MEM: Regression in model with errors

In Kaique-fhs/Error: Model with errors in variables

Regression in model with errors

Description

MEM is used to adjust models with errors in the variables. It can be used to perform regressions on variables that have measurement errors.

Usage

MEM(
  x,
  y,
  lambda_x,
  Correction = FALSE,
  method = "Bartlett",
  conf.level = 0.95,
  ...
)

Arguments

x	Values vector
y	Values vector
lambda_x	Ratio of variances between values and error
Correction	Small sample correction, use TRUE or FALSE
method	The correction method that will be considered "Bartlett", "B1" or "B2"
conf.level	Significance level for the hypothesis test
...	Not always use all arguments

Details

Only use the conf.level, beta_til e method arguments when running equals "Y", otherwise these arguments will not affect the results.

Value

Estimated Parameters x e y.

Author(s)

Kaíque Ferreira Henrique de Souza

Tatiane Ferreira do N. M. da Silva

References

Bartlett, M.S. (1937), Properties of sufficiency and statistical tests,Proceedings of RoyalSociety of London A, 160, 268-282.

Fuller, S. (1987),Measurement Error Models. Wiley, New York.

Melo, T.F.N., Vasconcellos, K.L.P., Lemonte, A.J. (2009). Some restriction tests in a newclass of regression models for proportions.Computational Statistics and Data Analysis, 53,3972–3979.

See Also

lm

Examples

x = seq(1,100,length.out = 100)
y = sort(runif(100,1,10))

MEM(x, y, lambda_x = 2 )
MEM(x, y, lambda_x = 1, conf.level = 0.95 ,Correction = TRUE)

Kaique-fhs/Error documentation built on June 6, 2023, 4:43 p.m.