Logistic regression and discriminant analysis of academic staff-mix by rank research performance in a university setting.

To safeguard the university system, guidelines for appointments and promotions of academic staff are designed to serve as a benchmark for assessing and appraising the staff. Appraisal of academic staff in Nigerian universities has become a subject of controversy in recent times. This paper is aimed at finding a classifier for the discrimination of academic staff of a university system of known states or staff categories, viz.: S1 - Lecturer I and below, S2 - Senior Lecturer and -S3 Professor/Associate Professor, into latent subgroups on the basis of their research proficiency. A combination of cluster analysis and linear discriminant analysis was first used as a framework for identifying three latent subgroups, viz., mover, mediocre and stayer, with the quantity of scholarly publications, quality of academic journals in the system publish in, and the author-level citation index as input variables. Principal component analysis in combination with logistic regression was used to investigate and classify a (training) data set of a cross-section of academics of several categories with diverse research features from different universities within Nigeria. The results revealed that there are more stayers in S2 , and more movers in S3 . A comparison of the staff categories indicates that the research performance of academics in S 3 outstrips those in S1 and S 2 , and that academics in S1 did better than those in S 2 .The methods reported here have potential utility for the latent intra-class categorisation of staff of the research oriented system within the mover-mediocre-stayer paradigm. The method is useful for shortlisting applicants for interview to a more appropriate staff category of the system.

Adomi, E. E.and Mordi, C. (2003). Publication in foreign journals and promotion of
academics in Nigeria, Learned Publishing, 16(4), 259–263. http://dx.doi.org/
10.1087/095315103322421991.
Affes, Z. and Hentati-Kaffel, R. (2019). Predicting US banks bankruptcy: logit versus
canonical discriminant analysis, Computational Economics, 54, 199–244.
https://doi.org/10.1007/s10614-017-9698-0.
Arevalillo, J. M. and Navarro, H. (2012). A study of the effect of kurtosis on discriminant
analysis under elliptical populations. Journal of Multivariate Analysis, 107, 53 – 63.
Bartholomew, D. J., Forbes, A. F. and McClean, S. I. (1991). Statistical Techniques for
Manpower Planning (2nd ed.), John Wiley & Sons, Chichester.
Bi, Y. and Jeske, D. R. (2010). The efficiency of logistic regression compared to normal
discriminant analysis under class-conditional classification noise, Journal of
Multivariate Analysis, 101, 1622 – 1637.
Boudaghi, E. and Saen, R. F. (2014). Developing a model for determining optimal  in
DEA-discriminant analysis for predicting suppliers’ group membership in supply
chain, OPSEARCH. DOI: 10.1007/s12597-014-0173-6.
Chauhan, S., Darvishzadeh, R., Boschetti, M. and Nelson, A. (2020). Discriminant analysis
for lodging severity classification in wheat using RADARSAT-2 and Sentinel-1
data, ISPRS Journal of Photogrammetry and Remote Sensing, 164, 138 – 151.
Choubin, B., Moradi, E., Golshan, M., Adamowski, J., Sajedi-Hosseini, F. and Mosavi, A.
(2019). An ensemble prediction of flood susceptibility using multivariate
discriminant analysis, classification and regression trees, and support vector
machines, Science of the Total Environment, 651, 2087 – 2096.
Cornell-Farrow, S. and Garrard, R. (2020). Machine learning classifiers do not improve the
prediction of academic risk: Evidence from Australia, Communications in Statistics:

Journal of the Nigerian Statistical Association, Vol. 34, 2022 Ekhosuehi & Iduseri
57
Case Studies, Data Analysis and Applications. DOI: 10.1080/2373748
4.2020.1752849.
Dattalo, P. (1995). A comparison of discriminant analysis and logistic regression, Journal
of Social Service Research 19, (3-4), 121 – 144.
De Feyter, T. (2006). Modelling heterogeneity in manpower planning: dividing the
personnel system into more homogeneous subgroups, Applied Stochastic Models in
Business and Industry, 22, 321 – 334.
Deming, K., Yaoyao, C., Lingfu, K. and Shutao, W. (2018). Classification of oil pollutants
based on excitation-emission matrix fluorescence spectroscopy and twodimensional discriminant analysis, Spectrochemical Acta Part A: Molecular and
Biomolecular Spectroscopy. https://doi.org/10.1016/j.saa.2019. 117799.
Demir, S.B. (2018). Predatory journals: Who publishes in them and why? Journal of
Informetrics, 12, 1296–1311. DOI: https://doi.org/10.1016/j.joi.2018.10.008
Ekhosuehi, V.U. (2017). On graduation of enrolment size in a multi-echelon educational
system, Mathematica Applicanda [Matematyka Stosowana], 45(2), 119 – 126. DOI:
10.14708/ma.v45i1.4378.
Ekhosuehi, V.U. and Omosigho, S.E. (2018). The use of certain staffing requirements as a
means of benchmarking academic staff structure, Mathematica Applicanda
[Matematyka Stosowana], 46(2), 259 - 272. DOI: 10.14708/mav46i2.5175.
Gao, J. and Li, L. (2019). A robust geometric mean-based subspace discriminant analysis
feature extraction approach for image set classification, Optik – International
Journal for Light and Electron Optics, 199, 163368.
Gaynanova, I. and Wang, T. (2019). Sparse quadratic classification rules via linear
dimension reduction, Journal of Multivariate Analysis, 169, 278-299.
Hardle, W. andSimar, L. (2007). Applied Multivariate Statistical Analysis (2nded.),
Springer-Verlag, Berlin.
Harris, G. and Kaine, G. (1994). The determinant of research performance: a study of
Australian university economists, Higher Education, 27, 191 – 201.
Katos, V. (2007). Network intrusion detection: evaluating cluster, discriminant and logit
analysis, Information Sciences, 177, 3060 – 3073.DOI: 10.1016/j.ins.2007.02. 034
Manjunath, B.G., Frick, M. and Reiss, R.-D. (2012). Some notes on extremal discriminant
analysis, Journal of Multivariate Analysis, 103, 107 – 115.
Maugis, C., Celeux, G. and Martin-Magniette, M. L. (2011). Variable selection in modelbased discriminant analysis, Journal of Multivariate Analysis, 102, 1374 – 1387.
McCullagh, P. (1980). Regression models for ordinal data, Journal of the Royal Statistical
Society, Series B, 42 (2), 109-142.
Minka, T.P. (2004). A comparison of numerical optimizers for logistic regression. Technical
Report 758, Carnegie Mellon University.
Nwosu, D.F., Onyeagu, S.I., Mbegbu, J.I. and Ekhosuehi, V.U. (2016). On refined principal
component method for factor analysis, Journal of the Nigerian Statistical
Association, 28, 16 – 29.
Nwosu, D.F., Ekhosuehi, V.U. and Mbegbu, J.I. (2020). Performance of some factor
analysis techniques, Annals of Data Science, 7, 209-242. DOI 10.1007/s40745-020-
00260-6.

Journal of the Nigerian Statistical Association, Vol. 34, 2022 Ekhosuehi & Iduseri
58
Oravec, M., Beganovic, A., Gal, L., Ceppan, M. and Huck, C.W. (2019). Forensic
classification of black inkjet prints using Fourier transform near-infrared
spectroscopy and linear discriminant analysis, Forensic Science International, 299,
128 – 134.
Parmet, Y., Schechtman, E. and Sherman, M. (2010). Factor analysis revisited – How many
factors are there? Communications in Statistics – Simulation and Computation, 39,
1893 – 1908.
Press, S.J. and Wilson, S. (1978). Choosing between logistic regression and discriminant
analysis, Journal of the American Statistical Association, 73, 699 – 705.
Rencher, A.C. (2002). Method of Multivariate Analysis (2nd ed.), John Wiley and Sons Inc.,
New York.
Renjith, G., Mary, D.P., Soe, K., Wan, M.Y., Beh, H.C., Phuah, W.H. andTseu, A.T.H.
(2019). Sex estimation by discriminant function analysis using anatomical location
of mental foramen, Forensic Science International: Reports, 1, 100018.
Setser, A.L. and Smith, R.W. (2018). Comparison of variable selection methods prior to
linear discriminant analysis classification of synthetic phenethylamines and
tryptamines, Forensic Chemistry. https://doi.org/10.1016/j.forc.2018.10.002
Tavernier, J., Simm, J., Meerbergen, K., Wegner, J.K., Ceulemans, H. and Moreau, Y.
(2019). Fast semi-supervised discriminant analysis for binary classification of large
data sets, Pattern Recognition, 91, 86 – 99.
Udom, A.U. and Ebedoro, U.G. (2019). On multinomial hidden Markov model for
hierarchical manpower systems, Communications in Statistics – Theory and
Methods. DOI: 10.1080/ 03610926.2019.1650185.
Xia, C., Yang, S., Huang, M., Zhu, Q., Guo, Y. and Qin, J. (2019). Maize seed classification
using hyperspectral image coupled with multi-linear discriminant analysis. Infrared
Physics & Technology. DOI: https://doi.org/10.1016/j.infrared. 2019.103077.
You, X., Xu, J., Yuan, W., Jing, X.-Y., Tao, D. and Zhang, T. (2019). Multi-view common
component discriminant analysis for cross-view classification, Pattern Recognition,
92, 37 – 51.
Zheng, J., Wang, H., Song, Z. and Ge, Z. (2019). Ensemble semi-supervised Fisher
discriminant analysis model for fault classification in industrial processes, ISA
Transactions, 92, 109 – 117.
Zhou, F., Guo, H.-C., Ho, Y.-S.and Wu, C.-Z. (2007). Scientometric analysis of geostatistics
using multivariate methods, Scientometrics, 73(3), 265 – 279. DOI:
10.1007/s11192-007-1798-5.
Zou, M., Sun, C., Liang, S., Sun, Y., Li, D., Li, L., Fan, L., Wu, L.and Xia, W. (2018).
Fisher discriminant analysis for classification of autism spectrum disorders based on
folate-related metabolism markers. Journal of Nutritional Biochemistry. DOI:
10.1016/j.jnutbio.2018.09.023.