METHODOLOGY
Corporate Strategy, Strategic Leadership, and Organisational Performance
Name
Liberty University - Lynchburg, VA
BUSI 101 - Introduction to Business
2022
1
RESEARCH METHODOLOGY
3.1 Introduction
This chapter describes the research methodology that will be in this study. It focuses on the
research philosophy, research design, population to be used in the study, data collection, the
operationalization of the key study variables and data analysis.
3.2 Research Philosophy
Social science inquiries have been generally guided by two broad research paradigm namely
phenomenology paradigms (Saunders, et al. 2007). Phenomenologist’s focuses on the
immediate experience where the researcher draws meaning by interpreting that which is
observed during his/her involvement in the phenomena (Blau, 1964). Phenomenological
research enables the researcher to acquire knowledge on the situation under investigation.
Phenomenon observations such as case studies provide qualitative data that illustrates and
explores the phenomenon in-depth thus providing more solid results (Zikmud, 2003).
Positivism paradigm also known as an acceptable knowledge of applying the methods of
natural sciences will be used by the researcher (Saunder et al, 2015). This study was guided
by a positivist paradigm approach because it sought to test various theories based on real
facts, neutrality and objectivity of the research. Positivism assumes that social reality is made
up of objective factors that can be precisely measured and statistics used to test causal theories
relating thereto. It holds a deterministic philosophy which pre-supposes that causes determine
effects or outcomes (Creswell, 2012).
3.3 Research Design
This study will use a cross-sectional descriptive survey research design. This is considered
appropriate since the concepts will be measured as they naturally exist without being
manipulated or controlled. This is given by the philosophical paradigm adopted for the
research as it was concerned with investigations in what, when and how much of phenomena
at one point in time (Cooper & Schindler, 2011).
The primary objective of this study is to establish the influence of organizational structure and
strategic leadership on the relationship between corporate strategy and performance of Kenya
state corporations. A cross-sectional descriptive survey offers the opportunity to collect the
data across different firms and test their relationships as per the research objectives and
2
hypothesis. It affords the researcher the opportunity to capture a population’s characteristics
and test the hypotheses quantitatively concerning the period over which data were collected
across various firms. The cross-sectional survey is appropriate for obtaining data at one point
in time. Cross-sectional research approaches are best used when obtained information
characterizes the events in a firm at an actual point in time, as in the case of this study,
(Bryman, 2004). Besides this design is deemed most fitting because of the scope of the
review, the nature of the data to be collected and the method of analysis to be used on the
obtained information.
3.4 Population of the Study
The population of the study will consist of all the Kenya state corporations as of 10th August
2018. The corporations are classified into: purely commercial state corporations (47); state
corporations with strategic functions (11); state agencies (62); independent regulatory
agencies (25) and state agencies –research institutions, public universities, tertiary education
and training institutions (44) total 189, (National Treasury, 2018). A complete list of all state
corporations in Kenya is attached on appendix 2.
Table 3.1: Distribution of State Corporations by Function
Classification of Kenya state corporations Population
Purely commercial state corporations 47
State corporations with strategic functions 11
Executive agencies 62
Independent regulatory agencies 25
Research institutions, public universities, tertiary education, and
training institutions
44
Total 189
3.5 Sample Design
The study will use the stratified random sampling technique. The classification of Kenya state
corporations by function will be used as stratum. In a cross-sectional descriptive survey
research design, a sample size of 50% is acceptable. The researcher applied the sampling error
3
formula (Creswell, 2011 pp. 609-612) to get a sample size of 97 State Corporations
respondents.
Table 3.2 Sample Size
Classification of Kenya state corporations (stratum) Population 50% Sample
Size
Purely commercial state corporations (Appendix 3, Table on
sample selection)
47 24 24
State corporations with strategic functions (Appendix 3, Table
on sample selection)
11 6 6
Executive agencies (Appendix 3, Table on sample selection) 62 32.8 33
Independent regulatory agencies (Appendix
3, Table on sample
selection)
25 12.1 12
Research institutions, public universities, tertiary education, and
training institutions (Appendix 3, Table on sample selection)
44 22 22
Total 189 97
Purely commercial state corporations, State corporations with strategic functions, Executive
agencies and Independent regulatory agencies were selected based on proximity and travel
time during data collection as shown in Appendix 3, Table on sample selection.
All five (5) research institutions will be selected for the study. Using the formula, N=n/P
(Tusell, 2012).
Where n is the strata while P represents the population, N=8. Therefore, 8 Public Universities
will be selected. The researcher will extrapolate the findings to describe the link between
corporate strategies and Performance. All the five (5) tertiary education providers were
selected. Finally, all the four (4) training institutions were selected. The sample size for
Research institutions, public universities, tertiary education, and training institutions was 22
as indicated in table 3.2.
4
3.6 Data Collection
The study will use both primary and secondary data. Primary data, typically quantitative, will
be on corporate strategy, organizational structure, strategic leadership, and firm performance.
Primary data will be collected using a questionnaire shown in Appendix 1. The questionnaire
is preferred because of a large number of respondents. Questionnaires will also provide
information which the respondent would not be comfortable providing in face-to-face
interviews. The questionnaire will be composed of both open and close-ended questions. Part
A will seek demographic information of the respondent. Part B, C, D, and E will explore
Corporate Strategy, organizational structure, strategic leadership, and organizational
performance respectively. Secondary data will be collected using secondary data collection
form and shall be obtained from the relevant documents of the firms on their performance
from 2011 to 2015. A form has also been prepared where companies unable to provide their
records can still provide the required secondary data by one senior manager from each
department filling the respective section in the form.
The respondents will consist of four (4) members of the management team per state
cooperation, specifically the chief executive officer, the human resources' manager, the
operations manager, and the finance manager. The respondents will be given time to fill out
the questionnaire. The questionnaire is based on a five-point Likert scale ranging from 1 (one)
represents a substantial extent and 5 (Five) strong extent. Likert type scale allows the
researcher to analyze data using inferential statistics.
Pretesting the questionnaire is vital to ascertain unclear questions since the respondents may
comprehend them in a differently thus allowing for improving the questionnaire. The
questionnaires will be distributed and collected using trained research assistants. The research
will request for information on whether the respondents had access to the internet. An internet
link pointing to the online survey administration tool known as survey monkey will be
provided for respondents who have access to the internet. A letter of introduction from the
university will be used to reassure contribution and as evidence of the credibiity of the survey.
Secondary data relating to financial performance for a five year period of 2010/2011 to
2014/2015 will be collected from firm’s records, industry records and Ministry of
Industrialization, Enterprise and Co-operatives Development. This departments are tasked
with evaluating performance and ranking performance of Kenyan State Corporations.
5
3.6 Operationalization of the Key Study Variables
The key variables in this study are a corporate strategy which is the independent variable of
the research and organizational performance which is the dependent variable. Organizational
structure and strategic leadership are the moderating variables which affect the association
between the independent and dependent variables as noted earlier. Table 3.2 below shows the
operationalization of stated variables.
6
Table 3.3: Operationalization of the Key Study Variables
Variable Indicator Measure of Indicator Questionnair
e Items
Supporting
Literature
Corporate
strategy
(Independen
t Variable)
-Growth
-Stability
-Retrenchment
5- point Likert Type
Scale
Section B of
the
questionnaire
in Appendix 1
Haythem (2015),
(Kyereboah and
Biekpe (2006),
Herath (2007), Lear,
(2009).
Organizationa
l structure
(Moderating
Variable)
-Complexity
-Formalization
-Standardization
-Centralisation
-
5-Point Likert Type
Scale
Section C of
the
questionnaire
in Appendix 1
Collins (2015)
Yazdani (2009),
Herath (2007),
Serfontein (2010),
Robbin and DeCenzo
(2005)
Strategic
Leadership
(Moderating
Variable)
-Shared vision
-Long term
orientation
-Concern with making
change that make a
difference
-Being innovative
5- Point Likert Type
Scale
Section D of
the
questionnaire
in Appendix 1
Nielsen (2010),
Mallin (2010), Rau
(2008), Wendy (2012)
Organizationa
l Performance
(Dependent
Variable)
-Financial and non-
perspective
-Customer perspective
-Internal process
perspective
- Learning and growth
perspective
5- Point Likert Type
Scale
Section E of
the
questionnaire
in Appendix 1
Richard et al. (2009),
Daft (2000),
Barney (1991),
Hefferman and Flood
(2000)
Kaplan and Norton
(1996)
Huang and Li (2009)
7
8
(Source: Researcher, 2018)
3.7 Test of Reliability
Reliability shows the extent to which the research is short of bias, guarantees uniformity in
measurement across time and several items contained in the research instrument (Sekaran,
2010). Reliability is a gauge of the stability and consistency within which the instrument
analyses the models and aids in evaluating the authenticity of the measure. The safety of the
research tool will be assessed using Cronbach’s α (alpha). The Cronbach coefficient will then
be applied in measuring the common association of objects.
The alpha coefficient values range from 0 to 1 strong coefficient suggests that the items or
objects relate to themselves, that is, there is consistency between items in measuring the idea
of interest (Mugenda & Mugenda, 2003). A Cronbach's Alpha Coefficient value of below 0.5
is estimated weak while a value above 0.5 is substantial. Nunnally, (1978) recommends
Cronbach’s Alpha Coefficient of 0.7 or above as the most appropriate. This study will assume
a coefficient of and above 0.7 to infer internal consistency reliability.
3.8 Test of Validity
Validity is the point at which the outcomes obtained from the scrutiny of the data collected
epitomizes the phenomena under study (Mugenda & Mugenda, 2003). It defines whether the
research instrument accurately examines what it is anticipated to be measured with accuracy
(Barbour, 1998). There are four forms of validity, namely; face validity, content validity,
criterion validity and construct validity (Gomez-Haro et al. 2011). Validity test will be
established by administering the questionnaire as a pilot test before data collection for the
entire study. Content validity will be ensured by the questionnaire getting tested by subjecting
it to double checks. This ensures that the questionnaire covers all the areas of study which
includes; strategic planning, organizational structure, strategic leadership, and firm
performance.
Content validity shall be tested by use of experts and supervisors in the relevant area. They
will look at the research questions and objectives to see whether the questions in the
questionnaires achieve the objectives or answer the research questions.
9
3.9 Data Analysis
After completing the field survey, the collected data shall be edited for accuracy, uniformity,
consistency, and completeness, organized, summarized and then arranged to enable coding
and tabulation before final analysis. The data will be transferred from the questionnaires into
the worksheet as a database file. The variable names within the database file will refer to the
numbers of each question in the questionnaire. The data will be divided into several sub-
topics by the structure of the questionnaire.
Descriptive statistics such as mean scores, standard deviations will be used to summarize both
the secondary and primary data to enable meaningful description. The use of inferential
statistics shall be used to help answer the objectives and research questions and find out if the
research can be generalized from the sample to the population.
3.9.1 Diagnostic tests
The diagnostic tests used in the study will be Statistical tests which rely upon certain
assumptions about variables used in the analysis. When the assumptions are not met, the
result may not be trustworthy resulting in either Type I or Type II error or over or
underestimation of significance or effect sizes. Normality will be undertaken through the use
of histograms and probability-probability (p-p). Non-normality distributed variables can
distort relationships and significance tests (Osborne & Waters, 2002). Visual inspection of
data plots will cater this, skew, and kurtosis. Shapiro-Wilk test will be done, and the results
plotted in the Q-Q plot to establish the normality of the data and statistical errors that need to
be checked. The Data is assumed to be normal when the histogram appears symmetrical, bell-
shaped, curved, with the highest frequency of scores in the middle and smaller frequencies to
the extremes. Data that exhibits non-normality characteristics may lead to inaccuracy of the
results.
Collinearity or multicollinearity refers to the assumption that the independent variables are
interrelated (Keith, 2006). The Variance Inflation Factor (VIF) will also be applied to asses’
multicollinearity. The VIF values should not exceed 10, and the tolerance values should not
be less than 0.10 (Keith, 2006). The homogeneity variance of the study variables will be
tested using Levine tests. If the Levine value is greater than 0.05, then the variability of
conditions is determined to be the same. Statistical procedures to be applied using correlation,
10
regression, t-tests and analysis of variance are based on the assumption that data follows a
normal distribution. The coefficient of variations, skewness, and kurtosis will be computed to
describe the characteristics of the variables of interest. The study will also test the data using
multicollinearity tests, normality tests and homogeneity of variance tests.
Table 3.4: Analytical Model for Corresponding Objectives and Hypotheses
The table below provides the research objectives with their respective hypotheses; analytical
technique uses and the final interpretation of the expected results.
Research
objective (s)
Hypotheses Analytical Technique Interpretation
Determine the
effect of
corporate
strategy on
the
performance
of Kenya
state
corporations
H01:
Corporate
strategy has a
significant
influence on
the
performance
of Kenya State
Corporations
Simple linear regression analysis:
Y1= βo1+1X1 +1
Where: βo = intercept
Y= Firm Performance
1, represents beta coefficients
for H01
X1 represents Corporate Strategy
is the error term
Βeta value – The effect of
independent variable
on the dependent
variable (Positive or
Negative)
t- test – Significance of
individual variables.
P-value –A smallPp-
valueP(typically ≤ 0.05)
indicates strong
evidence against the null
hypothesis, reject the
null hypothesis. A
largePp-valueP(> 0.05)
indicates weak evidence
against the null
hypothesis; fail to reject
the null hypothesis.
Establish the
effect of
organizational
structure on
the
relationship
between
corporate
strategy and
H02 l Structure
has a
significant
moderating
influence on
the
relationship
between
corporate
Performance = F (Corporate
Strategy + organizational
structure + (Corporate Strategy*
organizational structure)
Y2a = β2ao+ β2a1 X1+ β2a2 X2 + β2a3
X3+ β2a4 X4+ β2a5 X5+ β2a6 X6+ β2a7
X7+2a
Y2b = β2bo+ β2b1 X+ β2b2 Z+ β2b3
X*Z+2b
Βeta value – The effect of
independent variable
on the dependent
variable (Positive or
Negative)
t- test – Significance of
individual variables.
11
performance
of Kenya
state
corporations
strategy and
performance
of Kenya State
Corporations
Where, X=Composite Corporate
Strategy; Z =Composite
Organizational structure;
P-value –A smallPp-
valueP(typically ≤ 0.05)
indicates strong
evidence against the null
hypothesis, reject the
null hypothesis. A
largePp-valueP(> 0.05)
indicates weak evidence
against the null
hypothesis; fail to reject
the null hypothesis.
Establish the
effect of
strategic
leadership on
the
relationship
between
corporate
strategy and
performance
of Kenya
state
corporations
H03:
Strategic
leadership has
a significant
moderating
influence on
the
relationship
between
corporate
strategy and
performance
of Kenya State
Corporations
Multiple Regression analysis
Hierarchical Regression analysis,
Path analysis used to test and
determine the intervening effect.
Y3a = β3ao+ β3a1 X1+ β3a2 X2 + β3a3
X3+ β3a4 X4+ β3a5 X5+ β3a6 X6+ β3a7
X7+2a
Y3b = β3bo+ β3b1 X+ β3b2 Z+ β3b3
X*W+2b
Where, X=Composite Corporate
Strategy; W =Composite
Strategic leadership;
Βeta value – The effect of
independent variable
on the dependent
variable (Positive or
Negative)
t- test – Significance of
individual variables.
P-value –A smallPp-
valueP(typically ≤ 0.05)
indicates strong
evidence against the null
hypothesis, reject the
null hypothesis. A
largePp-valueP(> 0.05)
indicates weak evidence
against the null
hypothesis; fail to reject
the null hypothesis.
Determine the
joint effect of
corporate
strategy,
organizational
structure and
H4: There is a
significant
joint effect of
corporate
strategy,
organizational
structure and
strategic
leadership on
Multiple Regression analysis
Y4a = β4ao+ β4a1 X1+ β4a2 X2 + β4a3
X3+ β4a4 X4+ β4a5 X5+ β4a6X6+ β4a7
X7+ β4a8 X8+ β4a9 X9+ β4a10 X10+
β4a11 X11+4a
Y4b = β4bo+ β4b1 X+ β4b2 Z+ β4b3
W+ β4b4 X*Z+ β4b5 X*W+ 4b
Βeta value – The effect of
independent variable
on the dependent
variable (Positive or
Negative)
t- test – Significance of
individual variables.
12
strategic
leadership on
the
performance
of Kenya
State
Corporations
the
performance
of Kenya State
Corporations
Where, X=Composite Corporate
Strategy; W =Composite
Strategic leadership;
P-value –A smallPp-
valueP(typically ≤ 0.05)
indicates strong
evidence against the null
hypothesis, reject the
null hypothesis. A
largePp-valueP(> 0.05)
indicates weak evidence
against the null
hypothesis; fail to reject
the null hypothesis.
Using stepwise regression, the variables (or a subset of variables) that best predict the
performance of Kenya State Corporations will be analyzed. Thus, the dependent variable will be the
performance of Kenya State Corporations, and Corporate Strategy, Organizational Structure, and
Strategic Leadership will be the independent or predictor variables.
13
Table 3.4 Continued…
