Authors : Ackmed Chebli; Dr. Alhaji Hamza Conteh; Dr. Brima Gegbe

Volume/Issue : Volume 11 - 2026, Issue 8 - August

Google Scholar : https://tinyurl.com/4cfvcjbz

DOI : https://doi.org/10.38124/ijisrt/26aug705

Note : A published paper may take 4-5 working days from the publication date to appear in PlumX Metrics, Semantic Scholar, and ResearchGate.

Abstract : Background: Earlier versions of this article contained unsupported sampling claims, duplicated effect sizes, inconsistent duration statistics and an unstable multivariable specification. This reanalysis uses the supplied cleaned microdata and treats data quality as part of the inferential problem.  Methods: The workbook contains 1,342 records. Employment status was analysed using Firth bias-reduced logistic regression because no employed respondent lacked internet access, creating complete separation. The primary model used education level, age, coded gender, coded marital status, province, internet access and a three-item perception index. Education years were examined separately. Recorded non-employment duration was modelled exploratorily among 922 strictly eligible records using a Gamma log-link model.  Results: Employment was recorded for 415 respondents (30.9%). The perception items were highly redundant (alpha = 0.961; pairwise r = 0.871-0.912) and were combined. In the bias-reduced model (n = 1,339), higher perception-index scores were strongly associated with employment (OR = 18.33, 95% CI 12.04-27.89), diploma education differed from no education (OR = 2.04, 95% CI 1.24-3.35), age was inversely associated after adjustment (OR = 0.72 per year), and the Eastern Province estimate was positive relative to Northern Province (OR = 1.78). Internet access produced an extreme estimate because it perfectly separated the outcome and is not interpreted as a stable population effect. Model AUC was 0.935; HosmerLemeshow p = 0.068.  Conclusion: The reanalysis supports conditional associations within the supplied records but not causal or nationally representative conclusions. The dataset contains structural dependencies and coding conflicts that materially limit substantive interpretation. Recovering the original questionnaire, sampling protocol and source records is necessary before external publication.

Keywords : Youth Employment; Sierra Leone; Education; Skills Alignment; Firth Logistic Regression; Data Quality; Complete Separation.

References :

1. African Development Bank. (2012). Youth employment in Africa. African Development Bank.
2. Becker, G. S. (1964). Human capital: A theoretical and empirical analysis, with special reference to education. University of Chicago Press.
3. Card, D. (1999). The causal effect of education on earnings. In O. Ashenfelter & D. Card (Eds.), Handbook of labor economics (Vol. 3, pp. 1801-1863). Elsevier.
4. Doeringer, P. B., & Piore, M. J. (1971). Internal labor markets and manpower analysis. Heath.
5. Filmer, D., & Fox, L. (2014). Youth employment in Sub-Saharan Africa. World Bank.
6. International Labour Organization. (2022). Global employment trends for youth 2022. ILO.
7. Kuhn, P., & Skuterud, M. (2004). Internet job search and unemployment durations. American Economic Review, 94(1), 218-232.
8. McGuinness, S., & Bennett, J. (2007). Overeducation in the graduate labour market: A quantile regression approach. Economics of Education Review, 26(5), 521-531.
9. Psacharopoulos, G., & Patrinos, H. A. (2018). Returns to investment in education: A decennial review of the global literature. Education Economics, 26(5), 445-458.
10. Schultz, T. W. (1961). Investment in human capital. American Economic Review, 51(1), 1-17.
11. Spence, M. (1973). Job market signaling. Quarterly Journal of Economics, 87(3), 355-374.
12. Verhaest, D., & Omey, E. (2006). The impact of overeducation and its measurement. Social Indicators Research, 77(3), 419-448.