Penny Robinson; Week 5 MED1022; Epi
A parameter is a fixed numerical value that summarises a characteristic for a population. A sample statistic is characteristic of a smaller sample group. Greek letters denote a population; Latin a sample. Samples are used because measuring populations are not feasible. These samples are used to make inferences about populations.
The smaller the sample is, the greater the uncertainty. Standard error is the average amount of error in a sample statistic, depending on sample size and variability of measurements from study subjects. Standard error = sd/ square root of number of participants. Limits of CI are determined by confidence level percentage times standard error
T table is values of degrees of freedom: if CI is 95% read off 0.05 column, if 90% read off 0.1. Degrees of freedom is one less than sample size. This can be used to find the CI for difference between two means. If CI includes 0, there is no difference between the groups.