Sampling Methods
Definitions of Important Terms Population The total group of people from whom information is needed. Census Data obtained from or about every member of the population of interest. Sample A subset of the population of interest. Sampling Units (elements) – Person/object of interest for study Sampling Frame – All the elements that are available for selection
Parameter vs Statistic Parameter: The “true” population data. Statistic: The sample data.
Sampling Designs Nonprobability Probability Convenience Judgment Quota snowball Probability Simple Random Systematic Random Stratified Cluster
Probability Sampling Each unit in the sampling frame has an equal, non-zero chance of being selected for the study. Implies random selection. You can then say that the sample of representative of the target population (of interest)
Nonprobability Sampling Nonprobability sampling the probability of selection of each sampling unit is unknown. Data results can’t be used to make predictions about the defined target population; it is limited to just the people who provided the raw data in the survey.
Types of Probability Sampling Designs Simple Random Sampling Types of Probability Sampling Designs Systematic Random Sampling Every sampling unit making up the defined target population has a known equal, non-zero, chance of being selected into the sample. Stratified Random Sampling Similar to simple random sampling, but requires that the defined target population be naturally ordered in some way, i.e customer list. Requires the separation of the defined target population into different subgroups (strata), and the selecting of samples from each stratum.
Types of Nonprobability Sampling Designs Convenience Sampling (Accidental Samples) Types of Nonprobability Sampling Designs Judgment Sampling Drawn based on the convenience of the researcher or interviewer of when and where the study is being conducted. Quota Sampling Participants are selected based on an experienced individual’s belief that the prospective respondent will meet the requirements of the study. Selection of prospective participants based on pre-specified quota requirements. Ie., males and females
Total Error the difference between the true value in the population of interest (the parameter) and the observed value in the sample (the statistic). sources of error: Sampling error, sampling bias Non-sampling error
Total Error – when a sample is not representative Poor Logic Improper Use of Statistics Inadequate sample size design Poor data collection Improper design Poor problem formulation Poorly written report
Controlling Sampling Error Error due to sampling problems or sampling size. controlled by increasing the sample size. However, a larger sample can result in less quality control…in other words, more non-sampling error
Controlling Sampling Bias - systematic error Sample differs from the population in a systematic way not controlled by sample size but by better sampling controls (ie., defining population better).
Controlling Non-sampling Error All other error sources such as measurement error controlled by using good measurement principles and good data entry/analysis a census and a sample both have non-sampling error