Simulated datasets of "Measurement error in continuous endpoints in randomised trials: problems and solutions"

Files with results of the simulation study in "Measurement error in continuous endpoints: problems and solutions". <br>Four different types of files are saved in this folder, all consisting of 10,000 rows (number of replicates for each scenario):<b>1.</b> Results for the systematic measurement error scenario (prefix = sme). Each file has 12 columns: alpha_hat (estimated regression intercept after correction), beta_hat (estimated regression slope after correction), ci.zv.l (zero variance confidence interval lower limit), ci.zv.u (zero variance confidence interval upper limit), ci.delta.l (delta ci l limit), ci.delta.u (delta ci u limit), ci.fieller.l (fieller ci l limit), ci.fieller.u (fieller ci u limit), ci.bootstrap.l (bootstrap ci l limit), ci.bootstrap.u (bootstrap ci u limit), boostrap.R (number of replicates of bootstrap), seednr (number of seed).<br><b>2.</b> Results for the differential measurement error scenario (prefix = dme). Each file has 12 columns identical to the sme scenario.<b>3.</b> Results for the naive scenario of both systematic and differential measurement error (no measurement error correction applied) (prefix = naive_dme or naive_sme). Each file has 4 columns: alpha_hat, beta_hat, ci.l, ci.u and seednr.<b>4.</b> Results containing the estimated theta's that are used to correct for systematic measurement error (prefix = sme_thetas). Each file has 3 rows: theta0 (estimated theta0 = intercept of sme model), theta1 (estimated theta1 = intercept of sme model), seednr.<br>File names are constructed as: prefix_k_rsquared, with prefix as above, k the number of people in the external calibration set and rsquared the r-squared between y* (y with error) and y. <br>