Stratified Sampling in AP Stats- Rows of Trees Method
What Stratified Sampling Actually Is
In AP Statistics, stratified sampling means dividing your population into groups called strata, then randomly selecting samples from each group. The goal is to make sure every subgroup is represented in your sample.
Why bother? Because a simple random sample might miss an important subgroup entirely. If you're sampling students and accidentally grab mostly seniors, your results are trash.
Stratified sampling fixes this by guaranteeing representation from each stratum.
The Rows of Trees Method Explained
The Rows of Trees Method is a specific technique for applying stratified sampling to rectangular plots of land. It's called this because you literally divide the area into rows, treat each row as a stratum, and sample within each row.
This method works best when:
- Your population is arranged in a grid or regular pattern
- You suspect important differences exist between rows (like soil quality, sun exposure, or drainage)
- You need a systematic way to ensure every "section" gets represented
It's not the only way to do stratified sampling, but it's the method the College Board loves to throw at you on the AP exam.
When to Use This Method
You'll use the Rows of Trees Method when the question explicitly describes:
- A rectangular area divided into equal rows
- Some characteristic that varies by row (tree height, soil type, plant density)
- Instructions to sample a certain number of trees from each row
If the problem mentions "a grove of trees arranged in rows" or "a rectangular field divided into 12 rows," start thinking about this method.
Step-by-Step: How To Do It
Step 1: Divide Into Strata
Identify your rows. Each row becomes a stratum. If you have 12 rows, you have 12 strata.
Step 2: Determine Sample Size Per Stratum
Usually, you sample proportionally. If you need 30 trees total and have 12 rows, that's about 2-3 trees per row. The exact number depends on what the problem asks for.
Proportional allocation: Sample size for a stratum = (N_h / N) × n
Where N_h is the population size of that stratum, N is the total population, and n is your total sample size.
Step 3: Randomly Select Within Each Stratum
Use a random number generator or table. Pick your assigned number of trees from each row independently.
Step 4: Combine Your Samples
Put all your selections together. That's your stratified sample.
Example Problem
Problem: An orchard has 10 rows with 50 trees per row (500 trees total). You want to estimate the average number of apples per tree. Using stratified sampling with the Rows of Trees Method, select a sample of 20 trees.
Solution:
Step 1: 10 rows = 10 strata
Step 2: 20 trees ÷ 10 rows = 2 trees per row
Step 3: For each row, randomly pick 2 trees. Label trees 1-50 within each row. Use your calculator's randInt function:
For Row 1: randInt(1,50,2) → maybe you get 12 and 34
For Row 2: randInt(1,50,2) → maybe you get 7 and 41
Repeat for all 10 rows.
Step 4: Your sample is the 20 trees you selected across all rows.
Comparing Sampling Methods
Here's how stratified sampling (Rows of Trees) stacks up against other methods you'll encounter:
| Method | Best When | Downside |
|---|---|---|
| Simple Random Sample | No subgroups matter; population is homogeneous | May miss small but important groups |
| Stratified (Rows of Trees) | Important variation exists between rows/groups | Requires knowing strata boundaries |
| Systematic | Population has a natural order; you need efficiency | Hidden periodic patterns can mess you up |
| Cluster | Can't sample from entire population easily | Higher sampling error than stratified |
Common Mistakes Students Make
Mistake 1: Sampling the same position in every row instead of randomizing within each row.
Don't just pick tree #10 in each row. That's systematic sampling, not stratified. Pick randomly for each row independently.
Mistake 2: Forgetting to calculate proportional allocation.
If row sizes are unequal, you can't just take the same number from each row. Weight your sample by stratum size.
Mistake 3: Confusing cluster sampling with stratified sampling.
Cluster sampling picks entire groups randomly. Stratified sampling picks within every group. They're opposites in practice.
Why This Matters on the Exam
The AP Stats exam loves asking about comparing sampling methods and identifying sources of bias. When you see a grid or row-based setup, the answer usually involves stratified sampling.
If a question asks you to design a sampling plan for an orchard, forest, or field, the Rows of Trees Method is probably what they're looking for.
Know how to:
- Identify when stratified sampling is appropriate
- Calculate how many units to sample per stratum
- Execute random selection within each stratum
- Explain why this method reduces sampling error compared to simple random sampling
The Bottom Line
The Rows of Trees Method is stratified sampling applied to a rectangular layout. Divide into rows, sample proportionally from each, randomize within rows. That's it.
Don't overthink it. Don't add complexity that isn't there. The method is straightforward, and the exam questions reflect that.