There have been many statistical advances to address measurement error in the past few decades. With our activities we intend to raise the awareness for the topic and to overcome barriers to using measurement error adjustment methods, e.g.
Lack of, or problems with data for determining the measurement error model and estimating its parameters
Inadequate standards of "validation" of measurement instruments
Lack of appreciation of the biases and the loss of precision resulting from measurement error
Lack of software for implementing methods of adjustment
eICDAM, February 8-12, 2021: Categorizing variables measured with error (H. Boshuizen, slides)
eICDAM, February 8-12, 2021: Using a prediction equation variable as an outcome in a regression model is dangerous! (L. Freedman, slides)
eICDAM, February 8-12, 2021: Understanding and adjusting for the impact of Berkson error arising from prediction equations in nutritional and physical activity epidemiology (P. Shaw, D. Sotres-Alvarez, L. Freedman)
Article on the general problem of measurement error in epidemiology intended for the intelligent layman
Literature survey about current practice for addressing measurement error in four areas of epidemiology in which measurement error is known to be extensive
Main document on guidance on dealing with measurement error and misclassification intended for biostatisticians
Projects that are currently ongoing include the construction and use of prediction equations, handling error-prone continuous variables that have been categorized, a case study on how to do measurement error correction in practice and error in measurements of dietary intake used in nutritional epidemiology.
Contributions to workshops and conferences in order to increase the awareness of the topic
Collaborations with other topic groups and participation in committees of the STRATOS initiative