In this paper we show that omitted variables and publication bias lead to severely biased estimates of the value of a …

winsorizing may induce or exacerbate bias. We term this source of bias Iatrogenic' (or econometrician induced) error. The …

This paper develops and applies semiparametric econometric methods to estimate the form of selection bias that arises … econometric selection model which represents the bias solely as a function of the probability of participation; and (c) the method … of difference-in-differences. Using data from an experiment on a prototypical social program combined with unusually rich …

This paper considers causal inference and sample selection bias in non-experimental settings in which: (i) few units in …-treatment characteristics. We propose the use of propensity score matching methods and implement them using data from the NSW experiment … and compare the estimates of the treatment effect obtained using our methods to the benchmark results from the experiment …

Field experiments rely heavily on self-reported data, but subjects may misreport behaviors, especially sensitive ones …

to selection bias and a partial solution to publication biases. Second, it discusses various ways in which randomization …

and show they behave well in realistic scenarios, correcting the large bias problem of the full sample estimator. We use … data from the National JTPA Study and the Tennessee STAR experiment to demonstrate the performance of alternative …

teachers' classrooms, prior research has largely focused on the validity of value-added measures. We conduct an experiment … two previous experiments, we provide further evidence that value-added measures are unbiased predictors of teacher …

need for a nonlinear utility-of-wealth function, and instead results from a sort of perceptual bias — but one that … numerical cognition, and test its ability to explain the choice frequencies that we observe in a laboratory experiment …

experiments to isolate the role of one such mechanism—ethnic preferences—which has been central in both theory and in the … relative to non-coethnics. The results run strongly against the common presumption of extensive ethnic bias among ordinary …