Lecture 23: Chance l Flashcards
What is external validity?
The extent to which the findings of the study can be applied to a broader population
External validity is also known as
generalisability
What is internal validity?
The extent to which the findings of the study are free of chance, bias and confounding
What three things so we need to consider when looking at internal validity?
chance, bias, confounding
What is a population parameter and how does this differ from the estimate?
population parameter is the answer to the question - the true value of the measure in the population that the study is trying to discover but when we take a sample, we find the study’s estimate of the parameter and this is called an estimate.
What is sampling error and how does it come about?
the estimate is just an estimate of the parameter, it differs from the parameter and every sample has a different estimate. This is called sampling error, and is a form of random error. It is commonly just called chance.
What are the advantages of increasing the size of the study sample?
- it reduces the sample variability
- increases likelihood of getting a representative sample
- increases precision of a parameter estimate
What is the statistical definition of a 95% confidence interval?
If you repeated a study 100 times with a different random samples each time, and got 100 estimates and 100 confidence intervals, in 95 of the 100 studies, the parameter would lie within that study’s 95% confidence interval but in 5 studies, the parameter would not
What is our interpretation of confidence interval?
We are 95% confident that the true population value lies between the limits of the confidence interval
What is the template for the interpretation of a confidence interval?
We are [confidence interval level] confident that the parameter lies between [lower bound] and [upper bound]
What 6 things can confidence interval be applied to?
- means
- prevalence
- incidence
- RR
- OR
- RD
Which is more precise, a narrow or wide CI?
narrow
If an RCT shows a 10% reduction in the incidence of heat attacks in people taking the drug compared to people taking the placebo wold be clinically important, what would the RR have to be to be clinically important?
0.90 or less
How does relative risk and CI contribute to the clinical importance? Using 0.90 as the example
If the relative risk is lower that the clinically important value of 0.90 and the upper and lower limits of the CI is lower then 0.90
How does the CI relate to the statistical significance?
If the CI does not contain the null value (1) then it shows statistical significance