At Scientifically Sound, we have reviewed ongoing discussions on the benefits of confidence intervals (CIs) over p values for statistical analysis and reproducibility in research. In a short editorial, the statistician Doug Altman summarised why we need confidence intervals and showed how confidence intervals force investigators to consider sizes of effects. Here are the key points: Two different but complementary
Read more
The growing awareness of the need for reproducibility in research is encouraging, but what does reproducible research actually look like in practice? Marty and I recently had an interesting discussion on what it means for a study’s findings to be independently replicated, and the metrics scientists use to interpret reproducibility. I tend to interpret level of reproducibility using confidence intervals
Read more
Allocation concealment and blinding are characteristics that prevent bias in randomised controlled trials and experimental studies. However, these concepts are often confused. Using a randomised controlled trial as an example, the statistician Philip Sedgwick explains the differences between allocation concealment and blinding, and why these characteristics are important: Researchers investigated whether a nutritious meal and food packages was more effective
Read more
Sometimes when we plot data we want to indicate a relevant feature or have a user select a section of the data. For example, if we record electromyography (EMG) and force signals, we might want to have the user mark the baseline EMG level so that it can be removed, and determine the amount of EMG produced during a maximum
Read more
In randomised trials or repeated-measures experimental studies of randomised conditions, researchers often measure a continuous variable at baseline and at the end of the study at follow up. Examples of some outcomes include blood pressure, pain, physiological responses, range of motion, etc. In a BMJ statistics note, methodologists Andrew Vickers and Doug Altman explain how these outcomes can be analysed.
Read more
Confidence intervals are widely reported in published research and are usually thought to provide more information than p values from significance tests because confidence intervals indicate how precise an estimate is. Sometimes, however, investigators report an estimate (eg. a mean) and p value, but not the confidence interval about the estimate. In a BMJ statistics note, statisticians Doug Altman and
Read more
Scientifically Sound 