Methodological appraisal: Two examples

摘要: Drawing an inference about empirical situation requires posing a problem or asking question, collecting relevant data and choosing methods to analyse them. Though ideally under the spell of muse objectivity (but unrealistically) one imagines that should be determined by objective alone, former can depend on various aspects entire process precedes it. This is brought out two examples from earth science literature. They pertain seismicity classification based pattern recognition algorithm early palaeomagnetic studies. The are deliberately chosen disjoint fields so it clear not contribute any field science, but make essentially methodological points. These are: (i) no culmination lineal sequential process, part it; (ii) Therefore, return there preceding revise (iii) All intermediate secondary inferences, all disagreements between what was expected priori happened way symptoms effects format questions inferences critically scrutinized; (iv) total exercise viewed as only drawing external world, also learning doing so. latter experience generally transferable other situations.