Inference From Sample Statistics and Margin of Error Pattern - Estimation
Digital SAT® Math — Inference From Sample Statistics and Margin of Error
Inference: Estimation
This pattern tests one core skill: using a sample proportion to estimate a count in a larger population. A random sample reveals that some percentage (or fraction) of items have a certain property. You then apply that proportion to the total population to estimate how many items in the whole group share that property.
The Formula
$\text{Estimated count} = \text{population size} \times \text{sample proportion}$
The sample proportion can be given as a percentage (convert to a decimal first) or as a fraction.
Two Formats You Will See
-
Percentage given directly. "A random sample found that 4% of 1,500 smartphones had a defect." → $1500 \times 0.04 = 60$
-
Fraction computed from sample counts. "Out of 50 fish caught, 45 were trout. Estimate the trout in a lake of 1,200 fish." → Proportion $= \dfrac{45}{50} = 0.9$, then $1200 \times 0.9 = 1080$
Step-by-Step Method
- Identify the sample proportion. If given a percentage, convert to a decimal by dividing by 100. If given sample counts, divide the count of interest by the sample size.
- Multiply the proportion by the total population.
- Pick the answer choice that matches.
Common Gotchas
- Confusing the percentage with the count. If 4% of items are defective, the answer is not 4 — it is 4% of the total. The raw percentage number almost always appears as a wrong answer.
- Computing the complement. If 15% are infected, then 85% are not infected. The SAT puts $\text{population} \times 0.85$ as a trap answer. Make sure you are estimating the group the question asks about.
- Subtracting instead of multiplying. Some wrong answers are $\text{population} - \text{sample size}$ or $\text{population} - \text{count in sample}$. These operations make no statistical sense — always multiply.
Worked Example 1
A factory produced 1,500 smartphones. A random sample found that 4% had a screen defect. Estimate the total number of defective phones.
A) 4 $\quad$ B) 60 $\quad$ C) 1,440 $\quad$ D) 1,500
SOLUTION
Sample proportion $= 4\% = 0.04$
Estimated count $= 1500 \times 0.04 = 60$
Answer: B) 60A) 4 is the percentage, not the count.
C) 1,440 is the number of phones without defects ($1500 \times 0.96$).
D) 1,500 is the total population.
Worked Example 2
A biologist catches 50 fish from a lake of 1,200 fish. Of the sample, 45 are trout. Estimate the total trout in the lake.
A) 5 $\quad$ B) 1,080 $\quad$ C) 1,150 $\quad$ D) 1,155
SOLUTION
Sample proportion $= \dfrac{45}{50} = 0.9$
Estimated count $= 1200 \times 0.9 = 1080$
Answer: B) 1,080A) 5 is the number of non-trout in the sample ($50 - 45$).
C) 1,150 is $1200 - 50$ (subtracting the sample size — meaningless).
D) 1,155 is $1200 - 45$ (subtracting the trout count — also meaningless).
What to Do on Test Day
- Formula: $\text{Estimated count} = \text{population} \times \text{sample proportion}$
- Convert percentages to decimals before multiplying.
- If given raw counts in a sample, compute $\dfrac{\text{count of interest}}{\text{sample size}}$ first.
- The percentage itself, the complement count, and any subtraction-based number are always traps.
- These are fast questions — 20 to 30 seconds if you stay focused on the formula.
More Inference From Sample Statistics and Margin of Error Patterns