Academic journal article Environmental Health Perspectives

A Survey of Laboratory and Statistical Issues Related to Farmworker Exposure Studies

Academic journal article Environmental Health Perspectives

A Survey of Laboratory and Statistical Issues Related to Farmworker Exposure Studies

Article excerpt

Developing internally valid, and perhaps generalizable, farmworker exposure studies is a complex process that involves many statistical and laboratory considerations. Statistics are an integral component of each study beginning with the design stage and continuing to the final data analysis and interpretation. Similarly, data quality plays a significant role in the overall value of the study. Data quality can be derived from several experimental parameters including statistical design of the study and quality of environmental and biological analytical measurements. We discuss statistical and analytic issues that should be addressed in every farmworker study. These issues include study design and sample size determination, analytical methods and quality control and assurance, treatment of missing data or data below the method's limits of detection, and post-hoc analyses of data from multiple studies. Key words: analytical methodology, biomarkers, laboratory, limit of detection, omics, quality control, sample size, statistics. Environ Health Perspect 114:961-968 (2006). doi:10.1289/ehp.8528 available via http://dx.doi.org/ [Online 16 February 2006]

**********

Developing internally valid, and perhaps generalizable, farmworker exposure studies is a complex process that involves many statistical and laboratory considerations. Statistics are an integral component of each study beginning with the design stage and continuing to the final data analysis and interpretation. Similarly, data quality plays a significant role in the overall value of the study. Data quality can be derived from several experimental parameters, including statistical design of the study and quality of environmental and biological analytical measurements. Because these issues are so intricately intertwined and affect almost all aspects of the study, we chose to discuss these issues together. This survey of issues is intended as a guide or resource for epidemiologists, exposure assessors, and others who plan to undertake a study involving a highly mobile farmworker population.

A large number of different statistical issues arise in the development and analysis of farmworker pesticide exposures and resulting data. These discussions can be generally categorized as relating to a) sampling and design (i.e., pre-data collection) issues; b) laboratory quality issues; c) missing or misclassified data and subsequent biases, including limits of detection (LODs; i.e., primarily preanalysis issues); d) data reporting and reliability issues (i.e., primarily descriptive statistical issues); e) extensions to standard statistical models and analyses (i.e., primarily inferential statistical issues); and f) post hoc analyses and combining study results. It is also important to use the proper interpretation and context for analysis, including consideration of scientific information and understanding the direction of test statistics.

Statistical Issues at the Design Stage

A primary concern of studies associated with farmworker exposures is choosing the appropriate sampling strategy. Ideally, the objective of any sampling strategy is to collect data that are representative of the study population (Rothman and Greenland 1998). However, representativeness is often difficult to obtain. Although statistical representativeness is certainly a desirable attribute, it is neither a necessary nor a sufficient condition for a well-designed investigation. For example, case-control studies are rarely statistically representative. For cases in which very little information is available, it may be impossible to know whether the study is representative of the population. In such conditions, an exploratory investigation is warranted. Convenience or anecdotal samples may give very useful information. Because responses from convenience samples are likely to be better than that from a representative sample, they may actually be more "representative." Data gathered from such an investigation can prove invaluable in developing a better design in a representative sample. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.