Models of Psychometrics and SamplingAs the science of test theory and testing, psychometrics has two models, each of which is supported by a different set of assumptions. Based on a researcher's goals

Populations and Samples PSY3700 Multimedia Assessment and Psychometrics ©20 16 South University 2 Populations and Samples Objectives of Testing Sampling is a concern in survey research because sample data are meaningless until framed within the context of the population of interest. Unless the sample is representative of that population, data generated may not serve the objectives of the research study (Sue & Ritter, 2013). To draw a sample, we first need a list of all members of a population, which is called the sample frame.

For some populations, it is possible to obtain a complete list of all members; for most, this is not possible.

When you c an secure a sampling frame, you can cull a sample from that group. However, this sample may not be the actual group from whom you will receive data. Potential respondents from the culled sample may be unavailable, may refuse to participate, may be ineligib le, or may drop out from the study. To address this problem, you must initially recruit more participants than needed (Sue & Ritter, 2013). A variety of probability and nonprobability techniques for sample selection exist and should be considered in the p rocess of survey planning. Probability sampling affords the benefit of allowing the researcher to generalize about the population based on data from the sample, but contains a margin of error that can influence the representativeness of any one sample draw n from the population. Therefore, an adequate sample size should be calculated beforehand to reduce the margin of error. Nonprobability sampling, often used in exploratory research, does not carry the same sample -to- population generalizability as probabil ity sampling. Because most respondents of nonprobability samples have self -selected, error estimates are meaningless because we have no knowledge about variability among participants drawn from the population. Thus, choosing an adequate sample size is more taxing. Generally, it is suggested that the sample size be approximately 10% of the population of interest (Sue & Ritter, 2013). PSY3700 Multimedia Assessment and Psychometrics ©20 16 South University 3 Populations and Samples Objectives of Testing References Sue, M., & Ritter, L. (2013). Conducting online surveys (2nd ed.). Los Angeles, CA: Sage. © 201 6 South University