William Briggs explains the basics of calculating probability, which pieces of the calculation are subjective, and how the results can influence beliefs and actions.

Economist Gary Smith digs into a published study on green innovations that failed to disclose missing pieces of its dataset.

Breaking down composite endpoints to their component parts.

In this video Emily explains the difference between a Bayesian approach and a frequentist approach to analyzing statistics. A Bayesian analysis looks at prior probabilities combined with data to determine the probability that the hypothesis is true. A frequentist analysis compares the hypothesis to the null-hypothesis, a yes/no approach, to determine if the data could support the null-hypothesis. It then ranks the data with a P-value, but it actually says nothing about the hypothesis being true.

Harvard medical school 'research-integrity officer' was co-author of studies in question.

Data falsification claims against Harvard researchers have now expanded to Khalid Shah, a neuroscientist at Harvard Medical school. The claims include 44 instances of fraud spanning from 2001 to 2023.

Explaining how a clinical trial can report success – when there is none.

Scientific misconduct is ever becoming more recognized as a problem, yet what exactly constitutes misconduct varies by source.

Four senior researchers at Harvard's Dana-Farber Cancer Institute have been accused of forging data through “duplications of blots, bands, and plots within images.”