You're staring at a table. Consider this: two columns. In real terms, x on the left, h(x)* on the right. Still, the problem says h is increasing. That's it. Even so, no formula. No graph. Just a handful of rows — maybe five, maybe six — and a question that asks you to estimate something, or compare something, or explain why something must be true.
This is where a lot of students freeze. Not because the math is hard. Because the absence* of a formula feels like a trap Not complicated — just consistent..
It's not. Selected values of an increasing function h is one of those topics that sounds abstract until you realize it's just logic wearing a math costume. Once you see the pattern, you stop guessing and start reasoning.
What Is an Increasing Function, Really?
Let's clear the air first. A function h is increasing on an interval if, whenever you pick two inputs a and b in that interval with a < b*, the outputs satisfy h(a) < h(b)*. Always. In practice, no flat spots. On the flip side, no exceptions. Bigger input, bigger output. No dips.
Strictly increasing. In practice, that's the default assumption unless the problem says "non-decreasing" (which allows flat sections). In most high school and early college contexts, "increasing" means strictly increasing Nothing fancy..
Now, "selected values" just means you're given a subset of the function's domain and corresponding outputs. A table. In practice, maybe a graph with a few labeled points. A few coordinate pairs. You don't have the whole function. You have clues.
Your job: use the fact that h is increasing to deduce things about values you don't* have Small thing, real impact..
The Hidden Power of Order
Here's the key insight: an increasing function preserves order. On top of that, if x₁ < x₂ < x₃*, then h(x₁) < h(x₂) < h(x₃)*. The x-values and h(x)-values march in lockstep.
This means the table isn't just a list. It's a chain of inequalities. Every row gives you a boundary for the rows around it The details matter here..
Why This Shows Up Everywhere
You'll see selected-values problems in:
- AP Calculus (estimating derivatives, integrals, inverse function values)
- SAT/ACT (function behavior questions with tables)
- College algebra/precalculus (inverse functions, average rate of change)
- Data science interviews (monotonicity checks, interpolation logic)
The common thread: you're being tested on whether you understand what increasing* actually implies — not whether you can plug numbers into a formula And that's really what it comes down to..
A Quick Example to Ground Things
Suppose you're told h is increasing and given:
| x | h(x)* |
|---|---|
| 1 | 3 |
| 4 | 7 |
| 6 | 12 |
What can you say about h(5)*?
You know h(4) = 7* and h(6) = 12*. Since 4 < 5 < 6 and h is increasing, h(5)* must be between 7 and 12. That's it. 7 < h(5) < 12.
Could it be 7.1? Yes. 11.Plus, 9? Yes. So 7? So no — strictly increasing means h(5) > h(4). 12? No — h(5) < h(6) Simple, but easy to overlook..
That's the whole game. Bounding unknown values using known ones It's one of those things that adds up..
How to Work With Selected Values
Most problems fall into a few categories. Let's walk through them Surprisingly effective..
1. Estimating Function Values (Interpolation)
You're given a table. You're asked: "Estimate h(2.That's why 5)" or "Between which two integers does h(3. 7) lie?
The method: Find the two x-values in the table that bracket your target. The corresponding h(x)* values are your bounds.
If the table has x = 2, h = 5* and x = 3, h = 9*, then h(2.So naturally, h(3. 5)* is between 5 and 9. 7)* is between 9 and whatever h(4)* is (or unbounded above if 3 is the largest x given).
Pro tip: If the problem asks for a single number* estimate (not a range), the most common approach is linear interpolation — assume the function grows at a constant rate between the two known points. But only do this if the problem asks for an estimate or approximation*. If it asks "what must be true?" or "what is the range of possible values?", stick to inequalities.
2. Estimating Derivatives (Average Rate of Change)
Calculus favorite. "Use the data to estimate h'(3)*."
You don't have the derivative. Also, you have a table. The best you can do is average rate of change over a small interval containing 3.
If the table gives h(2.9)* and h(3.1)*, use those:
h'(3) ≈ [h(3.1) - h(2.9)] / [3.1 - 2.9]*
If you only have h(2)* and h(4)*, your estimate is cruder:
h'(3) ≈ [h(4) - h(2)] / [4 - 2]*
Crucial: Because h is increasing, h' > 0* (where it exists). Your estimate should be positive. If you get a negative number, you subtracted in the wrong order Surprisingly effective..
3. Inverse Function Values
If h is increasing, it's one-to-one. It has an inverse h⁻¹.
Problem: "Find h⁻¹(10)* given the table."
Look at the h(x)* column. Find 10 (or the two values that bracket 10). The corresponding x-value(s) give you h⁻¹(10)* It's one of those things that adds up..
If h(4) = 8* and h(5) = 12*, then h⁻¹(10)* is between 4 and 5.
This works because* increasing functions have increasing inverses. The order flips: y₁ < y₂* implies h⁻¹(y₁) < h⁻¹(y₂)*. Same logic, swapped axes Not complicated — just consistent..
4. Definite Integrals (Riemann Sums)
"Approximate ∫₁⁵ h(x) dx using a left Riemann sum with the given subintervals."
You're not integrating a formula. You're summing areas of rectangles whose heights come from the table Surprisingly effective..
Left sum: use the left endpoint* of each subinterval as the height. Right sum: use the right endpoint*. Because of that, midpoint: use the midpoint (if you have that data). Trapezoidal: average the left and right heights for each subinterval.
5. Accumulation Problems
These combine multiple concepts. On the flip side, initially there are 100 gallons. "Water flows into a tank at a rate given by the table. How much water is in the tank at time t = 6?
The method:
- Net change = integral of rate = area under the rate curve
- Final amount = Initial amount + accumulated change
- Use Riemann sums to estimate the integral from the table data
- Add to the initial condition
If the rate is always positive, the amount increases. If negative, it decreases. Watch the signs carefully.
6. Optimization from Discrete Data
"Given the table of values for h(x), what is the maximum value of h on [1,5]?"
The method: Scan the h(x) column for the largest value. The maximum occurs at the corresponding x-value.
For more complex problems like "find the maximum rate of change," compute average rates of change between consecutive points and identify the largest And that's really what it comes down to..
7. Behavior Analysis
Questions about where h is increasing, concave up, etc Simple, but easy to overlook..
Increasing: Look for h(x) values getting larger as x increases Decreasing: Look for h(x) values getting smaller as x increases Concavity: Examine the differences in h(x) values - if they're increasing, h is concave up
Key Strategies for All Table Problems
- Identify what you're looking for - function value, derivative, integral, or inverse
- Find the relevant rows in the table that contain or bracket your target
- Choose the right tool - interpolation, average rate of change, Riemann sum, or direct lookup
- Check units and signs - especially for rates and accumulation problems
- Estimate first - get a rough sense of what the answer should be before calculating
Practice Tips
Start by categorizing the problem type. Still, once you know the type, the solution path becomes clear. Most table-based questions fall into the seven categories above. Always read carefully to determine whether you need an exact value, an estimate, or a range of possible values Took long enough..
Not obvious, but once you see it — you'll see it everywhere.
Remember: Tables give you discrete snapshots. Your job is to extract meaningful information about the continuous function that generated those snapshots.
Working with function tables requires you to think like a detective - carefully examining the evidence provided and drawing logical conclusions. So by mastering these seven problem types and their associated techniques, you'll be equipped to handle virtually any table-based question thrown your way. The key is recognizing patterns quickly and applying the appropriate method with precision Simple as that..
