The USPTO designed the intervention as a randomized control trial, assigning incoming pro se patent applications randomly to either the treatment group, i.e., to the PSPEU, or the control group, the standard patent examination process. Since the USPTO does not collect information directly from applicants on whether they have legal representation, the PSPEU identified potential pro se applications using a search query executed on internal USPTO application data. The approach identified all applications that (1) were not yet assigned to an examiner, (2) did not have a power of attorney communication filed, and (3) were new applications, in the sense that the applications were not continuations, continuations in part, divisionals, reexaminations, requests for continued examination or reissues.
The RCT had two rounds. RCT Round 1 took place in October 2014, randomly assigning applications to either the treatment or control groups (using stratified randomization within the initial technology centers assigned to each application). RCT Round 2, which used the same assignment mechanism, was in October of 2015.