The NIEHS conducted a study on oil spill workers' health by analyzing exposure data. They compared methods for analyzing data with values below detection limits. The β-substitution method performed best in most scenarios, especially with small sample sizes and high variability. The ML method was good for large sample sizes with low variability, while the K-M method worked well with low censoring levels. More research is needed to improve estimation accuracy for small sample sizes and high censoring levels.