PNB 3RM3 Lecture Notes - Lecture 44: Open Science, Statistical Hypothesis Testing
Document Summary
Innovative ideas: contribute new knowledge to the world. Reproducible results: because we are studying phenomena that are presumably real they should be reproducible by other researchers. These capture what we want to convey with science. The ideal form of doing science is when its transparent, open and reproducible. Questionable research practices: people tend to just report the analysis that gives them the result they want vs the result of all the analyses that they did. Overabundance of positive results: biased of seeing only positive results. Lack of replication: not enough people are replicating their results. Limitations of null hypothesis testing: people tend to preclude too much from their results. What it is: all data processing and analytical choices made after seeing and interacting with your data. You want to have as high of a degree of freedom as possible to have as high of a power as possible.