Concentration models calculation

There are a number of concentration model types are available. A basic distinction is between using the empirical concentration data (empirical model), fitting a statistical model to the concentration data (parametric model), or to construct a model from (conservative) limit values. Settings relevant for some of these model types as well as other settings are described under concentration model settings.

Concentration data from Total Diet Studies does not include variability information. In order to include variability in the concentration models created for TDS samples, the variance of the TDS concentrations may be estimated using variances of the foods making up the composite TDS food.

• Concentration model types
  • Empirical model
  • Censored spike lognormal model
  • Censored-Spike Truncated lognormal model
  • Censored Lognormal model
  • Zero-spike censored lognormal model
  • Censored spike MRL model
  • Summary statistics model
• Concentration models
  • Mixture zero spike and censored lognormal
  • Censored lognormal
  • Mixture censored spike and truncated lognormal
  • Mixture censored spike and lognormal
• Imputation of non-detect measurements
  • No imputation
  • Impute all censored values by factor times LOQ
  • Impute censored values based on authorized uses
  • No imputation except for authorized uses
• Deriving the variance of TDS samples from monitoring