Whether a study is exploratory or confirmatory, experimental or nonexperimental, many resources are needed to develop and carryout a study or project. One question that often arises is: how big of a sample do I need for my study? Research has shown that certain rules of thumb (e.g., 10 participants per variable in multiple regression analysis) are problematic, given the various circumstances that may be unique for any given study. Thus, determining the sample size is a necessary step in the data analytic process so as to minimize: (1) oversampling and thus wasting resources or (2) undersampling and compromising interpretation of the data. Services offered in this area include:
- Power analysis to assess proper sample size given desired parameters (e.g., effect size, level of significance, etc) for the simplest univariate tests (e.g., two group design) up to multivariate designs (e.g., MANOVA, mixed ANOVA, Structural Equation Modeling).
- Sample size estimation for survey projects based on desired confidence levels and margin of error