Assume that you do not know anything about the true relationship between motivation and weight loss. Which of the following strategies, if any, might be successfully employed with the data observed by the researchers in order to produce valid causal estimands? Choose all that apply. Compute the difference in mean outcomes between the treatment and control groups Regress the outcome on the treatment indicator, adjusting for X Matching or forming blocks of observations based on propensity scores "Changing the estimand" to instead reflect some sub-population truth Trim the observations so that only areas of the covariate space where there is overlap are considered None of the above