Modelling Vehicular Noise Pollution Data in Some Parts of Warri, Delta State Using Multivariate and Confirmatory Factor Analysis

Abstract

Noise is inevitable. We come across it in our daily lives while
driving, working and many routine activities. Many interact daily
with machinery, either during working activities, or domestic
chores, which produce noise levels large enough to be sources of
environmental concern. The focus of the research is to study the
issue of noise pollution in some parts of Warri using exploratory
and confirmatory factor. In carrying out the noise level
measurements, 10 locations comprising of commercial, industrial
activities and busy roundabouts were selected. The measurement of
sound level was carried out using a type 1 integrated sound level
meter accompanied with a Garmin Oregon 650t hand-held GPS.
The CR811C noise level meter was held at a height of 1.2m above
ground level with the antenna pointing to the sound source. The
measurement process was carried out for the 10 locations at two
times a day which are: 7.00am – 9.00am and 5.00pm – 7.00pm. The
instrument was set at the A-weighting network and the equivalent
noise level (Leq) which is the constant noise level that expands the
same amount of energy over the same period, was measured for the
various locations. Noise measurements was done for ten (10) weeks
(70days) between march to May 2021 for each of the 10 locations
at mornings and evenings and the weekly average noise level in
(dBA) was recorded and employed for further analysis. Results of
the preliminary analysis of the data revealed that; though the noise
level data are significantly homogeneous and devoid of possible
outliers, they are not normally distributed owing to their stochastic
nature. Multivariate analysis of the data revealed that; the
calculated partial Eta squared of the Pillai’s trace is 0.379 which
indicates 37.90% variability among the dependent variables
occasioned by change in the period of measurement. In addition, a
Goodness of Fit Index (GFI) of 0.945, Normal Fit Index (NFI) of
0.900, Relative Fit Index (RFI) of 0.767, Comparative Fit Index
(CFI) of 0.941 and Tucker-Lewis Index (TLI) of 0.863 obtained
from the structural equation modelling (SEM) revealed that the
confirmatory factor analysis showed an acceptable overall model
fit and hence, the theorized model fit well with the observed data
and the hypothesized factor CFA model fits the sample data and
thus the null hypothesis was accepted and it was concluded that the
difference in the observable noise level data is significant.