Statistics and errors Flashcards
What is the definition of an experiment?
An experiment requires a variable to be changed. E.g. control arm, treatment arm.
What is difficult to determine from an observational study?
Causation, as the investigator has not changed any variable to establish causation.
What is per protocol design?
Dump patients out of the data if they fail to complete the study.
What is intention to treat study design?
Keeping participants who drop out of a study in the analysis at the end.
What is the difference between descriptive and inferential statistics?
Descriptive statistics tell you about the characteristics of the sample used (also called a sample statisitc or SS) and inferential statistics use the inforamation in the study to make an inference about the population (also known as a population prediction or PP). SS are reported in arabic text, PP in greek letters.
What is a type II error?
Incorrectly retaining the null hypothesis when the alternative hypothesis is in fact true.
What is a type I error in a study/test/stats?
Incorrectly rejecting the null hypothesis. If P<0.05, then type I errors will be made 5% the time.
Define a confidence interval
When the procedure is repeated, the true value will fall between the confidence interval, the XX% amount of the time.
What is relative risk?
The ratio of risk of something happening in one scenario (thing happening/ thing happening + thing not happening) to that thing happneing in another scenario (thing happening/ thing happening + thing not happening).
