Sampling Methods and Bias with Surveys - Crash Course Statistics

The video explores non-experimental methods in statistics, focusing on surveys. It discusses the importance of well-constructed surveys, highlighting potential biases in questions and sampling. Various sampling techniques, such as random sampling and stratified sampling, are explained. The video also covers the significance of a census and concludes with cautions about survey biases and their value in data collection.

Why can't certain experiments be conducted to answer specific questions?

• They are too expensive.

• They require unethical practices.

• They are too time-consuming.

• They are not scientifically valid.

What is a key factor that can make or break a survey?

• The time of day it is conducted.

• The color of the survey form.

• The length of the survey.

• The questions and the target audience.

What is a leading question in a survey?

• A question that is too difficult to understand.

• A question that suggests a desired answer.

• A question that is irrelevant to the survey topic.

• A question that is too long.

What is non-response bias in surveys?

• When those who don't respond differ from those who do.

• When the survey is too long to complete.

• When the survey is conducted at an inconvenient time.

• When respondents answer dishonestly.

How does stratified random sampling ensure representation?

• By randomly selecting individuals from the entire population.

• By selecting only the most common groups.

• By focusing on the largest demographic.

• By dividing the population into groups and sampling each group.

What is cluster sampling?

• Using a random number generator to select individuals.

• Sampling individuals from each demographic group.

• Selecting entire groups or clusters to survey.

• Surveying only the most accessible individuals.

Why might a census be preferred over sampling?

• It requires less manpower.

• It is faster to complete.

• It provides data from the entire population.

• It is cheaper to conduct.

What is a major challenge of conducting a census?

• It is not as accurate as sampling.

• It is hard to reach every individual.

• It is difficult to analyze the data.

• It is not legally required.

What should you check to avoid falling for a fake poll?

• The length of the poll.

• The number of questions.

• The professionalism and source of the poll.

• The time it was conducted.

Why are well-conducted surveys essential?

• They require less analysis.

• They provide information without the need for experiments.

• They are cheaper than experiments.

• They are easier to distribute.