This study demonstrates new methods of identifying and interacting with respondents who introduce high levels of measurement error in web surveys. JavaScript, a browser-incorporated scripting language, was used to translate Cronbach's Alpha, Spearman-Brown's Split-half reliability coefficients, and two different applications of within-subject standard deviation of responses, into reusable dynamic functions to be incorporated in web instruments. When respondents exhibit insufficient reliability in responses, high deviation on homogenous sets of items, or low deviation on heterogeneous sets of items, the functions embedded in the web instrument (which perform calculations instantaneously without communicating data to the server) trigger a message prompting for higher attentiveness in selecting responses. A study was undertaken to test the methods' effectiveness in which a previously validated instrument was administered to employees in a US office of an international human capital firm. Due to the small sample size and low incidence rate of inattentive respondents, structural equation modeling could not be used to evaluate changes in measurement error and factor structure. Instead, a qualitative approach was undertaken and comments and behavior from prompted respondents were examined. The results were not conclusive, but supportive of the use of the proposed method if liberal cutoff values are used in the formulas.