Nonprobability (or design-free) surveys are becoming more common place based on changes in the definition of fit for purpose for many studies. Nonprobability surveys are touted as offering both increased speed in obtaining data on emerging issues (e.g., an opt-in web survey) and decreased costs compared with probability-based surveys. However, evaluation studies have shown that many nonprobability estimates are biased because of errors associated with coverage, selection, and model misspecification. Calibrating design-based survey weights to control totals estimated from other surveys has been implemented for years. Referred to as estimated-control (EC) calibration, this technique has been shown to reduce bias for design-based estimates beyond levels seen when calibrating to typical controls alone such as demographic characteristics.By comparison, propensity score adjustments (PSA) are used to calculate estimates from nonprobability surveys, and may include questionnaire items (e.g., attitudinal questions) as model covariates. However, research to date on PSA shows mixed results. This research builds on prior results related to weighting for nonprobability surveys. We compare the reduction in estimated bias for PSA and EC calibration techniques through an empirical study, noting that the latter methodology has many benefits. Applications are applied where appropriate to a nonprobability survey of young adults enrolled in a smoking cessation program. Next, we outline steps in calculating precision for the estimates when using EC calibration. The discussion concludes with out next steps in the study of nonprobability surveys.