Which statement best describes p-hacking?

Which statement best describes p-hacking?

• A. A procedure for correcting p-values for multiple testing.
• B. A framework for preregistering hypotheses to avoid bias.
• C. Research practices that lead to selective reporting of significant p-values.
• D. A method for computing exact p-values.

Answer: (C)

P-hacking refers to manipulating data analysis to produce significant p-values rather than reflecting a true effect. It happens when researchers have flexibility in how they analyze data—trying different transformations, adding or removing covariates, making decisions about which data to include, or stopping data collection early—so that only analyses yielding p-values below the conventional threshold are reported. Because p-values are shaped by the analytic path, these practices increase the chance of finding a false positive and undermine replicability. This is why the description focusing on selective reporting of significant p-values captures the essence of p-hacking. The other statements describe legitimate procedures: correcting p-values for multiple tests, preregistering hypotheses to reduce bias, and computing exact p-values.

P-hacking refers to manipulating data analysis to produce significant p-values rather than reflecting a true effect. It happens when researchers have flexibility in how they analyze data—trying different transformations, adding or removing covariates, making decisions about which data to include, or stopping data collection early—so that only analyses yielding p-values below the conventional threshold are reported. Because p-values are shaped by the analytic path, these practices increase the chance of finding a false positive and undermine replicability. This is why the description focusing on selective reporting of significant p-values captures the essence of p-hacking. The other statements describe legitimate procedures: correcting p-values for multiple tests, preregistering hypotheses to reduce bias, and computing exact p-values.