Chapter
Three (Research Methodology)
By Kibash
This chapter is aimed
at explaining the methodologies to be adopted in collecting, analyzing,
interpreting and drawing conclusions from data used in the research or study.
The chapter basically comprises of; the nature and sources of data, population
and sample, method of analysis, model specification and sometimes the
theoretical framework on which the model is specified.
The major factors that
determine the content and nature of this chapter are:
i.
The research objectives
ii.
Unit and time dimension of research or
scope of the research
iii.
The level of measurement of data
(particularly the dependent variable)
Objectives
Techniques vary with
different objectives of the study, below are different objectives and various
techniques used in achieving them:
i.
Understand
Pattern/ Analyze trend: For this kind of objectives,
descriptive or summary statistics is recommended. This involves the use of
descriptive statistics tools like graphs, tables, natural text contextual
study, etc.
ii.
Measuring
Impact/ Effect/ Causality: Regression analysis is the most
recommended method of analysis in achieving any objectives relating to this. It
could be simple or multiple regression. This requires specification of a
regression model.
iii.
Examining
Relationship: to examine just relationship,
correlation or chi-square is considered to be okay. Correlation is used for
continuous variables while chi-square is used to examine the relationship
between discrete variables. To examine Long-run relationship, Cointegration
test is recommended.
iv.
Measuring
Difference/ Relativeness: T-test or Z-test is usually
adopted in achieving this objective. T-test is used when number of observations
is 30 and below. While Z-test is used when sample size is above 30.
v.
Evaluating:
To evaluate, regression method, T-test or Chow test may be used.
vi.
Measuring
Efficiency/ Effectiveness: Mathematical programming is
recommended.
Unit
and Time Dimension: When data is to be collected on a unit
over several time periods, such data is time series in nature while data on
many units at a particular point in time is cross-sectional data. Panel or
pooled data is data collected on different units collected over time.
Level
of Measurement: This has to do with the form of measurement
the data comes with. It could be in ratio, nominal, ordinal or interval scale.
Nature
and Sources of Data: There are two sources of data; primary
and secondary. Data collected by the researcher through on field through the
use of questionnaire, survey or interview are said to be sourced primarily. On the
other hand, secondary sources are database of organizations or government where
ready-made data are in stock. Examples of secondary sources are; CBN database,
NBS, World Bank data bank, IMF database, IFO, FAO, etc.
Determining Sources of Data: i. Time series data/ Panel
data: Secondary source
ii.
Cross-sectional data: Secondary/ Primary source
Structure of this Chapter
This
depends basically on the objective of the study, source and nature of data.
For
cross-sectional data sourced primarily, the common structure is:
Method of Analysis;
Nature and Source of Data; Population; Sample Size; Sampling Technique; Research
Instrument/ Design; Model Specification (If regression/ Correlation); Variables
in Model; Method of Estimation and Method of Evaluation
For
Time Series or Pooled data from
secondary source, the common structure is:
Method of Analysis;
Nature and Sources of Data; Descriptive Analysis (if necessary); Regression
Analysis, Model Specification, Variables in the Model; Method of Estimation;
Method of Evaluation; Correlation and Cointegration test (If necessary).
Source:
Author’s review of Dr. Kilishi
Presentation at project workshop organized by NESA UNILORIN chapter on March,
2016.
For
further assistance on writing project and data analysis, contact Kibash on 07068769014 (Whatsapp), 08174327803
(Call) or kbashib@gmail.com
