The Professional Counselor | Volume 12, Issue 3 219 Sampling Procedure and Sample Size Sampling procedures should be clearly stated in the Methods section. At a minimum, the description of the sampling procedure should include researcher access to prospective participants, recruitment procedures, data collection modality (e.g., online survey), and sample size considerations. Quantitative sampling approaches tend to be clustered into either probability or non-probability techniques (Creswell & Creswell, 2018; Leedy & Ormrod, 2019). The key distinguishing feature of probability sampling is random selection, in which all prospective participants in the population have an equal chance of being randomly selected to participate in the study (Leedy & Ormrod, 2019). Examples of probability sampling techniques include simple random sampling, systematic random sampling, stratified random sampling, or cluster sampling (Leedy & Ormrod, 2019). Non-probability sampling techniques lack random selection and there is no way of determining if every member of the population had a chance of being selected to participate in the study (Leedy & Ormrod, 2019). Examples of non-probability sampling procedures include volunteer sampling, convenience sampling, purposive sampling, quota sampling, snowball sampling, and matched sampling. In quantitative research, probability sampling procedures are more rigorous in terms of generalizability (i.e., the extent to which research findings based on sample data extend or generalize to the larger population from which the sample was drawn). However, probability sampling is not always possible and non-probability sampling procedures are rigorous in their own right. Readers are encouraged to review Leedy and Ormrod’s (2019) commentary on probability and nonprobability sampling procedures. Ultimately, the selection of a sampling technique should be made based on the population parameters, available resources, and the purpose and goals of the study. A Priori Statistical Power Analysis. It is essential that quantitative researchers determine the minimum necessary sample size for computing statistical analyses before launching data collection (Balkin & Sheperis, 2011; Sink & Mvududu, 2010). An insufficient sample size substantially increases the probability of committing a Type II error, which occurs when the results of statistical testing reveal non–statistically significant findings when in reality (of which the researcher is unaware), significant findings do exist. Computing an a priori (computed before starting data collection) statistical power analysis reduces the chances of a Type II error by determining the smallest sample size that is necessary for finding statistical significance, if statistical significance exists (Balkin & Sheperis, 2011). Readers can consult Balkin and Sheperis (2011) as well as Sink and Mvududu (2010) for an overview of statistical significance, effect size, and statistical power. A number of statistical power analysis programs are available to researchers. For example, G*Power (Faul et al., 2009) is a free software program for computing a priori statistical power analyses. Sampling Frame and Location Counselors should report their sampling frame (total number of potential participants), response rate, raw sample (total number of participants that engaged with the study at any level, including missing and incomplete data), and the size of the final useable sample. It is also important to report the breakdown of the sample by demographic and other important participant background characteristics, for example, “XX.X% (n = XXX) of participants were first-generation college students, XX.X% (n = XXX) were second-generation . . .” The selection of demographic variables as well as inclusion and exclusion criteria should be justified in the literature review. Readers are encouraged to consult Creswell and Creswell (2018) for commentary on writing a strong literature review. The timeframe, setting, and location during which data were collected are important methodological considerations (APA, 2020). Specific names of institutions and agencies should be masked to protect
RkJQdWJsaXNoZXIy NDU5MTM1