Good Analytical Research

Abstract

The purpose of this commentary is to address the issues raised by Ohlson from the point of view of analytical accounting research. The aim is not only to provide some input to young researchers who are going to publish good research using analytical methods, but also to give some hints to help users of analytical accounting research to understand and interpret the findings of this type of research. Ohlson has taken on a task of identifying a set of critical factors which are likely to lead to successful research. Good research is defined as research that makes an impression. Thus, it is not enough to get the research published – not even in a premier journal. The research should have an impact, the community should learn something. As Ohlson notes, there is enough ‘ordinary’ research. In my view this is the right attitude. Short-term optimization is also widespread in the research community and that is not what we should strive for. With the objective in place, I will continue to analyze the question in relation to analytical research. I start out discussing the aim of analytical research by providing a few examples of good models. The first is the Feltham-Ohlson model and the second is the agency model. Both are simple and elegant models dealing with difficult issues. The analysis proceeds to characterize good models. A good model is a simple model that zooms in on the problem under scrutiny. It is a ‘minimal’ model that contains the problem and nothing outside the problem. I then proceed to characterize good research in an analytical framework. This is research that tackles a problem that is of interest to the users and the researcher. In this process I also identify current notable analytical research. Finally, I contrast this to the recommendations of Ohlson.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 13 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.