Sampling Flashcards
Non-response bias
When a significant number of participants do not return their surveys. This could result in bias
Population
Total number of person of object of interest in a study.
Retrospective study
Pick a sample and gather data about their past
Placebo
The control treatment used in medical experiments
Simple Random Sample (SRS)
A sample in which each group of n units has an equal chance of being selected.
Systematic Sampling
When you randomly pick a starting point on a list and select every 5th person thereafter
Undercoverage
When individuals from a subgroup of the population are not represented
Cluster Sampling
When entire groups, which are representative of the population are chosen randomly for a survey
Representative Sampling
A hopefully representative subset of the population
Stratified Sampling
When the population is subdivided and random samples drawn from each group. The groups should each represent a group characteristic which may influence response.
Counfounded
When you can’t determine which variable is effecting the difference in the outcomes of an experiment
Response Bias
Anything in a survey design that influences the response
Statistic
A data value (such as a mean) calculated from the sample.
Randomized Blocking Design
In this experimental design, you first divide the participants into groups based on a variable or characteristic that you believe will affect the response. Then you randomly assign to subjects from each group to treatments.
Randomization
In an experiment, this evens out the effects of variables we cannot control or that we haven’t even thought of.
