A screening design is a preliminary evaluation of data to determine which factors are significant and worthy of further study. This allows a researcher to identify the phenomena of interest for the purpose of additional experimentation and research. It is not possible to differentiate between main factors causing a reaction and interactions with screening design, but the technique can assist with the process of finding out where to apply energies in future projects.
In screening design, the researcher wants to identify the most important factors contributing to a phenomenon. For example, a health and human services agency might want to know why low income members of a population tend to experience malnutrition. It can identify a number of factors like poor education, lack of stable access to food, underlying disease, cultural traditions, and so forth. By drilling down into these factors, the agency can identify the most important ones, which is critical for additional research in this area.
Statistical analysis can help with screening design. With the use of statistics, researchers can weigh the various factors in a project and determine which ones are of statistical importance. Something may not be significant when evaluated in this way, indicating that it is not a major factor. In other cases, the analysis may highlight an issue that was not previously identified, indicating a direction for future research.
Like all study designs, screening design has limitations. The inability to determine when a factor is independently causing a phenomenon versus when interactions are the cause is one issue. It is also possible to miss important factors or to overstate rather minor factors that are actually not as important as they may appear. Researchers must consider these issues when they discuss their data, and can provide a thorough analysis of what their findings mean for the benefit of other researchers who want to determine the validity of the results.
This is an example of a preliminary study. No firm findings will emerge from the screening design, but it will provide information for future research. Errors during this process could become costly in the long term, as an organization or researcher may devote substantial energy to following a red herring. For this reason, such studies are designed and executed with care, and researchers evaluate their data critically and aggressively to generate an honest assessment of the utility of their results before taking action on any of their findings.