There are a number of methodological practices commonly employed by CALL researchers that limit progress in the field. Some of these practices are particular to replication research, but most are more general and are found throughout the field. I describe in this paper two studies that are fabricated but that resemble much of what is found in published second language research. Each study corresponds to and contains a set of methodological issues. Following each study, I address the issues they illustrate, providing comments and suggestions for how the analyses could be modified to produce greater replicability and/or replicational value. I conclude with a summary of suggestions for quantitative reforms related to improving replication research and quantitative practices more generally in CALL and applied linguistics.