| dc.description.abstract | The main objective of the study was to determine the effect of the internal control system
on fraud prevention in the Banking Sector in Kenya. Studies have been carried out by
different researchers some focused-on fraud in commercial banks and others on internal
controls in government departments. This study filled the gap by focusing on the effect of
internal control system on fraud prevention in banking sector. Research was based on
control environment, risk assessment, communication of information, and monitoring as
independent variables, compliance with prudential regulations as moderating variable and
fraud prevention as dependent variable. Theories that guided the study include; agency,
fraud management lifecycle, fraud triangle and fraud diamond theories. The study
captured various empirical studies to provide more information on the study on this
research area. Pragmatism philosophy, the research designs applied were descriptive and
correlational. The study utilized mean and standard deviations for descriptive research
design. The study involved all banks registered and operating in Kenya and the
respondents were all branch managers, operations managers and cash managers or
supervisors in all headquarters or offices of these banks. Questionnaire was used for
primary data and secondary data schedule for secondary data correction. Coding of data
was undertaken for processing and analysis. Document analysis was used to analyse
qualitative data. ANOVA test was applied to test the overall significance of the model, to
test multicollinearity to determine whether there is inter-correlation among independent
variables, the variance inflation factor was used, to test autocorrelation and
heteroscedasticity, Durbin-Watson test and Breusch-Pagan test respectively were applied.
Exploratory factor analysis was applied to extract factors where principal component
analysis and varimax rotation methods were applied where seven factors were extracted.
Confirmatory factor analysis was used to determine construct bias and validity and was
concluded that there was no bias and the research instruments were valid. Further
analysis on determination of model fit involved chi-square, comparative fix index, root
mean square for error approximation, pclose and standardized root mean residual and was
found that the model fit well except for SRMR. Karl Pearson’s coefficients of correlation
were used to test the strength and to show the direction of association of variables and the tstatistics
was applied to test the hypotheses. The hypotheses tests were based on
structural equation modelling and it was found that risk assessment, communication and
monitoring had significant effect on fraud prevention whereas control environment and
control activities had insignificant effect. Further it was found that compliance with
prudential regulations has no significant moderating effect on the relationship between
control environment and fraud prevention and risk assessment and fraud prevention. The
study applied Beneish model and probit regression in analysis of secondary data and it
was found that financial statement manipulation is rife in seven banks thus those banks
have weak ICS and are not capable of preventing fraud. Recommendations from study
are that new ways of strengthening ICS should be established and that the government
should review regulations. Further research should be carried out with cross sectional
research design that will include questionnaire, interview and secondary data schedule. | en_US |
