October 2 2018 Soci 217 Sampling Introduction · The process fo selecting observations is called sampling · Key to generalizing from a larger population is probability sampling . This involves random sampling Why Sample? . We sample all the time · Need a process to select participants/observations · Studying a whole population may be cumbersome/impossible 2 Major Types of Sampling . Non Probability Sampling · Probability Sampling Probability Sampling . Samples selected in accord with probalility theory Every element of the population has a KNOWN NON-ZERO probability of being selected o To know someone's possibility of being selected, the tester must know the total # of people in the population o Leads to a problem because a lot of the time we do not know the exact population o Typically involving some random selection mechanism . Typically involves some random selection mechanism > representative of larger population o Most often, elements of a population stand equal chance of selection ... · Most common approach in quantitative research Non-Probability Sampling . Any technique for selecting samples that does not use probability theory ... · May be unrepresentative o Elements of a population do not stand equal chance of selection · Most common approach in qualitative studies · Reduce bias- larger sample, more representativeness Conscious and Unconscious Sampling Bias . A biased sample means that those selected are not typical or representative of the larger population that they have been chosen from · Bias is a frequent issue · If, for example, you select every tenth student who enters a uni library, you could not be assured of being representative because different types of students visit the library at different times How to Reduce Bias
October 2 2018 Soci 217 · Representativeness: the quality of a sample having the same distribution of characteristics as the population from which it was selected · Representativeness is enhanced by probability sampling and provides for generalizability and the use of inferential stats · ALL members of a given population should stand an equal chance of being selected for a particular sample. The term for this is equal probability of selection method or EPSEM · Probability sampling offers 2 special advantages: 1. Biases are avoided 2. It allows the researcher to have reasonable expectation that the sample reflects the population Two Types of Error Related to Sampling · Sampling Error The difference between a population value and an estimate of that value derived from a sample O Reduced by 2 Factors: · Sample Size (N) · A homogeneous population o We cannot eliminate sampling error, because when we draw a sample, there will always be some slight difference o We still want to reduce sampling error · Coverage Error All units in a target population not having a known non-zero probability of being included in the sample The sample doesn't cover the full population of interest ... o Go the extra mile to find hard to reach respondents Ex. If the prof gives a survey on student