2027 Singapore–Cambridge H2 Mathematics 9758 syllabus includes scalar products, vector products, angles and the geometric meaning of a·n̂ for a unit vector n̂. Teaching order may differ between junior colleges.

JC1 H2 Mathematics Tutor Bukit Timah | Signed Vector Projection vs Length