Summary, Conclusion and Recommendations
There is no doubt that this paper has been able to adequately address
the objectives it set out to achieve. It started by introducing the
subject of study and identifying the problem it set out to tackle. While
reviewing extant literature frameworks on the subject of multivariate
dependent variables, the paper introduced the various methods and
formula adaptations that can be employed to unify financial performance
data for the purpose of relational research analysis. From the
literature review and the earlier discussions of the results of research
data analysis, it is evident that financial performance research
utilizing multivariate dependent variable has very little space to
maneuver with the existing econometric methods of multivariate data
analysis. Financial and accounting information are made up of already
processed and validated economic data and values which require careful
and exact practical treatment which the abstract assumptions of most
econometric measurements and analysis may not adequately handle,
resulting in the need to adopt a more practical and less abstraction
assumptions and methods to manipulate them. The use of geometric means
to unify multivariate dependent variables into a single variable is
considered the best way to effectively execute a financial performance
relational research because, according to Transaction Processing
Performance Council (2011), the geometric mean presents a single figure
of merit which respects the intrinsic differences across all dimensions
of the data distribution (UNDP, 2011).
In view of the aforementioned findings, this paper recommends that more
attention be paid to the peculiar nature of financial performance
research and the important role that such researches ought to play in
the economy of every nation when deciding on the method and tool of
analysis to adopt. It will not be out of place if we have a statistical
analysis type that takes the peculiar interest of accounting and finance
into consideration. If chemistry can develop a special branch of
statistics called chemometrics , it is possible to have eitherfinometrics or accometrics .