Learn to collect reliable data! Understand different sampling methods, recognize bias in data collection, and evaluate the quality and reliability of your data sources.
• Every member has equal chance of selection
• Most reliable method
• Reduces bias
• Example: Drawing names from a hat
• Population divided into groups (strata)
• Random sample from each group
• Ensures representation of all groups
• Example: Sampling by grade level
• Select every nth member
• Regular interval pattern
• Easy to implement
• Example: Every 10th person in line
• Select easily accessible members
• Quick and inexpensive
• Often biased
• Example: Surveying friends
• Selection Bias: Sample doesn't represent population
• Response Bias: People give inaccurate answers
• Volunteer Bias: Only certain types of people participate
• Question Bias: Questions lead to specific answers
Scenario: A school wants to know students' opinions about homework. They survey only students in the library during lunch.
Problem: Selection Bias
Better Method:
Randomly select students from all grade levels and classes to get a representative sample.
Scenario: A survey asks: "Don't you think our school needs better sports facilities?" (Yes/No question)
Problem: Question Bias
Better Question:
"How would you rate our school's sports facilities?" (Excellent/Good/Fair/Poor)
• Consistency of results
• Same results when repeated
• Free from random errors
• Example: Scale gives same weight
• Measures what it claims to measure
• Accuracy of results
• Free from systematic errors
• Example: Scale measures actual weight
Scenario: A study claims that 80% of teenagers prefer online learning. The survey was conducted on social media with 100 responses.
Reliability Issues:
Validity Issues:
Conclusion:
The data has both reliability and validity issues. Results should be interpreted with caution.
• Use random sampling
• Ask neutral questions
• Include all response options
• Keep questions simple and clear
• Ensure anonymity
• Leading questions
• Double-barreled questions
• Missing response options
• Complex or confusing language
• Small, biased samples
Original Question: "Don't you agree that our cafeteria food is terrible and needs improvement?"
Problems:
Improved Question:
"How would you rate the quality of our cafeteria food?"
Options: Excellent / Good / Fair / Poor / Very Poor
Scenario: A school board wants to know if students support wearing uniforms. They survey 50 students who are waiting in the principal's office.
Survey Results: 70% support uniforms
Bias Analysis:
Better Approach:
Convenience sampling is quick but often biased. Use random sampling for reliable results.
People who don't respond may have different opinions than those who do respond.
Just because two things are related doesn't mean one causes the other.
Problem 1:
A survey asks: "How much do you love our amazing new cafeteria food?" What type of bias is this?
This is question bias. The word "amazing" leads respondents to give positive responses.
Problem 2:
A school surveys only honor roll students about study habits. What type of bias is this?
This is selection bias. Honor roll students may have different study habits than the general student population.
Problem 3:
Which sampling method is most likely to produce unbiased results?
Random sampling is most likely to produce unbiased results because every member has an equal chance of being selected.