Resumo:
Propsito do Trabalho:
In this paper I analyse 12 experimental researches published in Brazilian accounting journals from 2006 to 2012, in order to develop an epistemological critique of them. Firstly, I delineate the main characteristics of experimental research designs, the main validity threats they face, and discuss how the 12 articles under analysis dealt with these threats. Secondly, I contrast the positivist theoretical perspective common to these articles, characterised by a realist ontology and an objectivist epistemology, with constructionist conceptions of social sciences which claim that meaning is constructed (instead of discovered) out of the interaction between subjects and objects, between human beings and their world. Then I formulate a critique of these articles inspired by constructionist notions of research. The paper aims at contributing to improve the quality of Brazilian accounting research by alerting future researchers to the challenges they face to conduct proper experimental designs, showing how problematic experimental accounting researches can be from an epistemological point of view, and proposing a cross-paradigmatic debate which can lead the academic community to a higher level of maturity.
Base da plataforma terica:
The argument of the paper is developed in two steps. In a first moment, I analysed the experimental designs employed in 12 articles published in Brazilian accounting journals according to their own rationale. This analysis relies on descriptions of experimental research designs presented by Gall, Borg, and Gall (1996), and Smith (2003). According to Smith (2003), experiments are particularly suited to research questions that investigate causal relations between variables (p. 100), aiming to allow a researcher to control all possible variables which can affect a phenomenon, except from the ones he/she manipulates, thus allowing him/her to claim that the measured variations in this phenomenon are due only to the manipulated variables, i.e., the experimental treatments. He identifies three kinds of validity threats to experiments: construct validity, or the extent to which abstract concepts are successfully operationalised into reliable measurements; internal validity, i.e., how well the research design rules out the effects on the dependent variable of factors other than the experimental treatments; and external validity, the extent to which the findings of an experiment are applicable to individuals and settings other than those that were studied. Secondly, I contrasted the positivist theoretical perspective common to these articles with constructionist conceptions of social sciences. The arguments presented in this part of the paper draw mostly on Crotty (1998) and Jrgensen and Phillips (2002). Crotty (1998) presents as characteristics of positivism a realist ontology which affirms the existence of a real world of objects, and an objectivist epistemology which claims that these objects not only exist by their own, but also have meaning prior to, and independently of, any consciousness of them (p. 27), whereas costructionism, for him, is the view that all knowledge, and therefore all meaningful reality as such, is contingent upon human practices, being constructed in and out of interaction between human beings and their world, and developed and transmitted within an essentially social context. (p. 42). Jrgensen and Phillips (2002) point to key premises of constructionism: a critical approach to taken-for-granted knowledge; historical and cultural specificity; a link between knowledge and social processes; and a link between knowledge and social action.
Mtodo de investigao:
Resultados, concluses e suas implicaes:
In the first part of this paper, I analysed the experimental designs employed in 12 articles published in Brazilian accounting journals according to their own rationale. In general, this analysis revealed a lack of evidences about the validity of the constructs employed to measure abstract concepts, a difficulty to design internally valid experiments, and an inability to express confidence on the applicability of the results to organisational settings in terms other than as acts of faith. The aim of this part of the paper is to alert future researchers to the challenges posed by experimental designs and call on them to give a tougher consideration to these issues.
In a second moment I also discussed the epistemological foundations of experimental researches to show how problematic they are. I argued that the experimental designs employed in these 12 articles rest on a positivist theoretical perspective which underestimates the social aspects of the human experience. Thus, they are marked by a behaviourist approach which virtually ignores these aspects, focusing on individual behaviour instead; a reified notion of subjectivity, seeking to describe a sort of transcendental human nature situated in a vacuum of culture and history; a disregard for culture and history characterised by the lack of a contextual embeddedness of the investigations; and a commitment to render people controllable, evidenced by the common articulation of research objectives in terms of promoting managerial efficiency rather than social justice.
A good point against this second part of the argument could be that I am criticising the articles for not being something they are not meant to be. However, none of these articles clearly articulates what they mean, i.e., the epistemological foundations on which they rest, what is symptomatic of the taken-for-grantedness of their positivist premises amongst their intended audience, the academic accounting community in Brazil. But I believe that an explicit consideration of these issues can also contribute to improve the quality of future researches, and that cross-paradigmatic debates such as the one I propose in this paper can lead the academic community to a higher level of maturity.
The extent to which my critique of these 12 articles also applies to other researches is a matter of specific consideration. In terms of coverage, I believe that most of the experimental accounting researches ever published in Brazilian journals were included in this sample, even though I probably have missed some. But as much of my criticism is centred on the positivist groundings of these papers, they probably hold for the majority of the accounting research which is being conducted in Brazil as well.
As this paper aims at contributing to enhance the quality of Brazilian accounting research, a last consideration I present is about the very comprehension of the purpose of accounting expressed in all these articles, defined as to provide useful information for economic decision-making or some variation of that. Such a restricted view of accounting not just underestimates its social and organisational roles, but also prevents accounting research from being innovative. I suggest that future researchers need to see accounting in a more elaborated way, considering many more aspects of its operation.
Referncias bibliogrficas:
Burchell, S., Clubb, C., Hopwood, A., Hughes, J., & Nahapiet, J. (1980). The roles of accounting in organizations and society. Accounting, Organizations and Society, 5(1), 527. doi:10.1016/0361-3682(80)90017-3
Crotty, M. (1998). The Foundations of Social Research. Sidney: Allen & Unwin.
Gall, M. D., Borg, W. R., & Gall, J. P. (1996). Educational Research: An Introduction (6th ed.). New York: Longman.
Jrgensen, M. W., & Phillips, L. J. (2002). Discourse Analysis as Theory and Method. London: Sage.
Smith, M. (2003). Research Methods in Accounting. London: Sage.
Tinker, A. M., Merino, B. D., & Neimark, M. D. (1982). The Normative Origins of Positive Theories: Ideology and Accounting Thought. Accounting, Organizations and Society, 7(2), 167200. doi:10.1016/0361-3682(82)90019-8
|
