Confidence and significance Flashcards
How is confidence measured? (2)
What does it measure?
Measure variability
Within sample- standard deviation
Between samples- standard error
How measure confidence within sample?
How measure confidence between sample?
Within- SD
Between- standard error
What percentage data within one SD limit?
68%
What percentage data within two SD limit?
95%
Is high or low standard error better?
Low standard error- low variability
What does standard deviation do?
What does standard error do?
SD = describing the variability
SE = estimating the variability
Is SE normally high or low in large sample?
Low
What does 95% of data falls within 2 SDs from the mean mean?
95% change the means fall within the limits
If confidence interval includes 1:
NOT significant
I.e. 0.85-1.02 = NOT significant
I.e. 0.6-0.95 = significant
Define null hypothesis?
Statement tested
Statement of no association
Aim prove hypothesis wrong
State 3 common significance tests?
T test
Chi squared
Spearmann rank
Linear regression
If small p value: accept or reject null hypothesis?
If large p value: accept or reject null hypothesis?
what p value needed to be statistically significant?
Small p-value 🡪 reject the null hypothesis; significant data
Large p-value 🡪 accept the null hypothesis; not significant data
p<0.05 indicates significance
Why does statistical significance not mean clinical significance?
Doesn’t account confounding factors
Define confounding factor? Why issue?
Factor other than one being studied that associated both with disease (dependent variable) and with factor being studied (independent variable)
A confounding variable may distort or mask the effects of another variable on the disease in question
Define independent variable?
Define dependent variable?
Independent variable- stands alone, isn’t changed by other variables measuring eg age
Dependent variable- depend on factors- change
