Finish the solution
The expression
is equivalent to for all values of , where and are constants. What is the value of ?
First steps
- Make a list of inputs with
x_1=[1...10]. - Fit both sides with
(x_1+p)^2-(x_1-4)^2~12x_1+q.
Why this matters on the SAT
Some SAT questions say that two expressions are equal for all values of and hide one or two unknown constants inside them. Finding those constants by hand means expanding every product and matching the terms one power of at a time. It's easy to drop a sign along the way.
Here's a typical question.
SAT example
The expression
is equivalent to for all values of , where and are constants. What is the value of ?
For all values of means the two sides give the same number whatever is. So hand Desmos a whole list of -values, and ask for the and that make the sides match at every one of them:
x_1=[1...10]
(2x_1+p)(x_1-3)-(x_1+q)(x_1+1)~x_1^2-10x_1-11
Desmos reports and , so
By hand, you'd expand both products and match three pairs of coefficients. Here Desmos does all of that in two short lines.
Reach for a list fit when a question says two expressions are equal for every value of the variable and asks for a constant, or for something built from constants, like or .
Then check whether a shorter route settles it first.
The question says equivalent, for all values of , true for all or has infinitely many solutions, and the answer is a constant or a combination of constants.
Finding the constants by hand means expanding several products, clearing fractions or matching three or more coefficients.
Matching one coefficient by hand wins when a single product gives the constant away. In , the terms are , so .
Graph overlap wins when there's no constant to find and the question asks which expression is equivalent. That's the job of equivalent expressions by graph overlap.
Comparing coefficients wins for a system of two lines in and with infinitely many solutions, as in how many solutions.
The fit itself is the custom regression you already know, with one change. In custom regression with multiple conditions, each fact was one equation. An identity gives you an equation at every input, so you supply the inputs as a list.
Which would you fit with a list of inputs? (A) Which choice is equivalent to ? (B) The expression is equivalent to for all , where and are constants. What is ?
An identity is an equation that's true for every value of the variable, not just one or two. SAT questions signal one in a few ways:
The last one is the same idea in disguise. An ordinary equation like is true for one value of . If an equation is true for infinitely many values, its two sides are the same expression, so it's an identity.
Then look at the letters. Some are constants you have to find: the question names them, like where and are constants, and asks for their value or for something built from them. Others can stand for any number, like the in for a nonzero constant , which expression is equivalent…, where the answer choices still contain . There's no constant to find there, so it isn't a fitting job: compare the algebra, as in the graph overlap lesson.
The variable doesn't have to be . If the question uses , you'll still type x_1 in Desmos, the same way you renamed variables in the graph overlap lesson.
In the equation , is a constant, and the equation has infinitely many solutions. Is this an identity question? Which letter do you have to find?
The fit takes two lines.
x_1=[1...10]. Square brackets make a list, and the three dots fill in every whole number from to .x_1 wherever the question has , then the tilde, ~, where the equals sign would go, then the right side. The tilde tells Desmos to find the constants that make the two sides match.Desmos then shows the constants it found under the regression line.
Why the list matters. An identity has to hold at every input, and each input gives Desmos one more equation the constants must satisfy. With ten inputs, the constants have to make all ten equations work, and that pins them down in almost every SAT question. You want more inputs than the highest power of , and ten covers any SAT identity you're likely to meet.
Without the list, Desmos has nothing telling it that x_1 is a set of inputs. It treats x_1 as one more unknown, finds values that make the two sides match at a single point, and reports constants that don't work for every .
Delete the first line, x_1=[1...10]. Before you look, predict: will and stay at and ? Then compare, and reset the calculator.
Leaving the list out, or setting x_1=0, lets the constants match at one point only. Without the list, Desmos reports and ; with x_1=0, it reports and . Neither pair works for every . Give it a list.
As in the last lesson, check that no earlier line gives or a value. If one does, Desmos treats that letter as fixed and won't fit it.
A regression always reports numbers, even when no constants can make the two sides match. It gives the closest fit it can find. So before you answer, check that the fit is exact.
Type the left side minus the right side on a new line, still with x_1. Desmos shows the difference at every input. If every entry is , the two sides really are equal for all ten inputs, and the constants are right. A tiny entry like counts as too: it's the calculator rounding in its last digits, not a real gap. For the opening question, the line
(2x_1+p)(x_1-3)-(x_1+q)(x_1+1)-(x_1^2-10x_1-11)
shows ten zeros.
Here's a case where the check catches a problem. In the equation , is a constant. Is there a value of that gives infinitely many solutions?
The fit 2(kx_1+3)~4x_1+7 reports . That alone looks like an answer. But the check line shows entries like and , not zeros. No value of works, and you can see why: matching the terms needs , so , but then the numbers without are and , and they never match. Some SAT questions offer there is no such value as a choice for exactly this reason.
The check also settles decimals. A value like is Desmos rounding, so type in place of the letter in the check line. If it still shows zeros, is exact.
One limit: this "no such value" reading is safe when the constants are only multiplied by numbers, as is here. When a constant is squared or multiplied by another one, a nonzero check can also mean Desmos settled on the wrong fit, so match one coefficient by hand before you answer no such value.
You fit an identity and Desmos reports . The check line shows . What can you conclude? What if it showed ?
Every input in your list has to be one the expressions accept. A fraction has no value where its denominator is . So if a side contains , an input of breaks it.
When that happens, Desmos shows an error instead of fitted values. The fix is to start the list past the problem input. For denominators and , the list x_1=[2...11] avoids both and .
Two more cases to watch:
x_1=[1...10] works; for , start the list at .x_1=[6...15].Worked example
For all and , the expression is equivalent to , where and are constants. What is the value of ?
Step 1
Is equivalent to with for all makes this an identity, and and are the constants to find. The opening phrase also tells you which inputs are off limits: and , where a denominator would be .
Step 2
The usual list, x_1=[1...10], includes , and Desmos would show an error. Start at instead:
x_1=[2...11]
None of these inputs is or , and ten inputs are plenty for two constants.
Step 3
Type each fraction with parentheses around every top and bottom:
(7x_1+5)/((x_1-1)(x_1+3))~a/(x_1-1)+b/(x_1+3)
Desmos reports and .
Step 4
You can confirm the fit by hand at one input. At , the left side is , and the right side is . They match.
The question asks for , not or :
The answer is A.
Picking , which is . The question asks for the difference, , so watch the sign.
Squares hide a sign. Suppose is equivalent to for all , where and are constants and . Expanding gives , so , and can be or . Both make an identity: with , and with .
Desmos reports only one of them, and here it reports and . That breaks the condition . Add the condition to the end of the regression line in curly braces:
(ax_1+3)^2~16x_1^2+bx_1+9{b<0}
Now Desmos reports and . So when a question adds a condition like , or , check the fit against it, and put it in braces if Desmos lands on the other fit.
Sometimes two constants only ever appear multiplied together. Say and , and for all . Then , so the identity says and nothing more. The pairs and both work.
The fit a(bx_1)+3~12x_1+3 reports and , one pair out of many. If the question asks for , type a*b and you get , which is right for every pair. Don't answer with or alone. If a question does ask for alone, it has to give you another fact about or . Put that fact in too: for , type b=2 on a line above, and Desmos fits only , here .
In , for all , and are constants. Desmos reports and . Can you answer What is ? What about What is ?
The list fit is quick, but it still takes two lines and a check. Sometimes one look at the expressions gives the constant faster.
Take , which is equivalent to for some constant . The constant sits in front of , and only two products make an term: and . So . That's one step.
A good habit before you open Desmos: find the term that holds the constant you want. If one or two products build it, match it by hand. If the constant hides in several products, fractions or squares, or there are two or more constants tangled together, let the list fit do the algebra.
The same goes for infinitely many solutions. In , the terms already match, so the numbers must match too: , and .
The list and the regression line are already in the calculator, and Desmos has found and . Use them to answer the question.
Finish the solution
The expression
is equivalent to for all values of , where and are constants. What is the value of ?
x_1=[1...10].(x_1+p)^2-(x_1-4)^2~12x_1+q.For each one, decide first: is it an identity, which letters are the constants to find, and does one coefficient settle it by hand? Not every problem needs Desmos.
Practice problem
In the equation , and are constants. If the equation has infinitely many solutions, what is the value of ?
Practice problem
For all , the expression is equivalent to , where is a constant. What is the value of ?
Practice problem
The expression is equivalent to for all values of , where and are constants and . What is the value of ?
Practice problem
In the equation , is a constant. For which value of does the equation have infinitely many solutions?
Practice problem
The expression is equivalent to for some constant . What is the value of ?
Practice problem
The expression is equivalent to for all values of , where and are constants. What is the value of ?
Finish the lesson
Finish the remaining questions correctly to complete this lesson.
x_1=[1...10], then both sides with x_1 and a tilde between them. Skip any input that makes a denominator .Next lesson
Graph how a point moves as its parameter changes and test which answer choice satisfies a linear system.
Start next lessonPractice
341 SAT questions use what this lesson teaches. Practice a few in a study session at the difficulty you choose.
Start practice