Sampling Methods and Bias with Surveys (Statistics)

Crash Course: Sampling Methods and Bias with Surveys (Statistics)

Sampling Methods and Bias with Surveys: The Crash Course

Opening Hook

Imagine you're trying to figure out what your friends think about pineapple pizza. You ask 10 of them, and 9 say it's disgusting. But what if those 10 friends are all from the same dorm, and they're all obsessed with pineapple? That's basically what's happening with surveys, folks. Only 3% of Americans eat pineapple on pizza, but if you only surveyed college students, you'd get a very different answer.

The Core Idea

Sampling methods and bias are crucial when it comes to surveys. A survey is only as good as its sample, and if that sample is biased, the results are basically useless. Think of it like trying to take a selfie with a bunch of people who all look like you – it's not going to give you a very diverse or representative picture.

Key Facts & Figures

• The first survey was conducted by the ancient Greeks around 400 BCE, where they asked citizens about their opinions on various issues.
• The term "statistics" was coined by Sir Francis Galton in 1883, who was a British scientist and statistician.
• The first modern survey was conducted by Gallup in 1935, where they asked Americans about their opinions on politics and social issues.
• Only 1 in 5 Americans trust the government to do what's right, according to a 2020 survey by Gallup.
• The average American spends around 4 hours and 38 minutes per day on their phone, according to a 2020 survey by Deloitte.
• The most common reason people don't participate in surveys is because they don't have time, according to a 2019 survey by the Pew Research Center.
• The most common way people participate in surveys is through online questionnaires, according to a 2020 survey by the American Association for Public Opinion Research.
• The accuracy of surveys can be affected by factors like social desirability bias, where people answer questions in a way that they think is socially acceptable.
• The accuracy of surveys can also be affected by factors like non-response bias, where people who don't respond to surveys are different from those who do.
• The most common type of survey bias is sampling bias, where the sample is not representative of the population.
• The most common way to reduce sampling bias is through stratified sampling, where the population is divided into subgroups and a random sample is taken from each subgroup.
• The most common type of survey question is the multiple-choice question, where respondents are asked to choose from a list of options.
• The most common type of survey analysis is descriptive statistics, where the data is summarized and described.

Thought Bubble

Imagine you're a market researcher trying to figure out what kind of coffee people like. You decide to survey 100 people at a coffee shop, but you only ask people who are wearing a "I'm a coffee snob" t-shirt. You get back 90 responses, and 80 of them say they like artisanal coffee. But wait, what if those 90 people are all just trying to fit in with the coffee snob crowd? That's basically what's happening with sampling bias. You're only getting responses from people who fit a certain profile, and that's not representative of the whole population.

Why This Matters

• Surveys can be used to inform policy decisions, but only if they're accurate and unbiased.
• Surveys can be used to understand public opinion, but only if they're representative of the population.
• Surveys can be used to predict election outcomes, but only if they're accurate and unbiased.
• Surveys can be used to understand consumer behavior, but only if they're representative of the target market.
• Surveys can be used to inform business decisions, but only if they're accurate and unbiased.
• Surveys can be used to understand social trends, but only if they're representative of the population.
• Surveys can be used to predict economic trends, but only if they're accurate and unbiased.

Crash Course Recap

• ⚠️ Sampling bias is a major problem in surveys, where the sample is not representative of the population.
• Stratified sampling is a way to reduce sampling bias, by dividing the population into subgroups and taking a random sample from each subgroup.
• Descriptive statistics are a way to analyze survey data, by summarizing and describing the data.
• Multiple-choice questions are a common type of survey question, where respondents are asked to choose from a list of options.
• Surveys can be used to inform policy decisions, but only if they're accurate and unbiased.
• Surveys can be used to understand public opinion, but only if they're representative of the population.
• Surveys can be used to predict election outcomes, but only if they're accurate and unbiased.
• Surveys can be used to understand consumer behavior, but only if they're representative of the target market.
• Surveys can be used to inform business decisions, but only if they're accurate and unbiased.
• Surveys can be used to understand social trends, but only if they're representative of the population.
• Surveys can be used to predict economic trends, but only if they're accurate and unbiased.

Quiz Yourself

1. What is the most common type of survey bias? a) Sampling bias b) Non-response bias c) Social desirability bias d) All of the above

Answer: a) Sampling bias

1. What is a way to reduce sampling bias? a) Stratified sampling b) Random sampling c) Convenience sampling d) All of the above

Answer: a) Stratified sampling

1. What is the most common type of survey question? a) Multiple-choice question b) Open-ended question c) Likert scale question d) All of the above

Answer: a) Multiple-choice question

1. What is the most common type of survey analysis? a) Descriptive statistics b) Inferential statistics c) Predictive analytics d) All of the above

Answer: a) Descriptive statistics

1. What is the most common reason people don't participate in surveys? a) They don't have time b) They're not interested c) They're not representative of the population d) All of the above

Answer: a) They don't have time