Statistics Flashcards
What is sensitivity?
Proportion of patients who have condition who have positive test result
How is sensitivity calculated?
True positive / true posoitive + false negative (all patients with condition)
What is specificity?
Proportion of patients without condition who have negative test result
How is specificity calculated?
True negative / true negayive + false positive (all patients without condition)
What is positive predictive value?
Chance patient has condition if diagnostic test is positive
How is positive predictive value calculated?
True positive / true positive + false positive
What is negative predictive value?
Chance patient does not have condition if diagnostic test is negative
How is negative predictive value calculated?
True negative / true negative + false negative
What is likelihood ratio for positive result?
How much odds of disease increase with positive test result
How is the likelihood ratio for a positive test result calculated?
Sensitivity / (1-specificity)
What is the likelihood ratio for a negative test result?
How much odds for a disease decrease with a negative test result
How is likelihood ratio for a negative test result calculated?
(1-sensitivity) / specificity
What is a null hypothesis?
States no difference between 2 groups
What is an alternate hypothesis?
States there is a difference between groups
What is a type 1 error when testing null hypothesis?
Null hypothesis is rejected when null hypothesis is actually true
(saying there’s a difference when there’s not)
What is the probability of making a type 1 error?
Alpha
What is a type 2 error when testing a null hypothesis?
Null hypothesis accepted when it is false
What is the probability of making a type 2 error?
Beta
What is the p value?
Assuming null hypothesis is true, p value is probability of obtaining a result by chance at least as extreme as one observed
What is the power of a study?
Probability of correctly rejecting null hypothesis when it is false
How is the power of a study calculated?
1 - beta (probability of making type 2 error )
What are the 2 types of significance tests?
Parametric tests - for quantitive data in normal disturbution (mean=median=mode)
Non parametric - if qualitative data or shewed distrubution
What are the types of parametric significance tests?
Student’s T test
Pearson product-moment coefficient - shows correlation
What are the types of non parametric tests?
Mann-Whitney U test
Wilcoxon signed rank test
chi-squared test
Spearman or Kendall rank
What does Mann Whitney U test do?
Compares ordinal, interval or ratio scales of unpaired data
What does Wilcoxon signed rank test do?
Compares 2 sets of observations on single sample before and after intervention
What does chi-squared test do?
Compare proportions or percentages
What does Spearman or Kendall rank show?
Correlation
What is reliability?
Consistency of measure
What is validity?
Accuracy of measure
What is relative risk?
Ratio of Experimental event rate to control event rate
How is relative risk calculated?
EER / CER
How is is relative risk reduction or increase calculated?
Dividing absolute risk change by control event rate
(EER-CER) / CER
What is a confidence interval?
Range of values within which the true effect of intervention is likely to lie
What is the standard error of the mean?
Measure of th spread for the mean of the observations ie how accurate calculated sample mean is from true population mean
How is standard error of the mean calculated?
SEM = standard deviation / square root of sample size
How is a 95% confidence interval calculated?
Lower limit = mean - (1.96 x standard error of mean)
Upper limit = mean + (1.96 x standard error of the mean)
