Sampling Methods and Random Digit Tables
Day 2: Sampling
Do Now (May 5th, 2025)
- Problem: Selecting 3 people randomly from a class of 10 to represent the class at a math competition.
Homework Review
- Last Night's HW:
- 1a) Experiment: Treatment is being imposed (mints in soda).
- Treatment: Mints dropped in the soda.
- Response Variable: Height of the soda.
- 1b) Observational Study: No treatment is imposed.
- Population of Interest: Fifth-grade students.
- 2) Conclusion: It is not reasonable to draw a cause-and-effect relationship because it is not an experiment. Only experiments can demonstrate causation.
- 3a) Issue: Rats don't normally use drugs.
- 3b) Method: Randomly assign 10 rats to each group (0 mg, 1 mg, 2 mg). Measure the time taken to complete the maze.
- 4) Identifying studies:
- A) II - Treatment is being imposed.
- B) III - Depends on the population of interest.
- C) A treatment is not being imposed.
Random Digit Tables
- Definition: Lists of digits 0-9.
- Property 1: Each digit has the same likelihood of appearing in each spot.
- Property 2: Each entry is independent of the next.
- How to Use: Assign each member of the population a unique number and select numbers from the random digit table until the desired amount of distinct subjects is obtained.
Example: Mrs. Isaac & Syosset HS Sophomores
- Scenario: Mrs. Isaac wants to gauge the proportion of Syosset HS sophomores (612 total) who think the pool should be replaced with a student arcade.
- a) Assign numbers to the students.
- b) Use the random digit table to select the first 5 students.
Representative Samples
- If a sample is taken to generalize to a population, the sample must be representative (similar to the population, but smaller).
Example: Senior Class President Election
- Scenario: Campaign manager wants to know what proportion of students would vote for their friend running for senior class president (314 students total).
- a) Polling everyone in the friend’s math class is not representative.
- b) Assign every student a number from 1 to 314, then use a random number generator to select 10 students to poll. This is a better approach.
- c) Asking every student going through the lunch line is not ideal as it may not be fully representative.
- Ramon wants to know if students at Syosset High School would be more likely to go to home football games on Friday nights.
- Problem with Method: Standing outside the boys' locker room at the end of the school day is biased because it's not representative of the entire school population.
Example: Nomar & Netflix Accounts
- Nomar wants to know the proportion of adults ages 18-25 that have a Netflix account. He decides to go to the phone book and randomly select 50 landlines to call.
- Problem: This won’t accurately represent the population because many people in that age group may only have cell phones. Therefore, using landlines will exclude a significant portion of the target population.