Lecture 2A Flashcards
What are the three domains of public health?
Health improvement
Health protection
Healthcare public health
What is epidemiology?
Branch of medicine that deals with the incidence, distribution and possible control of diseases and other factors relating to health.
What is the difference between probability and statistics?
Statistics is where you have information from a sample and use it to generalise about a population.
Probability is where you have information about a population and use it to predict about a sample
What makes a good sample?
Representative
Unbiased
Precise
Precision and bias?
A sample that is precise and biased is worse that a sample that is not biased but not precise
What are the two types of error in a study that can influence results?
Chance (random error)
Outliers-reduced as sample increases
Bias (systematic error)
Difference between true and expected value, increasing sample size will not reduce it
What is incidence?
Rate of occurance of new cases
What is prevalence?
proportion of cases in population at a given time
What is the use of observed values?
Help with knowledge of true values
Create hypothesis to test
Can calculate confidence intervals
Tendency vs observation
Our observation is our best estimate at the true or underlying tendency
Tendency for coins to flip half heads half tails
May not be observed
What is a hypothesis?
Educated guess based on an observation
What do you do with a hypothesis?
Reject or fail to reject
What is the P-value
Probability of observation as extreme or more extreme than the hypothesis occurs. If P value very small then either very unlikely event occurred or hypothesis is wrong.
Eg hypothesis coin is fair. If 10 heads probability of that less than 0.05 and so can reject hypothesis that coin is fair
P-value problems?
0.05 is an arbitrary number. 0.049 find and 0.051 isn’t?
Also depends on statistical significance. Coin might have only been flipped twice
What is a 95% confidence interval?
95% sure that true value lies within interval. Based on our observed value. 95% certain that mean of population lies within the interval. Larger confidence interval means smaller sample use or greater variation in population values.
