Description Usage Arguments Value Author(s) Examples

This function fits several measurement error models to model the relationship between the test measurement (error-prone measurement) and the reference measurement (true measurement) in a validation set.

1 2 |

`MeasError` |
object of class MeasError |

`differential` |
vector containing the differential measurement error variable |

`model` |
optional character string or vector of character strings indicating the assumed measurement error model: "all" (default), "cme" (classical measurement error), "sme1" (systematic measurement error with zero intercept), "sme2" (systematic measurement errror with non-zero intercept) and "dme" (differential measurement error). |

`robust` |
boolean indicating whether robust standard errors need to be calculated, the "HC3" robust standard errors from vcovHC are used for the heteroskedasticity-consistent estimation of the covariance matrix of the coefficient estimates. |

`plot` |
boolean indicating whether one wants plots of the residuals (one plot for each tested model) |

`mefit`

returns an object of class "mefit".

An object of class `mefit`

is a list containing, depending on the models tested for, the following components:

`cme` |
a list containing the coefficients, sigma, df and model of the classical measurement error model |

`sme1` |
a list containing the coefficients, sigma, df and model of the systematic measurement error model with zero intercept |

`sme2` |
a list containing the coefficients, sigma, df and model of the systematic measurement error model with non-zero intercept |

`dme` |
a list containing the coefficients, sigma, df and model of the differential measurement error model |

`lrtest1` |
a list containing the likelihood ratio test of, depending on which models are tested, cme vs sme1 vs sme2; cme vs sme1; cme vs sme2; sme1 vs sme2 |

`lrtest2` |
a list containing the likelihood ratio test of, depending on whether one tests for differential measurement error, sme2 vs dme |

Linda Nab, [email protected]

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 | ```
##measurement error in endpoint
X <- c(rep(0, 50), rep(1, 50))
Y <- X + rnorm(100, 0, 1)
Vcme <- Y + rnorm(100, 0, 3) #classical measurement error (cme)
Vsme <- 1 + 2*Y + rnorm(100, 0, 3) #systematic measurement error (sme)
Vdme <- 2 + 2*X + 3*Y + 2*X*Y + rnorm(10, 0, 3*(1-X) + 2*X) #systematic differential measurement error (dme)
##measurement error in exposure
X <- rnorm(100, 0, 1)
Y <- 1.5*X + rnorm(100, 0, 0.5)
W1 <- X + rnorm(100, 0, 0.5)
W2 <- X + rnorm(100, 0, 0.5)
fit1 <- mefit(MeasError(Vcme, Y), plot = T)
fit2 <- mefit(MeasError(Vsme, Y), plot = T)
fit3 <- mefit(MeasError(Vdme, Y), X, robust = T, plot = T)
fit4 <- mefit(MeasError(W1, X), model = "all")
#fit5 <- mefit(MeasError(cbind(W1, W2), NA)) #not yet supported
``` |

Embedding an R snippet on your website

Add the following code to your website.

For more information on customizing the embed code, read Embedding Snippets.