• No results found

5. Chapter Five: Research Methodology and Data Collection

5.4 Research Methodology

In any framework of research procedure, the researchers should reveal their viewpoint, the goals and methods of their research, how they collect and analyse the data and display the results. Therefore, the researchers have to consider the procedures that they need to follow by identifying the models that they decide to use, which in turn allows them to identify and determine the path of the research. The idea of using two approaches, quantitative and qualitative, is a controversial issue.

The previous arguments on paradigm incommensurability notwithstanding, we decided on a mixed methodology study for this research, using both the qualitative method and the quantitative methods, as we wished to investigate not only the relationships (between internal control, corporate governance mechanisms and earnings management in Jordan), but all the mechanisms through which these processes occurred and this relationship was created the issues of paradigm incommensurability, the nature of the assumptions behind the use of different methods and the approach to the use of theory were all reflected upon in the analysis of our results and in the conclusions drawn from them.

Following an analysis of mixed methodology by Tashakkori and Teddlie (1998) a number of

researchers in accounting have chosen to use such an approach Brown and Brignall (2007)

justify their use of mixed methodology by proposing the use of Critical Realism as a

133

considered these arguments, we constructed a research design to investigate the relationship

between corporate governance mechanisms and earnings management in Jordan, which has

four stages:

First stage

The researcher use three earnings management models: the Jones Model (Jones 1991); the Modified Jones Model (Dechow et al, 1995), and the Margin Model (Peasnell et al, 2000) to estimate earnings management values for Jordanian firms. These models used financial data extracted from Jordanian industrial firms’ annual reports, which are available from DATASTREAM, the Amman Stock Market (ASM), and Jordan Securities Commission (JSC). Each of these models uses a different way to calculate the accruals, which are detailed in chapter three.

Second stage

The researcher use published annual financial reports from the Amman Stock Exchange

website to collect information on internal control and corporate governance, such as: board of

directors’ characteristics (e.g. board size, board meetings, board independence, board

outsiders); audit committee characteristics (e.g. audit committee size, audit committee

meetings, audit committee independence and audit committee expertise); external audit

factors (e.g. external auditor reputation, external auditor tenure, external auditor opinion and

external auditor change), and ownership structure (e.g. managerial ownership, family

ownership, blockholder ownership, individual ownership, institutional ownership and state

ownership). These variables, as suggested by previous studies, examine the relationship

between internal control and corporate governance mechanisms and the earnings

management estimates from the accruals models, by using quantitative method.

Third stage

On the basis of the results from the quantitative analysis and knowledge of the firms’

134

These semi-structured interviews, conducted with general managers, financial managers and

internal auditors. To complete the qualitative analysis, the data from these interviews are

combined with the models of accruals, analysis of the firms’ published financial statements,

review of published Amman stock market information, reviews of corporate governance

statements in each firm, and reviews of internal documentation.

Fourth stage

At this stage, the researcher considers the extent to which the results of the quantitative and qualitative analysis could be combined in order to produce a contribution to knowledge in the area that reflected both different approaches and types of results. In general, when using dual methodologies, thought needs to be given to which methodology is going to be allowed to predominate. The two can be combined by: (1) applying the qualitative method before the quantitative method to enable the researcher to develop and present a suitable conceptual framework, hypothesis, or necessary tools for the quantitative study, or (2) applying the qualitative method after the quantitative method to clear up any confusion in the quantitative findings and to support an interpretation, which means that merging the results of the quantitative and qualitative methods will provide a more exact overall picture of the research problem (Srnka and Koeszegi, 2007).

Existing literature suggests three typical reasons for using mixed methods (Caracelli and

Greene 1997), and these are: (1) testing the agreement of findings obtained from different measuring instruments, (2) clarifying and building on the results of one method with another method, and (3) demonstrating how the results from one method can impact subsequent methods or inferences drawn from the results.

Steckler, et al (1992) discuss how qualitative and quantitative methods can be integrated and

present the results. Their approach suggests that while the qualitative and quantitative

paradigms separately have some weaknesses, which can be attenuated by combining the two.

Furthermore, they indicate that the quantitative method produces accurate, consistent results

135

method generates rich, in-depth, suitable results that usually support the more complex issues

that the researchers cannot measure in numeric ways (i.e. using quantitative methods).

Accordingly, they provided four models, as shown in below Figure (5.2) that explain possible

integrated relationships between quantitative and qualitative methods.

Figure 5-3 Integrated Quantitative and Qualitative Method

Brown et al (2007) identify three ways in which the choices made during the process of designing a dual-methodology research project can affect the way in which the results can be combined. They suggest that the results could complement each other, challenge each other, or talk past each other.

Model (1): Qualitative methods are used to help developing quantitative measures and instruments.

Model (2): Qualitativemethods are used to help explain quantitative findings.

Model (3): Quantitativemethods are used to establish a primarily qualitative study.

Model (4): Quantitative and Qualitativemethods are used equally and parallel.

Source: Steckler et al (1992)

Qualitative Quantitative Result

Quantitative Result Qualitative Qualitative Quantitative Result

Qualitative Result Qualitative

136

5.4.1 Quantitative and Qualitative Methods Complementing Each Other

This section discusses the quantitative and qualitative methods as complementary approaches. Several researchers in prior literature have argued that if quantitative and qualitative analyses could be designed to complement each other, this will lead to producing a unified body of knowledge, e.g. Brown et al (2007), and this in turn will lead to a better fit and understanding between research objectives and their problems (e.g. Brown et al 2007; Burgess 2002). Furthermore, Johnson, Onwuegbuzie and Turner (2007) state that different methods can be based on different strategies and methods of collecting data, in ways that reflect complementary strengths and non-overlapping weaknesses.

Cupchik (2001) suggests that the notion of a qualitative-quantitative sequence could be replaced altogether by using a complementary approach. Often quantitative research yields statistically significant results, but not a full explanation of the underlying relationship powerful enough for the results to be generalised. Therefore, if the research is designed well enough, qualitative research can provide a rich source of data, that can clarify the unclear results produced by the quantitative methods and uncover the process by which the results are produced. Therefore, if qualitative and quantitative methods are deemed to be complementary, this means that the richness of information from both methods can enhance the quality of research results, since the in-depth information from the qualitative method will support the results of the quantitative method. Also, the large amounts of descriptive data produced by qualitative research can shape the choice of variables in quantitative research and provide a contextual explanation for the quantitative findings.

Existing studies have presented several advantages of using quantitative and qualitative together as complementary approaches, and Philip (1998) summarises these advantages in three points:

• The use of more than one method of gathering evidence helps to minimise the risk of generating erroneous findings.

• Qualitative research may be carried out in order to establish research questions which will subsequently be addressed by means of quantitative methods; or a quantitative study may suggest questions suitable for further investigation by means of in-depth qualitative methods.

137

• The combination of methods may allow a broader range of issues to be addressed during the course of conducting a research project than would be possible if exclusive use of either quantitative or qualitative methods had been made.

5.4.2 Quantitative and Qualitative Methods Challenging Each Other

Despite the existence of many similarities between qualitative and quantitative research methods, there are some measures and procedures that are extremely different under these two methods, due to both their natures and assumptions regarding the data and questions to be answered, which can lead to a challenge between these methods’ results (Malterud 2001). Johnson et al (2007) argue that the adoption of mixed methods is an unsuitable way to present results, since the quantitative and qualitative methods study different aspect of the phenomenon under investigation.

From a research point of view, prior studies have argued that a combination of the results of quantitative and qualitative methods might lead the researchers into reaching the wrong conclusions. In this regard, Brown and Brignall (2007) state that when quantitative and qualitative results contradict each other: “the combination of the results of quantitative and qualitative methods will be, at best, misguided and, at worst, misleading.”

5.4.3 Quantitative and Qualitative Methods “Talk Past Each Other”

In this context, “talk past each other” means that the researcher thinks that since each of the two methods are studying the same subject, they will support each other, but in fact while each of them produces results that are credible, they are so different in nature that they cannot be combined in any reasonable way.

Barkin (2010) documents that there are two reasons which can lead to “talking past each other”. The first reason concerns the nature of the paradigmatic debate, which results in different researchers approaching their research with different mindsets and thus conducting their studies in different ways. The second reason Barkin (2010) describes as “simple terminological confusion, which results from different interpretations of important words, caused partly because scholars redefine terms and different authors use them in different ways.

In the context of a research in Anthropology Metge (1978) points out that “talking past each other” is often based on the anthropological definition of culture, which is familiar to some researchers but new or strange for others. Accordingly, “talking past each other” is one of the

138

challenges of mixed methodology research approaches, which derives from paradigm incommensurability.