In the linked article regarding TOMS shoes: "While statistical significance is nowhere to be seen, by dint of their consistency the results are reasonably convi

In the linked article regarding TOMS shoes: "While statistical significance is nowhere to be seen, by dint of their consistency the results are reasonably convincing." That's not how statistical significance works; if you don't reach it, you can't say anything about whether or not there is an effect. The actual research paper is here: http://repository.usfca.edu/cgi/viewcontent.cgi?article=1005&context=econ http://repository.usfca.edu/cgi/viewcontent.cgi?article=1005.... In it, the authors say "After controlling for household and parental characteristics as well as ADP-level fixed effects, we can uncover no statistically significant negative impact on shoe purchases in local markets."