Medical Statistics Flashcards
(25 cards)
Define population
The total group of individuals about whom you wan tot make conclusions
What is a statistic
Numerical summary of a sample
What is a parameter
The numerical summary of the population
What is a sample
A subset of the population for whom we actually have data
What type of variable are blood group, ethnicity and country of birth?
Categorical nominal
What type of variable are disease stage, education level and heart murmur grade
Categorical ordinal
What type of variable are number of meds and number of hospital admissions
Numerical discrete
What type of variable are age, weight and blood pressure
Numerical continuous
What is time to event and give some examples
How long it takes for a binary event to happen
Eg time to death, time to recovery
How do you standardise with z scores
Z = [observation - mean] / standard deviation
What does a positive z score mean
The observation is above the mean
What does a negative z score indicate
The observation is below the mean
What does a z score of 2 mean
Data value is two standard deviations above the mean
What is a p value
The probability of obtaining a result as more extreme than that, if the null hypothesis is true and there is no bias
What happens if the results of a hypothesis test are >0.05
They’re not statistically significant
The results are likely due to chance
What happens if the results of a hypothesis test are
</= 0.05
The results are statistically significant
The result is unlikely due to chance
What happens to the p value as the sample side goes up
Goes down
What is a confidence interval
The range of values that is likely to contain the true parameter value
What is the difference between statistically and clinically significant
Statistically - the effect is not explained by chance
Clinically - the effect is clinically important
What are the four typers of bivariate teste
Correlation/regression
Chi squared/McNemar
T test
Anova
What stats tests could you do if you had two numerical variables
Correlation/regression
What stats test could you do if you had two categorical variables
Chi squared
McNemar
What stats tests ould you do of you had one categorical and one numerical variables
T test
ANOVA
What is correlation coefficient
Measures the strength of the linear association between two numerical variables
