Specifically, Allan Franklin is concerned with two problems in the use of experimental results in science: selectivity of data or analysis procedures and the resolution of discordant results.