Publication Title A Comparison of Logistic Regression and Discriminant Analysis in Predicting Student’s Academic Outcome. Download PDF
Publication Type journal
Publisher International Journal of Innovative Mathematics, Statistics & Energy Policies 10(2):1-9, April-June, 2022
Publication Authors F. Meka1 & R.N. Nwaka
Year Published 2022-06-09
Abstract The need to predict the academic outcome of new intakes in institutions of learning may arise at sometimes. The most widely used statistical methods for prediction when categorical outcome variables are involved includes Linear Discriminant Analysis and Logistic Regression. The question of which classifies better now comes to mind. Data of students who performed academically above average and that of those that performed academically below average was retrieved from their admission forms and analyzed with both Logistic Regression and Discriminant Analysis. The contribution of this work is twofold. First it compares Linear Discriminant Analysis and Logistic Regression on academic outcome variables in institutions of learning revealing a higher predictive accuracy in Logistic Regression than in Discriminant Analysis. Secondly, the study revealed that the marital status of parents and mother’s occupation are key pointers to the academic achievement of learners.
Publication Title AN ANALYSIS OF RURAL POVERTY IN OYO STATE: A PRINCIPAL COMPONENT APPROACH Download PDF
Publication Type journal
Publisher JORIND 9(2) December, 2011. ISSN 159
Publication Authors O.I. Osowole, Rita Ugbechie
Year Published 2011-12-02
Abstract The 2003/2004 NLSS data was used in this paper to consider the status of rural household poverty in Oyo State, Nigeria and to identify the relevant determinants using the. Principal component analysis (PCA) was employed to determine the important factors associated with rural household poverty. Household size, mother’s educational level, age of the household head, father’s work and mother’s work were the key factors found to be major determinants of rural poverty incidence in the state. The choice of expenditure as a proxy for measuring poverty was further corroborated. These findings indicated that factor analysis is very helpful in poverty -targeting and alleviation.
Publication Title Analytical Comparison of Model Information Criteria in ASARIMA Identification: Evidence from Benin Monthly Precipitation Download PDF
Publication Type journal
Publisher International Journal of Innovative Mathematics, Statistics & Energy Policies 11(2):10-22, April-June, 2023
Publication Authors Amaefula Chibuzo G 1* & Nwaka Rita N.2
Year Published 2023-06-22
Abstract The paper is concerned with optimal forecasts performance for Benin monthly rainfall (BMRF) pattern in Edo State, Nigeria using adjusted SARIMA (ASARIMA) model. It employed Amaefula forecast criterion (AFC) to compare the models selected by different information criteria such as Akaike information criterion (AIC), Schwarz Bayesian information criterion (SBIC), Hannan and Quinn(HQ), and forecast prediction error(FPE) to identify the optimal model. . The BMRF data span from 1981M1 to 2016M12. The order of integration test (OIT) adopted shows that BMRF data is integrated order zero (I(0)) in its regular series. The evidence of significant seasonal oscillation in the ACF and PACF reveal the need for seasonal differencing of MRF data. Using an iterative algorithm for calculating least squares estimates of the model parameters, 19 possible ASARIMA (P, D, Q)12 models were compared; AIC, SBIC and HQ preferred ASARIMA(1, 1, 2)12, ASARIMA(0, 1, 1)12 and ASARIMA(1, 1, 5)12, respectively. FPE supports the choice of SBIC. The three selected models were subjected to forecast optimality test using the AFC and ASARIMA (1, 1, 2)12 is found most appropriate. The diagnostic test indicates adequacy of the fitted model. Hence, ASARIMA (1, 1, 2)12 model is recommended for forecasting BMRF pattern and creating inter-mediate warning against erosion and flood in Benin
Publication Title Another Statistical Distribution: The Exponentiated Complementary Mukherjee-Islam Distribution Download PDF
Publication Type journal
Publisher EJERS, European Journal of Engineering and Technology Research Vol. 6, No. 1, January 2021
Publication Authors Oyedeji Isola Osowole, Ngozi Nzelu, and Rita Nwaka
Year Published 2021-01-02
Abstract Abstract — This study considered a newly proposed Exponentiated Complementary Mukherjee-Islam distribution obtained by exponentiating the Complementary MukherjeeIslam distribution. Some properties of the new distribution were derived and results from the distribution indicated that the distribution is a better alternative than its baseline distribution. The new distribution is therefore a creditable addition to the existing family of exponentiated distributions.
Publication Title ASSESSMENT OF THE BASIC DIMENSIONS OF POVERTY TO CHOICES OF THE POVERTY LINE IN NIGERIA. Download PDF
Publication Type journal
Publisher African Journal of Science and Research,2016,(5)4:09-13 ISSN: 2306-5877
Publication Authors 1,*Osowole, O. I., 2Bamiduro, T.A., 3Nwaka, Rita N., 4Adeniyi, O.S
Year Published 2016-05-06
Abstract Most poverty measures require poverty lines for their computations.This study undertakes an assessment of the basic dimensions of poverty, namely incidence, depth and severity to varying multiple poverty lines in a uni-dimensional setting. Results of the computationally appealing unidimensional poverty analysis indicate that the estimated basic dimensions of poverty are sensitive to the choices of the poverty lines with the two-third median per capita household expenditure poverty line having the highest gain in precision when the multiple poverty lin es are compared with the conventional two-third mean per capita household expenditure poverty line.The use of a fraction of the poverty line has been validated in the study and a plausible poverty line for future studies is the two-third median per capita poverty line.
Publication Title Improved Generalised Foster-Greer-Thorbecke Index In One-Dimensional Poverty Download PDF
Publication Type journal
Publisher International Journal of Innovative Mathematics, Statistics & Energy Policies 10(2):10-24, April-June, 2022
Publication Authors 1Nwaka, R. N. , 2Osowole O. I & 3Meka, F.
Year Published 2022-06-24
Abstract The use of income or expenditure as a measure of poverty is known as a one-dimensional study of poverty. This method relies largely on the use of a poverty line as a classification criterion to classify households into poor and non poor groups. The main issue with this approach is its sensitivity to extreme observations. Additionally, it ignores variability among the poor. These issues make the Foster-GreerThorbecke (FGT) generalized uni-dimensional poverty index to be restricted in its usage. Therefore, this study intends to access the suitability of a hybrid poverty line and an Improved Generalized Foster-GreerThorbecke (IGFGT) poverty index in a one-dimensional poverty study.
Publication Title Modelling Predictors for Usage of facemasks among the students of the Federal Polytechnic Ado-Ekiti, Nigeria Download PDF
Publication Type journal
Publisher International Journal of Novel Research and Development (www.ijnrd.org)
Publication Authors Ajao, I. O., Nwaka R. N., Abereifa, O. L.
Year Published 2023-05-05
Abstract Face cover use by the overall population for restricting the spread of the COVID-19 pandemic has been causing arguments across the globe, however progressively suggested, and the capability of this mediation is not surely known. This paper addressed the use of facemasks among the students of Federal Polytechnic Ado-Ekiti, Nigeria. Three factors investigated to be predictors for use of facemasks in the institution are gender of the students, prior keeping to the rule of social distancing and their various schools or faculties. The log-linear model was employed in order measure the students’ tendencies on the use of facemasks and relationship among the variables. Proportional allocation sampling technique was used to carry out the survey on the students from four faculties in the institution. The variable facemasks has three ordered levels (always, sometimes, never) upon which frequencies were predicted from the identified variables. Data collected revealed that 47% of the students claimed to always use facemasks while 24% claimed they have never used it. Results revealed that there is a strong relationship between use of facemasks and those who keep social distancing rules and faculties (schools) at 5% level having p-values 3.42e-12 and 0.018 respectively. Results of analysis on three-factor terms shows that model (GFS, GD, FD, DS) having residual deviance value 17.4 on 17 degree of freedom and p = 0.43 gave the closest fitted values to the observed frequencies, and consequently the best model. This suggests that the institution should continue to encourage the students to maintain social distancing in order have reasonable number of them complying with the rule of facemasks usage.
Publication Title ON THE IDENTIFICATION OF CORE DETERMINANTS OF POVERTY: A LOGISTIC REGRESSION APPROACH Download PDF
Publication Type journal
Publisher Mathematical Theory and Modeling www.iiste.org ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online) Vol.2, No.10, 2012
Publication Authors 1Osowole, O.I., 2Ugbechie, Rita, 3Uba, Ezenwanyi
Year Published 2012-10-10
Abstract This study carried out a logistic regression modelling of poverty status of households in Nigeria to identify possible determinants of poverty using the 2003/2004 National Living Standard Survey (NLSS) data. This approach differs from classical regression methods that place inappropriate restrictions on the residuals of the model. The results of the logistic regression showed that household size and educational group for highest level attained by the household head were the most significant determinants of poverty. Others include sex of the household head, age in years of the household head, father’s education level, father’s work, mother’s work, and occupation group of the household head. It strongly recommends that moderate household size and acquiring of formal education be encouraged to reduce poverty prevalence.
Publication Title On the k-Modified Generalized Uniform Distribution Download PDF
Publication Type journal
Publisher International Journal of Innovative Information Systems & Technology Research 8(2):85-99, April-June, 2020
Publication Authors 1Osowole, O. I. & 2Nwaka, Rita
Year Published 2020-06-06
Abstract The newly proposed k-Modified Generalized Uniform Distribution was considered. Some fundamental properties of the generalized distribution were derived. The theoretical and simulation results from the kModified Generalized Uniform Distribution indicated reasonable degree of concordance. This implies that the k-Modified Generalized Uniform Distribution can be successfully used to model observations with limited support.
Publication Title POVERTY INCIDENCE IN NIGERIA: AN APPLICATION OF THE BINOMIAL TEST Download PDF
Publication Type journal
Publisher ICASTOR Journal of Mathematical Sciences Vol. 10, No. 1 (2016) 5 - 16
Publication Authors O. I. Osowole, Rita N. Nwaka,, J. Kayode, F.O. kazeem,
Year Published 2016-01-01
Abstract This study considers an alternative estimation of the incidence of poverty based on the binomial test. The core objective is to explore an available alternative in the presence of limitations posed by consumption measurement problems, inadequate technical capacity and paucity of funds to conduct reliable household surveys. It is believed that the alternative approach considered will be less demanding and cost effective. This is the Nigerian situation with dwindling resources for the collection of necessary household survey data. The incidence of poverty is estimated from the binomial test using a recent general household survey panel data with per capita expenditure as the proxy for poverty. The estimate of the incidence of poverty from the binomial test compared well with the estimate expected conventionally. The alternative estimation procedure from the binomial test may therefore be adopted by socio and economic researchers especially in less developed nations because it gives an indication of the increase or decrease in poverty incidence. This will indeed aid assessment of poverty intervention programmes.
Publication Title PROXY MEANS TEST (PMT) ANALYSIS OF POVERTY IN OYO STATE Download PDF
Publication Type journal
Publisher International Journal of Basic, Applied and Innovative Research IJBAIR, 2018, 7(3): 62-67 www.arpjournals.com; www.antrescentpub.
Publication Authors 1,*Osowole, OI, 1Awomokun, G, 2Nwaka, RN, 3Balogun, KO
Year Published 2018-08-08
Abstract This study attempts an assessment of the effectiveness of the proxy means test (PMT) procedure in achieving better targeting of the poor in Oyo State, Nigeria. This is because proper identification and corresponding targeting of the poor is still a challenge in poverty analysis in Africa. The PMT method was therefore used to estimate household expenditures, corresponding poverty statuses of the households, inclusion and exclusion rates using data from the National Living Standard Survey (NLLS) for the five poverty lines considered. The PMT method gave higher percentages of the poor compared to the conventional method for all the five poverty lines. The implication of this finding is that the PMT method could indeed be used alternatively for improved targeting of the poor, especially in Oyo State.
Publication Title Statistical Analysis of Income Poverty and Inequality in Nigeria Download PDF
Publication Type journal
Publisher / Elixir Statistics 76 (2014) 28464-28465
Publication Authors Osowole, O.I1,*, Uba, Ezenwanyi2 , Ugbechie, Rita3 , Aghamie, S.O3 and Olusola, T.J1
Year Published 2014-04-04
Abstract The negative implications of rising poverty and inequality have awakened research concern in recent time due to the perceived correlation between both issues. This study was carried out with the aim of investigating the effect of the relationship between income poverty and inequality in Nigeria because few researchers have studied them together in past studies. The head count index was used to estimate the incidence of poverty while the Gini coefficient was used to estimate income inequality based on household data from a Living Standard Survey of households in Nigeria. The national estimates of the head count and Gini coefficient were 0.34 and 0.69. For the six geo-political zones, the estimates were North East (0.36, 0.81); North West (0.30, 0.76); South South (0.36, 0.66); South West (0.33, 0.63); North Central (0.24, 0.61) and South East (0.26, 0.57) respectively. The results indicated that poverty was highest in the North East (0.36) and lowest in the North Central (0.24). Also, income inequality was highest in the North East (0.81) and lowest in the South East (0.57). This study has revealed the positive correlation between poverty and inequality because results had shown that poverty was highest where inequality was also highest. Therefore to reduce poverty, efforts should be made to ensure reduction in income inequality.
Publication Title Statistical Analysis of Poverty in Oyo-State: A Q2 -Approach Download PDF
Publication Type journal
Publisher West African Journal of Industrial & Academic Research Vol.12 No.1 December 2014
Publication Authors , Osowole, O.I, 2Rita, Ugbechie, 3Aghamie, S.O., 4Nwoke, Fransisca, 5Balogun, K.O., 6Kazeem, F.O., 7Abolarin, Mofoluwake
Year Published 2014-12-12
Abstract The limited scope of uni-dimensional money-metric approach of analyzing poverty based on either expenditure or income has made approaches that allow poverty to be studied at several dimensions a necessity. It is in the light of this, that this study was aimed at analyzing poverty in Oyo State using a Q2 -approach. This approach combined both quantitative and qualitative aspects of poverty. The quantitative aspect was based on the per capita expenditure head count index while the qualitative aspect was based on fuzzy set poverty index derived from the following pre- selected indicators: major source of water for drinking and cooking, type of fuel used for cooking, toilet facilities, electricity supply and information and communication technology (ICT). Estimates of both quantitative and qualitative poverty indices were obtained from the National Survey of households’ data having a total of 508 households. The per capita head count and fuzzy set poverty indices showed that 196 and 316 households were respectively poor. These estimates implied that approximately 39% and 62% percent of households were poor. The qualitative estimate further indicated that more than 50% of the households lacked access to the pre-selected deprivation indicators with lack of access to quality water ranking highest. The Q2 Approach has therefore provided better insight into the nature of poverty in Oyo State. It has also identified provision of quality water as being paramount to any government-designed poverty intervention policy.
Publication Title CENSORED FOSTER GREER THORBECKE (FGT) INDEXES IN POVERTY ANALYSIS IN NIGERIA Download PDF
Publication Type journal
Publisher International Journal of Basic, Applied and Innovative Research IJBAIR, 2018, 7(4): 129 - 133
Publication Authors 1*Osowole, O.I., 1Ajibola, J., 2Nwaka, R.N., 3Balogun, K.O
Year Published 2018-08-08
Abstract This study is an attempt to assess the prospects of censoring in one-dimensional poverty analysis in Nigeria. The conventional method involves truncation of the non-poor households and this often leads to measurement error. Censoring instead of truncation was considered using data from the Nigeria General Household Survey (NGHS) Panel Survey. The estimated poverty indices obtained via censoring gave higher precision estimates compared to the indices obtained traditionally. This finding implies that censoring could reduce the problem of measurement error involved in the traditional poverty analytical procedures.