DEVELOPMENT AND UTILISATION OF WORK SKILLS SCALE FOR ASSESSING JOB PERFORMANCE OF SENIOR UNIVERSITY NON-ACADEMIC STAFF IN SOUTHWESTERN NIGERIA

Abstract

Job performance entails scalable activities and behaviours that are assessed with the Annual Performance Evaluation Report (APER) form among University Senior Non-academic Staff (USNS) on grade levels 6-12 in Nigeria. However, the APER form has been regarded deficient in objective assessment for its lack of job factors, and indices and rubric on work skills. Previous studies focused largely on work performances of service oriented employees such as teachers, engineers, nurses and doctors but little on work skills scale that can objectively measure USNS job performance. This study was, therefore, designed to develop a work skills scale to measure USNS performance and determine the predicting effects of demographic factors (highest qualification, school ownership, age of staff, age of university and years of experience) on their skills.

Katz’s Theory of Work Skills and Koopmans’ Heuristic Performance Framework guided the study. The analytical survey of the ex-post facto design was adopted. Three phases and three sets of samples were involved after establishing the scale’s content validity (0.9) with the Lawshe ratio method. Phase 1 entailed pilot testing of an initial pool of 202 work skills items using 251 randomly selected respondents (r=0.95). The six states in southwestern Nigeria were sampled. One each of federal and state universities per state were purposively selected owing to existence of one each per state, while a private university was randomly selected per state. Phase II entailed item selection, validation and calibration of the scale. Random sampling was adopted to select 2,748 USNS from 12 non-academic units across the universities. Phase III entailed utilisation of the scale involving 305 respondents. Data were analysed using Exploratory Factor Analysis (EFA), Parallel Analysis (PA), Confirmatory Factor Analysis (CFA), Graded Response Model (GRM), descriptive statistics and Partial Least Squares (PLS).

Nineteen factors were extracted through EFA, but 10 factors comprising 60 items, were retained for further analysis. The remaining items were reduced to 39 which loaded on three factors denoted as Basic Skills (BS)=(20), Personal Attitude to Work (PAW)=(13) and Workplace Value (WpV)=(6) through PA. The factor structure was confirmed by CFA giving moderate fit indices, Root Mean Square Error of Approximation (0.8) and Root Mean Residual (0.03). Item calibration using GRM, slope (a1- a3) ranged from 1.75 to 4.80 and intercept (c1 - c5) from 0.26 to -11.21. These indicate that the items possessed high discrimination and model fit was fair and satisfactory. Ordinal reliability index of the multidimensional scale was 0.97. Composite reliability of sub-scales were BS=(0.97), PAW=(0.95) and WpV=(0.79). The group X̅=3.87 revealed that the USNS job performance is high. Regression of the sub-scales and demographic variables, using PLS indicate that PAW contributed mostly to work skills (0.93, t=89.63). Also, academic qualification with (t=4.1; p<0.05) predicts work skills, while school ownership, staff age, years of experience and university age did not.

The developed Work Skills Scale effectively measured job performance of University Senior Non-academic Staff with academic qualification predicting their work skills in the selected universities in southwestern Nigeria. The scale is recommended for university stakeholders.