I Endorse This Message: Learning From Endorsement Experiments
Blair et al. 2013 report results from an endorsement experiment in Pakistan. Respondents are asked their level of support for four policies (on a 5-point scale, rescaled to [0,1]). In the treatment group, each policy is randomly paired with an endorsement from one of four militant organizations. The control group receives no endorsement. The difference in average support between treatment and control reflects respondents’ attitudes toward the endorsing groups.
They summarize their topline results as follows: "We find that Pakistanis in general are weakly negative toward Islamist militant organizations, as shown in Table 1."
The summary may be too rosy.