Academic writing should be organized logically, as this chapter is. Academic research, on the other hand, is part of life, and therefore is not logically organized by nature, however much researchers might wish it were. Although we try to organize our research logically, we have to gather data whenever, wherever, and however we can. Thus we must be ready to modify our research and writing plans according to contingency, and always be ready to improvise.

Researchers put order into the confusion of information we have gathered from various sources at various times, in order to tell a story from, or answer a question with, that information in ways that make sense to readers and listeners. This chapter is organized logically, to explain the stages of data collection and presentation in a logical order, but in fact the logic of real life research is rarely so neat and tidy. Instead, each stage blends into another, overlapping, with unexpected gaps and breaks throughout the process. At any moment the researcher must be ready and willing to jump into any other stage of research and writing, whether in logical sequence or not.

Scholarly research in any field involves collection of data, analysis of those data according to an algorithm or method, and derivation of conclusions based on methodical analysis of the data. This is true whether the field is English literature or astrophysics. The data are not necessarily facts, of course. If one works in literature, one's data are fiction, not the facts that are the obsession of the historian. Nonetheless, even when the methods, data, and purposes of research are different in different fields, the basic process is the same.

The logical order of research is also the same: decide on a topic, gather data, evaluate the data, organize the data and, finally, write up the results. But of course the actual process of producing history is more complicated and messy, and the chronological order of the stages is almost always very mixed, rarely following this logical order exactly.

Because natural scientists use generally accepted methods, they can expect re-analysis or reproduction of the same data to produce the same conclusions, a process referred to as replication and an important means of testing scientific findings. On the other hand, although certain historical methods, such as source criticism, are generally accepted, different historians using the same data may come to radically different conclusions.