Last post was about an approach to accounting research that reduces the isolation that frequently occurs in a serious research department. I introduced the idea of an area based (accounting subject matter, or research methods or research methodology – whatever works for your setting) informal meeting groups that allows members of an area of interest to spend some time together on those interests.
The advantages are huge in my mind:
- it aids doctoral students and new faculty in being socialized into the local community of researchers
- it helps with informal accountability as no one wants to meet and say they accomplished nothing in the three or four weeks between meetings
- it makes us aware of what our colleagues are doing and allows for suggestions about how things could be approached differently
- it allows for a SAFE environment to experiment with early drafts of papers, research ideas etc.
While I know nothing is perfect, our brownbag group is the envy of many.
We can always find an excuse to say we are too busy, but if approached with good will and long term enlightened self-interest, this is an approach worth exploring – even for those rational utility maximizing folks.
When I became a full professor (many moons ago) the first thing I did was adopt the “lab” approach that is common in psychology. The principle is simple: researchers with a given area of interest commit to meeting together at defined intervals to discuss what they are working on and a detailed enquiry into some specific piece of work.
At Queen’s we adopted the research methods approach to organizing (alternatively one could organize by subject matter) and our “lab” is called the “social and behavioral accounting brownbag.” It is composed of the four faculty in the area and the six doctoral students as well as any visiting faculty, adjunct faculty or others that are interested in field or experimental research based on economic, psychology or sociology theories. We commit to meeting four or five times a term, normally at intervals of three weeks.
A typical “lab” meeting is done over lunch, for two hours, and starts with a roundtable “update” by each person about what has been occupying their academic time (teaching/taking classes, research and service) since the last time we met (i.e. “the appetizer course”).
Then we delve into the “main course” that could be discussing a working paper by one of the members, previewing a presentation that a member wants to get feedback on (e.g. proposals, road shows), discussing a preliminary idea for carrying out research or discussing a particularly interesting article that someone suggests we might enjoy reading.
One of the keys to success is to be inclusive. Everyone talks because they are asked to! Further, we separate the discussion into two phases – what we like about the subject matter and then, after all have spoken on our likes – we move onto what accountants do best, constructive criticism. To me, this two phase approach is an article of faith – when we drop it – we lose the constructive nature of the discussion. Further, it makes us think deeply about what we like about the research – something that we have a hard time to do in accounting!!!
In my last post I noted the growing trend from editors and reviewers of positivist field research in accounting to require that it contribute to “theory”. In my mind that is an inappropriate methodological (not methods) importation from interpretive field research in accounting.
WHY? In interpretive research, the dance between theory creation and research reporting is an ongoing one as reality is socially constructed. So some argue it is natural that research reporting should engage with a dialogue with theory. In other words, it should not be a one way street given the social construction of reality argument.
However, positivist field research while understanding to some extent that argument, is focused on using theory to help make tentative predictive statements about the accounting world. Positivists believe that there are regularities out there that can be discovered via research and predictions developed via theory. Here, while there might be a dance between theory and research findings, the normal science approach does not require it except in cases of anomalies.
So editors and reviewers that are insisting on theory contributions from positivist field research are missing the point. Positivist field research is about understanding in light of theory leading to tentative predictions. Requiring more, as is often being done these days, is in my mind completely inappropriate.
That’s my thought!!!
Readership is up and I thank you for reading these musings!!!
In an attempt to be more user friendly I have added buttons to allow followers to receive updates by email. Thanks for the suggestions from readers to do this!!!!
I also have made it easier to post blog content to social media, especially to LinkedIn and Twitter.
I have been noticing a real problem with evaluation of field research in accounting that accepts positivist view of the world (i.e. that there is something out there that we can find and reliability report on in such a way that it may lead to predictions about general behavior within a context). This is due to editors of these pieces taking on the practice that I established at CAR of involving one interpretivistic researcher and one substantive area researcher in evaluating manuscripts.
The idea of the pairing of researchers as reviewers on positivist field research was good in the early days (in my biased view given I invented it and preached about it). Positivist field research had much to learn about research methods from interpretive researchers and reviewers. While there is still things to learn, a great amount of positive transfer has occurred and some transfer has also gone back to interpretive research.
One flaw in this review practice is emerging across all journals that encourage field research, that flaw is requiring positivist field research in accounting to contribute to theory. This, I think, can be attributed to the interpretive accounting field research community focusing more on a contribution to theory for studies within their domain (and while I think this is problematic I leave it to those researchers to back themselves out of the corner I think they are creating for themselves). That transfer of norms is in my mind an inappropriate methodological importation (NOT methods) and unfortunately many editors of these papers come from interpretive backgrounds and are somewhat oblivious to the harm it creates.
In the next post will discuss the harm more explicitly and what needs to be done about it.
Increasingly I am seeing claims in experimental papers about their “contribution to theory”, normally psychology. I find this strange, except for when the contribution is about a boundary condition to psychology theories.
If we believe the accounting setting is worthy of study, then that means that there must be something unique about the human activities in the setting that causes us to doubt that psychology research can be simply transferred and applied to the setting. Indeed, one of the critical points for psychology research in accounting came when Katherine Schipper, in a letter that is one of the most misunderstood missives in the history of accounting research, called for accounting researchers to stop simply transferring and applying psychology research (especially heuristics and biases research) to the accounting setting. In essence, she was willing to accept that accountants and auditors are human and if you transfer and apply psyhcology theory to a setting where there is no countervailing force, auditors and accountants are going to act as humans.
But what is interesting about the accounting and auditing setting is that they are often implicitly designed to try and reduce/eliminate these biases (one cause of ineffectiveness) while recognizing where the heuristics work well (one cause of efficiency). Of course, by trial and error systems were put in place to achieve that goal. As psychology based researchers we are attempting to see if that goal has been achieved (among others).
Hence, I find that most claims to contribute to psychology theory based on accounting specific research are problematic. For if the accounting researcher is studying a setting where theory applied as tested can be generalized to all people, then by definition, the researcher is not studying a setting in the accounting context. So when I read about contributions to psychology theory, it starts me thinking about whether this research is within the domain of accounting at all. Of course, I am always open to an argument to the contrary but a “throwaway line” about contributions to psychology by itself is more of a “red flag” than anything else.
That is my take anyhow.
Every now and again I get papers submitted to BRIA that are based on agency theory, signalling theory, and other microeconomic theory that is tested based on publicly available archival data. Rarely, I get challenged by authors you say, “you examine these in the experimental lab in BRIA, why not archival???”
A perfectly fair question. In my mind the answer is in two parts. First, BRIA is foremost about human behavior, it is not about theories of human behavior based on a set of assumptions that even their creators admit are significantly oversimplified. The school of Milton Friedman et al evaluates the success of these theories based on their predictive validity in actual settings and state the assumptions about human behavior in their theories do not matter if they can predict well enough (albeit “well enough” seems a very slippery concept to me). BRIA and similar journals respect that tradition, but it is not he basis for articles that we are interested in publishing. There are many other outlets for such papers.
Second, why does BRIA publish experimental economics pieces? While I cannot speak for all concerned, the difference between archival publicly available data and laboratory experiments reduces substantially the number of steps one has to take between the theory and the actual human behavior. Experimental economics allow us to examine the factors that need to be in place in order for actual individuals to respond to economic incentives in those theory settings. In other words, it stripes away the veneer that these simple assumptions about human behavior in economic theory are easy to implement and that they invoke a whole set of assumptions about the context where the decision is made.
Anyhow, that is my take on the issue.