Mathematical Approach to Recruitment & Selection

Preliminaries of Mathematical Models

Human resource professionals derive their practices from previously conducted research that has yielded numerous theories (Sulich 2015). Practitioners utilize those theories to implement strategies designed to best manage their workforces (Ardichvili, Michell & Jondle 2009). Different mathematical models and theories which may be commonly used and developed in the recruitment area. Considering job market as a two-sided model (Arcaute & Vassilvitskii 2009) is not enough, because relations between participants are complex. This statement demonstrates how mathematical techniques have evolved throughout time in order to avoid the complexity of the hiring process.

The success of hiring process rests on selecting the proper approach and methodology (Johanson 2009). The algorithms and models are secret and proprietary for all companies - issuers and their clients. Although there are some scientific publications describing preliminaries of mathematical models used in recruitment (Weber 2012). 

Statistical Techniques

Structural Equation Modeling (SEM) is a family of statistical models that seek to explain the relationship between multiple variables and are more effective when testing models that are path analytic with mediating variables and contain latent constructs that are being measured with multiple indicators (Hair et al. 2006). SEM researchers propose a two-step procedure when testing theoretical models (Medsker et al. 1994).

1. The first step is to examine and validate the measurement model
2. The second step tests the structural model and conducts hypothesis tests

(Garver and Williams, 2009).

    Measurement model: Validity and Reliability

Confirmatory Factor Analysis (CFA) was used for a simultaneous assessment of overall and specific elements of measurement validity and reliability. CFA showed that all factor loadings and path coefficients were statistically significant. The t-values were above the required value of 1.96. Convergent validity is good as all the items and variables has high and significant factor loadings greater than 0.60 (Bagozzi and Yi 1988). Discriminant validity was measured by comparing the square root of the average variance extracted to the correlation between constructs (Braunscheidel and Suresh 2009). Moreover, the correlation coefficients among the constructs do not exceed 0.85, indicating that multicollinearity is not a problem (Kline 2005). Reliability estimates for the three scales, based on the CFA, all exceeded the 0.70 cut-off value suggested by Hair et al. (2006) providing evidence of scale reliability (RMs-0.92, SMR-0.89, OLR-0.72, ADV-68, ICS-0.81and ROs-0.83). The combined variance accounted for was 77.52%. Common method bias is less of an issue if more than one factor is identified, and none of the factors account for the majority of the variance explained (Patel and Conklin 2012). The overall fit statistics of the CFA are χ2 = 83.107; d.f.= 76; p =0.084; CFI = 0.815; GFI = 0.817; RMSEA = 0.076; CMIN/df = 2.358; RMR = 0.034; and NFI = 0.938.

1. Social Media Recruitment (SMR)
2. Recruitment Methods (Rms)
3. Online Recruitment (OLR)
4. Job Advertisement (JA)
5. Recruitment Outcomes (Ros)
6. Information Credibility and Sufficiency (ICS)
7. Perspective of Pre-Hire Outcome (PEHO)
8. Post-Hire Outcome (POHO)
9. Model Fit Indices (CFI, GFI, and NFI)

Measurement fit model and the mean, standard deviations, and correlations among variables (Muduli and Trivedi 2019) 

Hypothesized model (Muduli and Trivedi 2019)

    Structural Model and Findings

The c2 statistic was non-significant (χ2 = 103.347; d.f. = 75; p = 0.01), indicating an acceptable fit (Kline 2005). Each of the remaining models fit indices shown in Table 3(CFI, GFI, and NFI) exceed the acceptable fit level of 0.90 (Kline 2005). The RMSEA does not exceed the acceptable fit measure of 0.08 (Browne and Cudeck 1993), nor does the RMR exceed 0.05 (Kline 2005). The probability value that the model is a close fit is convincing at 0.950. Joreskog and Sorbom (1996) suggested that the p-value for this test should be >0.50.

Measurment structural model (Muduli and Trivedi 2019)

The path estimates show that RMs significantly relates with RO (β = 0.176, p < 0.001). In detail, SMR significantly relates with PEHO (β = 0.236, p < 0.005) and POHO (β = 0.186, p < 0.005), OLR significantly relates with PEHO (β = 0.108, p < 0.005) and POHO (β = 0.132, p < 0.005), JA significantly relates PEHO (β = 0.088, p < 0.001). Only JA is insignificantly related with POHO (β = 0.055, T = -0.853) (Muduli and Trivedi 2019).

Path Coefficients Model (Muduli and Trivedi 2019)

Mathematical Approach Case Example (Survey)

A survey done by Chakravarthy and Selvakumar (2018), evolved a study has been taken on the topic “A Study on effectiveness of Recruitment and Selection Process at Jiffy solutions, Chennai”, to analyze the effectiveness and various sources of the recruitment and selection process. Analysis and interpretation are based on the responses of 150 respondents who were administered a questionnaire that contains personal data and their views about the recruitment and selection process and procedure. By analyzing the data, the researcher came to the findings that 100 percent of the respondents are aware of the recruitment and selection process and 70.67 percent of the respondents say that HR policies are a major factor influencing the recruitment and selection process.

The mathematical approach of this survey has found the following findings.

• 100 per cent of respondents are having awareness of recruitment and selection process of the organization.
• 90 per cent of respondents preferred both sources of recruitment.
• 100 per cent of respondents said that employee referral is the main source of internal recruitment.
• 56 per cent of respondents came to know about vacancy through consultants.
• 71.333 per cent of respondents feel that consultants play a vital role in recruitment process.
• 70.67 per cent of respondents agreed that HR policies are the factor that influences the recruitment process.
• 77.33 per cent of respondents believe that recruiter should be knowledgeable and experienced personnel.
• 43.33 per cent of respondents are satisfied with the recruitment and selection process of the organization.
• 5.33 per cent of respondents are disagreeing in saying that recruitment is a challenging task for the recruiter.
• 30.67 per cent of respondents prefer technical interview in case of selection process.
• 31.33 per cent of respondents prefer HR interview in case of selection process.
• 52 per cent of respondents are highly satisfied with the candidate eligibility verification followed in the organization.
• 94.67 per cent of respondents agreed that selection of candidates is strictly adherence to the HR policies of the organization.
• 70.67 per cent of respondents said that no preference is given to female candidates.

List of References

1. Arcaute, E and Sergei, V. (2009). Social Networks and Stable Matching in the Job Market
2. Ardichvili, A., Mitchell, James and Douglas Jondle (2009). Characteristics of ethical business cultures. Journal of Business Ethics, 85(4), 445-451.
3. Bagozzi, R.P. and Yi, Y. (1988), “On the evaluation of structural equation models”, Journal of the Academy of Marketing Science, Vol. 16 No. 1, pp. 74-94.
4. Braunscheidel, M.J. and Suresh, N.C. (2009), “The organizational antecedents of a firm’s supply chain agility for risk mitigation and response”, Journal of Operations Management, Vol. 27 No. 2, pp. 119-140.
5. Browne, M.W. and Cudeck, R. (1993), “Alternative ways of assessing model fit”, Sage focus editions, Vol. 154, pp. 136-162.
6. Chakravarthy, C.S. and Selvakumar (2018) “International Journal of Pure and Applied Mathematics,” A study on recruitment and selection process in jiffy solution, 119(12). Available at: https://www.ijpam.eu/.
7. Garver, M.S. and Williams, Z. (2009), “Examining a model of understanding customer value and satisfaction data”, Marketing Management Journal, Vol. 19 No. 1, pp. 113-132.
8. Hair, J., William, C.B., Barry, J.B. and Rolph, E.A. (2006), Multivariate Data Analysis, 7th ed., Prentice Hall, Englewood Cliffs, NJ.
9. Johanson, P. (2009). Human Resource Management in changing organizational contexts. [In:] D. G. Collings & G. Wood (Eds.), Human resource management: A critical approach (pp. 19-37). London: Routledge.
10. Joreskog, K.G. and Sorbom, D. (1996), PRELIS 2 user’s reference guide: a program for multivariate data screening and data summarization: a preprocessor for LISREL, Scientific Software International, Lincolnwood, US.
11. Kline, R.B. (2005), Principles and Practice of Structural Equation Modeling, second ed., Guildford Press, New York, NY.
12. Medsker, G.J., Williams, L.J. and Holahan, P.J. (1994), “A review of current practices for evaluating causal-models in organizational-behavior and human-resources management research”, The Journal of Management Development, Vol. 20 No. 2, pp. 439-464.
13. Muduli, A. and Trivedi, J.J. (2019) “Recruitment methods and outcomes,” Recruitment methods, recruitment outcomes and information credibility and sufficiency.
14. Patel, P.C. and Conklin, B. (2012), “Perceived labor productivity in small firms—the effects of high–performance work systems and group culture through employee retention”, Entrepreneurship: Theory and Practice, Vol. 36 No. 2, pp. 205-235.
15. Sulich, A. (2015). Mathematical models and non-mathematical methods in recruitment and selection processes.
16. Weber, L. Your Résumé vs. Oblivion, Inundated Companies Resort to Software to Sift Job Applications for Right Skills. The Wallstreet Journal. January 2012.