The Logic of Sampling
Methods of Sampling Nonprobability samplesNonprobability samples –Used often in Qualitative Research Probability or random samplesProbability or random samples –Every person has an equal chance of being included in the sample
Sampling of Participants Try to obtain a representative sample – –Representative samples allow us to generalize findings to the larger group Sampling is often not under the control of the researcher in low-constraint (field) research – –Therefore, caution is required in interpreting the results – –Generalize only to similar participants and NOT to the general population
Sampling Terminology Populations Sampling Element Target Population Sampling Frame Parameters and Statistics
Non-Probability Sampling Convenience or Accidental or Haphazard Quota Purposive or Judgmental Snowball
Non-Probability Sampling Deviant cases Sequential Theoretical Use of Informants
Theory & Logic of Probability Sampling Sampling Distribution Central Limit Theorem Sampling Error
The Normal Distribution Represents the actual distribution of naturally occurring data Real distributions do not conform completely to the normal distribution Inferential statistics takes a set of data and “normalizes” it so comparisons can be made
Characteristics of the Normal Distribution Bell shape Unimodal Mean is located at the center of the bell curve Area under the curve is 100% of the data The 50th percentile or the median, is the same value as the mean
The Standard Deviation and the Normal Distribution Direct relationship between the standard deviation and the curve The same number of observations will always fall within the same standard deviation units from the mean of the distribution – –68% lie within -1 to +1 s.d.’s from the mean – –95% lie within -2 to +2 s.d.’s from the mean – –99.8% lie within -3 to +3 s.d.’s from the mean
Probability Sampling Simple Random Sample Systematic Sampling Stratified Sampling
Probability Sampling Cluster Sampling – –Within Household Sampling – –Probability Proportionate to Size (PPS) Random-Digit Dialing
Hidden Populations Targeted Sampling Respondent Drive Sampling
Sample Size Degree of precision or accuracy needed – –Larger samples will provide more precise estimates of population parameters Variability or diversity in the population Number of different variables Costs and time constraints The larger the sample, the more narrow the confidence intervals
Drawing Inferences Inferential Statistics Sampling Error