The Classroom Argumentation Observation Protocol developed in LAMP measures teachers’ pedagogical practices in terms of teachers providing opportunities for students to engage in the mathematics learning experiences specified in the logic model. The protocol provides quantified scores for the types of claims a teacher uses, the explicitness of claims, the sophistication of the warranting, and the use of warrants and data. Open‑ended questions ask for the extent to which the observed lessons address LLAMA lesson objectives. LAMP established content validity of this protocol through an expert panel. At the onset of the LLAMA project, the protocol was revised. The primary revisions include (a) the inclusion of our recent understanding of generic example arguments (see Yopp and Ely, 2016 and Yopp, Ely, and Johnson‑Leung, 2016) and (b) asking about the percentage of students engaged in the classroom argumentation episode. The protocol was further revised during weekly Principal Investigator meetings and was piloted in Year 1. During Year 2, the research teams participated in an observation training to ensure interrater reliability among observers and maintain a codebook of decision rules pertaining to the coding of the observations. Minor modifications were made to the protocol during this training period. The team watched videos and scored the videos during multiple sessions and modified the rubric and wording accordingly, following those sessions.