Data Screening and Preliminary Analysis on the Impact of Rural Infrastructure on Rice Productivity in Kano State, Nigeria

Yusuf Tanko,Rabiul Islam,Chea Yong Kang

doi:10.31580/jei.v6i2.900

Abstract

Despite having vast farmland suitable for paddy rice farming, local production in the country is weak, especially in Kano where the state has the most abundant farmland put to rice farming and the most extensive rice farmers in the 36 states of the country. As such, over 4mm/t of milled paddy rice has to be imported annually into Nigeria to supplement home production. The economy cannot sustain rice import because it depends on crude oil revenue; thus, leading to scarcity of rice at an exorbitant price. The study was conducted in the 2018 cropping season for rainfed and irrigated paddy rice, to identify the impact of rural infrastructure on the productivity of rice farmers in Kano State, Nigeria. There are seven local governments with 17 rice clusters in the state that are cultivating rice. A random sample of 768 rice farmers was selected in 9 rice clusters from the population of 135,895 rice farmers using multistage and purposive sampling. Using the Statistical Package for Social Science (SPSS) software version 22, data screening and preliminary analysis was conducted, aimed at satisfying the assumptions of the multivariate analysis. Thus, missing data analysis was performed to identify univariate outliers and multivariate outliers. Likewise, normality skewness and kurtosis, as well as multicollinearity issues, were checked. The preliminary analysis indicates that the data fulfil the conditions of multivariate analysis, thus, suitable for inferences.

Highlights

In any multivariate analysis, data screening is an important aspect that lays the foundation of the results of a quantitative study
The quality of the outcome of the study and acceptability depends on the data screening though; researchers mostly skip this stage of research due to its burden (Gorondutse & Hilman, 2014; Maiyaki & Moktar, 2011)
The sample for the study was collected from the rice farmers in the 17 rice clusters of the 7 local governments of Kano state that are cultivating paddy rice

Summary

Introduction

Data screening is an important aspect that lays the foundation of the results of a quantitative study. It is imperative to identify any likely violation of the underlying assumptions of multivariate analysis through data screening (Hair Jr., Black, & Anderson, 2010). The quality of the outcome of the study and acceptability depends on the data screening though; researchers mostly skip this stage of research due to its burden (Gorondutse & Hilman, 2014; Maiyaki & Moktar, 2011). The absence of data screening often leads to poor quality of the result and accuracy of the type of analysis used. Tabachnick and Fidell (2013) believed that the screening could be performed by proofreading of the original data generated from the computer file. The need to examine the data using computer software for descriptive statistics, since the computer software exposes hidden errors that might not be discovered through proofreading (Hair Jr. et al, 2010)

Methods

Results

Conclusion