Define p-hacking and its impact on research integrity.

Define p-hacking and its impact on research integrity.

• A. P-hacking is increasing sample size; improves accuracy
• B. P-hacking is selective reporting or stopping data collection to achieve statistical significance; biases conclusions
• C. P-hacking is random sampling; has no impact
• D. P-hacking refers to data encryption; protects privacy

Answer: (B)

P-hacking refers to manipulating data analysis or reporting to push a result below the threshold of statistical significance, rather than reflecting a true effect. This includes stopping data collection early or continuing until a significant result is found, and selectively reporting only analyses or outcomes that cross the significance line. These practices inflate the false-positive rate and bias conclusions, which undermines trust, hurts reproducibility, and compromises research integrity. The described option best captures both the actions (selective reporting or stopping data collection) and the harm (biased conclusions) that come from p-hacking. Other choices miss the core idea: increasing sample size can be a legitimate design choice, random sampling is a standard method, and data encryption relates to privacy, not statistical integrity.

P-hacking refers to manipulating data analysis or reporting to push a result below the threshold of statistical significance, rather than reflecting a true effect. This includes stopping data collection early or continuing until a significant result is found, and selectively reporting only analyses or outcomes that cross the significance line. These practices inflate the false-positive rate and bias conclusions, which undermines trust, hurts reproducibility, and compromises research integrity. The described option best captures both the actions (selective reporting or stopping data collection) and the harm (biased conclusions) that come from p-hacking. Other choices miss the core idea: increasing sample size can be a legitimate design choice, random sampling is a standard method, and data encryption relates to privacy, not statistical integrity.