AP Statistics: Sampling Designs, Random Digit Selection, and Stratification

Identification of Populations and Samples

  • High School Cafeteria Satisfaction Study

    • Context: A study examining whether student duration of attendance at a high school correlates with reduced satisfaction regarding cafeteria food.

    • Data Collection Method: A two-question survey administered to the first 100100 individuals entering the cafeteria on a given day.

    • Question 1: "How long have you attended school here?"

    • Question 2: Numerical rating of cafeteria food on a scale from 11 to 55.

    • Sample: The 100100 individuals who entered the cafeteria first on that specific day and answered the two-question survey.

    • Population: All students currently attending the high school (or all students who use the cafeteria facility).

  • Sugar Maple Tree Diameter Study

    • Context: Estimation of the mean trunk diameter of Sugar Maple trees in a large forest conducted by a maple sugar producer.

    • Data Collection Method: Partitioning the forest into several hundred plot units measuring 10 m×10 m10\,\text{m} \times 10\,\text{m}, selecting a random sample of 2525 plots, and measuring trunk diameters for every Sugar Maple tree residing within those chosen plots.

    • Sample: All individual Sugar Maple trees located within the 2525 randomly selected 10 m×10 m10\,\text{m} \times 10\,\text{m} forest plots.

    • Population: The complete aggregate of all Sugar Maple trees in the entire large forest.

Classification of Sampling Designs

  • Convenience Sampling

    • Definition: A non-probability sampling methodology where individuals are chosen based on ease of access, proximity, or immediate availability.

    • Application: The high school cafeteria study utilizes a convenience sample by selecting the first 100100 people entering the cafeteria door.

  • Cluster Sampling

    • Definition: A sampling procedure wherein the population is divided into smaller geographical or natural subgroups called clusters, a probability sample of clusters is randomly selected, and a complete census is conducted on every unit within each selected cluster.

    • Application: The Sugar Maple tree study employs cluster sampling by treating each 10 m×10 m10\,\text{m} \times 10\,\text{m} plot as a cluster, selecting a Simple Random Sample (SRS) of 2525 clusters, and measuring every tree inside those clusters.

Potential Drawbacks and Biases in Sampling Methods

  • Undercoverage and Selection Bias in Convenience Samples

    • Early-Arrival Bias: Students arriving first to the cafeteria may possess systemic differences compared to the general student population (e.g., specific grade levels with early schedules, higher enthusiasm, or stronger preference for cafeteria food).

    • Exclusion of Subgroups: Students who bring lunch from home, students who eat off-campus, students who arrive late, or students with open campus privileges during lunch periods are excluded from selection.

    • Confounding Variables: Grade level or length of enrollment directly dictates scheduling and arrival times; thus, time of entry acts as a confounding variable that distorts the true relationship between school tenure and cafeteria satisfaction.

Simple Random Sampling Protocol Using Random Digits Tables

  • Study Parameters

    • Target Population: N=975N = 975 students who have completed at least one year of study at a university but have not yet graduated.

    • Target Sample: Simple Random Sample (SRS) of size n=50n = 50 students.

    • Sampling Frame: Alphabetized roster of all 975975 eligible students.

  • Standardized Protocol for Selection

    1. Labeling: Assign a unique three-digit numerical code from 001001 to 975975 to each student on the alphabetized list in sequence.

    2. Unassigned Digits: Reserve codes 000000 and 976976 through 999999 as invalid, unassigned labels.

    3. Reading Digits: Starting at a designated location in the random digits table (e.g., Line 120), read consecutive digits in blocks of three from left to right across each line.

    4. Inclusion / Exclusion Rules:

    • If a three-digit block falls in the range 001001 to 975975, select the student assigned that code.

    • If a three-digit block equals 000000 or falls in the range 976976 to 999999, ignore the block and advance to the next three digits.

    • If a three-digit block repeats a code previously selected, ignore the duplicate and advance to the next three digits.

    1. Stopping Rule: Continue reading three-digit blocks until 5050 unique valid student labels are identified.

Step-by-Step Execution of Random Selection from Table

  • Random Digits Table Reference

  

Random Digits Table Lines 120 to 122
  • Line 120 Digit Stream Analysis

    • Raw Stream: 35476 55970 39421 65850 04266 58435 43742 11937

    • Sequential 3-Digit Parsing:

    • Block 1: 354 →\rightarrow Valid label 354≤975354 \le 975 →\rightarrow Selected Student 1: 354354

    • Block 2: 765 →\rightarrow Valid label 765≤975765 \le 975 →\rightarrow Selected Student 2: 765765

    • Block 3: 597 →\rightarrow Valid label 597≤975597 \le 975 →\rightarrow Selected Student 3: 597597

    • Block 4: 039 →\rightarrow Valid label 039≤975039 \le 975 →\rightarrow Selected Student 4: 039039

    • Block 5: 421 →\rightarrow Valid label 421≤975421 \le 975 →\rightarrow Selected Student 5: 421421

    • Block 6: 658 →\rightarrow Valid label 658≤975658 \le 975 →\rightarrow Selected Student 6: 658658

    • Block 7: 500 →\rightarrow Valid label 500≤975500 \le 975 →\rightarrow Selected Student 7: 500500

    • Block 8: 426 →\rightarrow Valid label 426≤975426 \le 975 →\rightarrow Selected Student 8: 426426

    • Block 9: 658 →\rightarrow Duplicate label 658658 →\rightarrow Skipped

    • Block 10: 435 →\rightarrow Valid label 435≤975435 \le 975 →\rightarrow Selected Student 9: 435435

    • Block 11: 437 →\rightarrow Valid label 437≤975437 \le 975 →\rightarrow Selected Student 10: 437437

  • Selected Sample Summary

    • The first 1010 students selected using Line 120 of the random digits table correspond to student IDs: 354,765,597,039,421,658,500,426,435,437354, 765, 597, 039, 421, 658, 500, 426, 435, 437.

Stratified Random Sampling Implementation

  • Principles of Stratified Random Sampling

    • Definition: A probability sampling methodology where the target population is subdivided into mutually exclusive, homogeneous groups called strata based on a specific characteristic, followed by independent Simple Random Samples taken within each stratum.

    • Primary Advantage: Reduces sampling variability of the estimator compared to Simple Random Sampling, provided the stratifying variable is strongly associated with the response variable.

  • Optimal Stratification Variables for Summer Earnings

    • Option A: Class Year / Grade Level (e.g., Sophomore, Junior, Senior)

    • Justification: Summer job opportunities and wage rates differ systematically by academic class standing. Seniors and juniors typically qualify for higher-paying professional internships or full-time seasonal positions, whereas sophomores may be limited to entry-level or minimum-wage employment. Stratifying by class year ensures proportional representation across academic levels and decreases variability in total estimated earnings.

    • Option B: Major / Academic Field of Study

    • Justification: Summer earnings vary significantly across fields of study (e.g., engineering, computer science, and business majors often receive substantially higher hourly pay or stipends compared to humanities majors). Stratifying by major accounts for wage disparities across academic disciplines.