0414 BIostatistics 3 Flashcards
Identify the role of biostatistics in epidemiological and medical research Describe the steps in hypothesis testing Identify the role of chance in epidemiological studies and interpret p-values and confidence intervals Classify type 1 and type 2 errors
What is chance and what is it’s importance in a clinical context?
Chance is random error that appears to cause an association between exposure and outcome.
It is important because a principle assumption in research is that we can infer population characteristics from a sample. However, there is always a possibility our results are affected by chance. It is always best to account for chance in research
List the 7 steps of hypothesis testing
- Research question
- Sample and conduct study
- Null & alternative hypothesis
- Identify level of significance / probability
- Calculate test statistic
- Obtain p-value / confidence interval
- Interpret and make conclusions
What are the 3 key components required in every research question
Identify the population (and therefore a sample) of interest. Define outcome and parameters. Define factors and parameters
What is the impact of sample size in an experiment
Large sample = greater representation, reduces random error, increases power, more expensive/ time consuming
Small sample size = less representation, more chance of random error, less power, less expensive/ time consuming
What is null hypothesis and what is the alternative hypothesis
Null hypothesis = postulation that there is no difference between groups wtih in a population (any difference due to chance/sample variation)
Alternative hypothesis = Postulation that there is a difference between groups
What is the alpha value and what is the general set value
Alpha value is the probability that we will obtain an extreme test statistic assuming null hypothesis is true (i.e. a type 1 error). It is generally set to 0.05
What is a type 1 error and what is a type 2 error
Type 1 error = rejecting null hypothesis when it is actually true
Type 2 error = accepting the null hypothesis when it is actually false
What is meant by the ‘power’ of a study/experiment
Power is the probability of finding an association (result) in our sample is there is a true association in the population
List some examples of statistics tests in continous data and categorical data
Continuous data tests - t-tests, ANOVAs, regressions
Categorical data tests - chi-square, logistic and multinominal regressions
What is the function of the p-value in statistics
The p-value indicates whether you should accept or reject your null hypothesis based on your alpha value. If p < alpha you reject the null. If p > alpha you accept the null
What is the principle of the Central limit theorem
Central limit theorem - the sampling distribution of the mean approaches normal as the number of samples (n) increases – regardless of the underlying distribution in the population
What is meant by the 95% confidence interval in a samples data set
95% Confidence Interval = Probability that the sample mean lies within 1.96 SE of population mean
It is a measure of precision and is based on standard error
What is the implication of a wide confidence interval? What about a narrow confidence interval (think in terms of precision and random error)
Wide confidence interval = less precise and more random error
Narrow confidence interval = more precise and less random error
