Chapter three

Research Methodology

Introduction

This chapter outlines the methodology which the present study took. Outlined here are the research
design, population and sample, data collection, and data analysis.

Research Design

(Rajasekar, 2016) provided a definition of research design as a conceptual framework guiding the
conduct of a research endeavor. It entails formulating clear research objectives, establishing methods
for data collection, and selecting the population and samples to be studied. Similarly, (Almalki, 2016)
characterized research design as a process facilitating the planning of information collection and analysis
in terms of how, when, and where it occurs. Within this framework, descriptive and explanatory
research designs are prominent types (Almalki, 2016).

This study will utilize a descriptive research design to offer a comprehensive overview of the impact of
cost control on profit maximization. As stated by Churchill (2021), this approach is suitable for describing
the characteristics of specific groups, estimating the prevalence of certain attributes, and making
predictions. The descriptive methodology facilitates the acquisition of information to better understand
the challenges facing manufacturing companies and devise effective strategies to enhance profitability.
Furthermore, the study's use of interviews and questionnaires falls under the umbrella of the descriptive
research design, enabling the collection of data to assess the prevailing circumstances during the
research period. The study intends to gather data from manufacturing companies at a single time point
to evaluate the influence of management accounting practices on the financial performance of
manufacturing firms in Kenya.

Population

In statistics, the target population refers to the specific group of individuals or entities for which
information is sought. As outlined by Ngechu (2018), a population is a clearly defined set of people,
services, elements, events, groups, or households under investigation. This definition underscores the
importance of homogeneity within the population of interest. Population studies, also known as
censuses, are deemed more representative because every individual has an equal opportunity to be
included in the final sample drawn, as described by Mugenda and Mugenda (2023).

Sample design
A portion of the target population, known as the accessible population, is systematically chosen to serve
as a representative sample. While the total number of individuals in the population is known,
determining the sample size is unnecessary because the accessible population can be interviewed
comprehensively, as stated by Alvi (2016).
In this study, the target population consisted of 55 selected cost accountants at Bamburi Cement, with a
sample size of 48 respondents. The sample size was calculated using Slovene's formula, which is:

Where:

N = Total Population

n = Sample size

e = is the confidence level at 0.05

Substituting into the formula,

Data Collection

The research gathered primary data from participants, encompassing both quantitative and qualitative
data types. Qualitative data involves non-numerical descriptions, while quantitative data consists of
numerical measurements. Data collection was facilitated through a questionnaire, structured into two
sections. The initial section focused on demographic and operational traits, aiming to ascertain basic
information such as respondent demographics. The subsequent section delved into inquiries regarding
the impact of management accounting practices on the financial performance of manufacturing firms in
Kenya. Additionally, secondary data was acquired from the published annual reports of Bamburi Cement
spanning five years (2019-2023).

Data Analysis

Data analysis involves organizing, structuring, and deriving meaning from the extensive information
gathered (Mugenda & Mugenda, 2023). As the study employed a mixed-method approach, combining
both quantitative and qualitative data (as defined in the "research design" section), analysis was
conducted using the Statistical Package for the Social Sciences (SPSS) alongside coding and theming
techniques. SPSS facilitated the presentation of information through tables and figures.

Conceptual Model

The conceptual model in this study is specified as follows:

Y=ƒ (CAT)

Where Y is the financial performance;

CAT is the Cost Accounting Techniques which include direct costing, standard costing, marginal costing,
and historical costing.

Empirical model

The research employed a regression model to forecast the impact of the identified independent
variables on the dependent variable. SPSS version 18 was utilized for regression analysis and computing
coefficients. The regression line is depicted by the following model:

FP = X1 + X2 + X3 + X4+ et

Where;

FP = financial performance, ROA=Net Income

Total assets

X1= Direct Costing

X2= Standard Costing

X3= Marginal Costing

X4= Historical Costing

e t = Error term

To investigate the correlation between cost accounting techniques and the financial performance of
Bamburi Cement, regression analysis was conducted. Cost accounting techniques were regressed
against financial performance to identify which practices exert a significant influence. The interpretation
of regression analysis results relied on the R-squared value, the significance of F-statistics, and the
significance of beta values derived from the coefficients of the independent variables (X). Significance
was assessed at the 5% level.

Data Validity and Reliability

Certain constructs in this research conceptual model were established as new scales and drawn from
previous studies. The face and content validity of these constructs were assessed by accounting
academic experts. Confirmatory and exploratory factor analyses were employed to explore the
underlying relationships among numerous items and ascertain whether they could be condensed into a
smaller set of factors. Factor analyses were conducted separately for each group of items representing a
specific scale, a methodology chosen due to the limitations in observations. Factor loading values
exceeding 0.50 are typically deemed essential for practical significance.
References

Rajasekar, S. (2016). Reasearch Methodology. ResearchGate

Almalki, S. (2016). Research Methodology. ResearchGate, pp.55-62.

Alvi, M. (2016). Manual for Selectig Sampaling Techniques in Research. Munich Personal RePEc Archive,
p 10.