Practise the mathematics behind this
The practice portal carries the algebra, logarithms, calculus and vector work this article uses, with worked solutions that explain the reasoning rather than just the answer.
Open the practice portal →Only about a third of students beginning an introductory physics course can add two vectors correctly, and instruction does not reliably fix it. The reason is that vectors are taught as procedures rather than as questions. Here is what the operations are actually asking, and a worked orbital burn in which pointing the engine the wrong way costs 99 percent of the result.
I have watched capable sixteen year olds answer a vector question correctly and then, three minutes later, get a nearly identical question badly wrong. It is not carelessness and it is not a gap in effort. It is that most students are taught vectors as a set of procedures for manipulating pairs of numbers, and are never told what the operations are actually asking. This article is the version I wish somebody had given me before I started using vectors for a living.
The physics education research on this is unusually clear and unusually depressing, so I will start there rather than pretend the difficulty is unique to any one student.
Richard Knight tested more than two hundred students beginning an introductory physics course and found that only about one in three could add two vectors correctly in two dimensions. Nguyen and Meltzer ran a larger study across several institutions and found the same pattern persisting after instruction: a substantial proportion of students could not add or subtract vectors graphically even having just been taught to. Flores, Kanim and Kautz found that students who could perform vector operations in a pure mathematics context frequently failed to apply them in a mechanics context, which tells you the difficulty is not arithmetic.
The consistent finding across this literature is that students treat a vector as a pair of numbers with an arrow drawn near it, rather than as a single object with a direction that means something. Once you see that, most of the specific errors stop looking random.
A student who thinks of a vector as two numbers will happily write that the magnitude of (3, 4) is 7, because adding is what you do to numbers. The correct statement is:
and the reason is that the two components are at right angles to each other, so they do not accumulate the way lengths along a line do. That is Pythagoras, which every student in the room already knows. What they have not been given is the reason Pythagoras is the relevant theorem here.
Here is the sentence I most want a student to carry away. The dot product is not a rule for multiplying vectors. It is the answer to a single question: how much of this vector points along that one?
The is doing all the work. If the two vectors point the same way, is 1 and you get the full product. If they are at , is zero and you get nothing at all. If they oppose, you get the full product with a minus sign.
Now look at the definition of work in mechanics:
Students are usually told that work is force times distance, and then told that you have to use the component of the force along the motion, as if that were an awkward correction. It is not a correction. It is the entire content of the idea. Work asks how much of the force pointed the way the object went, and the dot product is the operation that asks that question. Once a student sees this, the standard puzzle of why carrying a heavy bag horizontally does no work stops being a trick and becomes obvious: the force is vertical, the motion is horizontal, is zero.
The cross product is the mirror image, and students often meet it as a second, unrelated way of multiplying vectors with a confusing right hand rule attached.
Where the dot product has a cosine and is largest when the vectors are aligned, the cross product has a sine and is largest when they are perpendicular. The dot product asks how much they agree. The cross product asks how much they are at odds. That is why one gives you a number and the other gives you a vector pointing out of the plane they define: the answer to "how much are these two at odds" needs an axis to be about.
This is why torque and angular momentum are cross products. Pushing a door at the hinge does nothing, pushing at the handle perpendicular to the door does the most, and the sine captures exactly that. In orbital work the relevant quantity is:
and the fact that this vector stays constant for a body orbiting under gravity alone is the reason an orbit stays in one plane. A student who understands the cross product as "how perpendicular are these" has been handed the reason planets do not wander out of the ecliptic, for free.
In my previous work, getting a vector direction wrong is not a lost mark. Here is the cleanest example I know, and it is one a sixteen year old can compute.
A spacecraft is in low Earth orbit moving at about 7,660 metres per second. It fires its engine and produces a velocity change of 100 metres per second. The new velocity is a vector sum:
so the new speed comes from the cosine rule:
where is the angle between the burn and the direction the spacecraft is already moving. Now put numbers in for three choices of that angle.
Burn straight ahead and you gain the whole 100 metres per second. Burn straight backwards and you lose the whole 100. Burn sideways, at , and you gain 0.7 metres per second. Not seventy. Not fifty. Under one.
The propellant is identical in all three cases. The engine ran for the same time and burned the same mass. The only difference is direction, and direction cost 99 percent of the result. This is what I mean when I say vectors are not a topic. A student who has done this calculation understands, in a way that no verbal explanation achieves, that a vector quantity carries information that its magnitude alone does not.
The same reasoning explains something students often find counterintuitive: why a burn is more valuable when a spacecraft is moving fast. Kinetic energy goes as the square of speed, so adding a fixed velocity change to a large existing velocity adds more energy than adding it to a small one. This is why manoeuvres are performed at the lowest point of an orbit where the vehicle is moving fastest, an effect named after Hermann Oberth. It is a direct consequence of the vector sum above, not a separate rule.
To raise an orbit you fire twice. Once to stretch the circle into an ellipse, and once at the far end to round it back off. This is the Hohmann transfer, and Prussing proved in 1992 that for two impulsive burns between circular coplanar orbits it is genuinely the cheapest possible route, which is a satisfying thing for a student to know: the obvious answer is, for once, provably the optimal one.
What matters for this article is that both burns are prograde, meaning along the existing velocity, and both are performed at specific points. Swap where you fire them and you do not get a slightly worse transfer. You get a different orbit entirely, possibly one that intersects the atmosphere. The magnitudes were right. The directions were not. In spacecraft operations that distinction is the difference between a mission and an incident.
Three changes make most of the difference, and none of them requires extra syllabus time.
The first is to stop introducing the dot and cross products as two ways of multiplying and start introducing them as two different questions. Students who know which question an operation answers can reconstruct the formula. Students who only know the formula cannot recover the question.
The second is to insist on drawing before calculating. The research is emphatic that graphical vector addition is where students fail, and the fix is not more component arithmetic. It is making the student put the arrows head to tail on paper and estimate the answer before touching a calculator, so that a wrong answer looks wrong.
The third is to give the vector a job. Vectors in a textbook chapter are labelled a and b and mean nothing. Vectors in the burn calculation above decide whether a spacecraft reaches its orbit. My experience is that the second kind is remembered and the first kind is not, and I do not think that is a matter of enthusiasm. A student who has seen direction cost 99 percent of a result has a reason to be careful that no amount of exam pressure supplies.
Because vectors are usually taught as procedures for handling pairs of numbers rather than as single objects whose direction carries meaning. Knight found only about a third of students entering an introductory physics course could add two dimensional vectors correctly, and Nguyen and Meltzer found the difficulty persisted after instruction. Flores and colleagues showed students who could do the mathematics often failed to apply it in mechanics, so it is not an arithmetic problem.
It answers one question: how much of one vector points along another. That is why the cosine appears. When two vectors are aligned the cosine is one and you get the full product; at ninety degrees it is zero and you get nothing. Work is exactly this question applied to force and displacement, which is why carrying a bag horizontally does no work against gravity.
The dot product uses a cosine and is largest when two vectors are aligned, so it measures how much they agree. The cross product uses a sine and is largest when they are perpendicular, so it measures how much they are at odds. That is why the cross product returns a vector: the answer needs an axis to be about. Torque and angular momentum are cross products for exactly this reason.
Because velocities add as vectors, not as numbers. A 100 metre per second burn applied to a spacecraft already moving at 7,660 metres per second gives the full 100 if fired along the existing velocity, loses 100 if fired backwards, and gains only about 0.7 if fired sideways. The propellant spent is identical in all three cases.
The observation that a given burn adds more energy when the spacecraft is already moving fast, because kinetic energy grows as the square of speed. This is why manoeuvres are performed at the lowest and fastest point of an orbit. It follows directly from adding velocities as vectors rather than being a separate rule.
Insist on a drawing before a calculation. Vector errors show up visually long before they show up in components, and a student who has estimated the answer from a head to tail diagram will notice when the arithmetic disagrees. Then attach the vectors to something real, so that getting the direction wrong has a visible consequence.
The practice portal carries the algebra, logarithms, calculus and vector work this article uses, with worked solutions that explain the reasoning rather than just the answer.
Open the practice portal →