RM Dissert

CHAPTER 3: RESEARCH METHODOLOGY
Introduction
This chapter aims at giving an overview of the two methodologies that will be used in this study to
conduct the analysis. In this study, secondary data will be used to gain data and information to
perform both a systematic review and meta-analysis.
Systematic review
Systematic Literature Reviews (SLRs) represent a meticulous and structured method for synthesizing
previous research, capturing, and critically evaluating pertinent papers on a certain issue. Formulated
by Liberati et al. (2009), this approach stands out for its rigorous processes geared towards
minimizing biases, thus providing a more dependable and comprehensive summary than standard
literature reviews (Okoli, 2015). Particularly beneficial in domains like medicine and business, SLRs
facilitate the amalgamation of numerous methodologies and address the challenges posed by the swift
generation and dispersion of knowledge in the digital era. They play a crucial role in pinpointing
research gaps and spotting research gaps and establishing a firm footing for future studies especially
for newcomers to a field. When using SLRs in research, it is crucial to approach them critically,
acknowledge their limitations, and provide a balanced approach.
A systematic review was carried out on rated research papers concerning tax compliance, adhering to
the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020
guidelines. This structured approach involves identifying, selecting, and analyzing data from previous
studies. The databases used included Emerald Insight (Emerald), Science Direct (Elsevier), Wiley
Online Library (Wiley), Google Scholar, and JSTOR. Journals were chosen based on the ABDC
journal list criteria, focusing on English-language articles published between 2014 and 2024, spanning
an 11-year period.
The systematic reviews’ search criteria included keywords, title, abstract and study purposes. Various
search terms such as "Tax compliance " "Tax enforcement," "Tax reporting " "Tax evasion," "Tax
avoidance ", "tax reforms", “tax administration” and “tax fairness” were utilised in databases. The
articles were categorised into 8 focus areas with an analysis of 100 publication journals. A summary
list of the articles used in the review is available in Appendix X detailing the authors, years, and
findings. Following screening 100 research articles met the eligibility criteria for assessment as shown
in Figure 1.
This study uses the PRISMA methodology to analyse the determinants of tax compliance. This
technique provides a thorough and unbiased examination of the current literature in this quickly
evolving field. The PRISMA technique organises and evaluates relevant papers to gain a full grasp of
the issue. Figure 1 shows our methodology for conforming to PRISMA criteria. This chart illustrates
our systematic review approach, from identifying sources to final inclusion of studies.
*TO ADJUST FOLLOWING DIAGRAM*

Meta-Analysis
After conducting a systematic review, the next method used to evaluate the collected data is meta-
analysis. Meta-analysis is a statistical technique that combines and analyzes multiple previous studies
to assess their overall effect (Kristalyn Salters-Pedneault, 2022). Initially introduced and applied in
the fields of medicine and science, one of the pioneers of this method was Karl Pearson in 1904. Since
then, it has been widely used in research across various fields (Lau, J., Rothstein, H.R., and Stewart,
G.B., 2013). This technique is considered less critical and subjective compared to other literature
analysis methods because it focuses on data and research results, examining effect sizes to determine
the overall significance of the articles (King, W.R. and He, J., 2005).
There are two main models for meta-analysis: the fixed effects model and the random effects model.
The fixed effects model assumes that a single true effect size is obtained when studies are combined
and that any observed differences in effect size are due to sampling error (Dettori, J.R., Norvell, D.C.,
and Chapman, J.R., 2022). Conversely, the random effects model, which is more commonly used,
assumes that effect sizes vary across studies due to inherent differences between them (Borenstein, M.
et al., 2010).
The first step in the meta-analysis is to apply the random effects model to synthesize and analyze
systematic reviews of tax compliance. Comprehensive Meta-Analysis V4 software was used to
calculate the effect size of the articles that met the eligibility criteria for quantitative analysis, as
illustrated in Figure 1. Next, publication bias will be assessed using a funnel plot and the classic fail-
safe N to evaluate the effectiveness of the studies and determine whether they are free from bias.
The effect size
Determining the effect size is crucial in the meta-analysis process, as it combines all eligible studies
into a collective mean effect size. This metric measures the correlation between variables or sample
sizes. In this study, the effect size will be assessed using Cohen's d technique, which standardizes the
mean difference of an effect observed between two distinct sample groups. The formula for this
calculation is as follows:
The effect size is calculated by subtracting the mean values of two samples and then dividing this
difference by the pooled standard deviation, which represents the average deviation within the groups.
This calculation is represented by the following formula:
Where ds is the effect size of a sample using Cohen's d, the numerator represents the mean difference
between the two samples, and the denominator is the pooled standard deviation formula. Furthermore,
when interpreting the effect size for different studies, three interpretations are considered: small size,
medium size, and large size (Vaidya, D, 2023).
Table 1 illustrates that an effect size falling within the range of 0.2 to 0.5 indicates a small effect
between the two samples, which may be considered negligible and not readily apparent. A medium
effect size is denoted by values between 0.5 and 0.8, suggesting a potentially significant impact on the
overall outcome. A large effect size indicated a substantial overall impact that can be observed
without detailed analysis.

RM Dissert

Uploaded by

Copyright:

Available Formats

RM Dissert

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

RM Dissert

Uploaded by

Copyright:

Available Formats

CHAPTER 3: RESEARCH METHODOLOGY

TO ADJUST FOLLOWING DIAGRAM

You might also like