P-Hacking - Crash Course Statistics

The video discusses the concept of p-hacking, where data is manipulated to achieve significant p-values, and its impact on research credibility. It explains p-values, null hypothesis testing, and the consequences of p-hacking, using real-world examples. The video also covers multiple testing, family-wise error rates, and methods like the Bonferroni correction to ensure valid research results.

What is p-hacking?

• A method to increase sample size

• A way to improve data accuracy

• Manipulating data to achieve significant p-values

• A technique to reduce error rates

In null hypothesis significance testing, what does it mean to 'fail to reject the null'?

• The alternative hypothesis is proven false

• The null hypothesis is proven true

• There is no evidence to support the alternative hypothesis

• There is evidence supporting the null hypothesis

Why might researchers feel pressured to find significant results?

• To avoid statistical errors

• To ensure their research is published

• To reduce the cost of research

• To increase the sample size

What can be a consequence of p-hacking?

• Reduced sample size

• Improved data accuracy

• Publication of false research results

• Increased research funding

What is the family-wise error rate?

• The rate of data collection errors

• The probability of making at least one Type I error in multiple tests

• The likelihood of a Type II error

• The chance of correctly rejecting the null hypothesis

In the jellybean experiment, what was the main issue with conducting multiple tests?

• It improved the accuracy of the data

• It increased the sample size

• It led to a higher chance of false positives

• It reduced the significance of results

What is a Bonferroni correction used for?

• To improve the accuracy of statistical models

• To reduce data collection errors

• To adjust p-values for multiple comparisons

• To increase the sample size

Why is it important to correct for inflated error rates?

• To increase the number of significant results

• To improve data collection methods

• To ensure research results are accurate

• To reduce the cost of research

How can researchers avoid p-hacking?

• By using more complex statistical models

• By reducing the number of tests conducted

• By choosing analysis methods before seeing the data

• By increasing the sample size

What impact can false research results have on society?

• They can lead to incorrect public policies

• They improve scientific understanding

• They reduce research funding

• They increase the accuracy of future studies