Master™ Recruitment Decision Tool

Developed by Laurits Skovlund Kilpinen Psychometrician at Master International

What is it?

The Recruitment Decision Tool helps you estimate the expected financial and organizational gains from switching from various recruitment methods to General Mental Ability (GMA) recruiting. It allows you to compare and analyze the benefits of using GMA-test in a recruitment scenario of your choice.

Schmidt, Oh, and Shaffer (2016) concluded, based on 100 years of research, that a GMA-test is the single best predictor of job performance. This fact is well-known in the HR-profession. The practical implications, however, can be difficult to interpret and difficult to communicate to decision makers. Because of this, we have developed the Recruitment Decision Tool.

The tool compares GMA-recruiting to one of 19 possible recruitment methods in a chosen recruitment scenario. The tool then presents the result of the comparison-analysis in the form of three practical and easily understood outputs.

The tool also has more advanced outputs targeting HR-professionals who want to go a step further in their analysis. For instance an applicant analysis showing the effect of adding more applicants – a knowledge that is important when planning a recruitment strategy. It also has a simulation analysis showing you simulations of the financial benefit which enables you to understand and explain the uncertainty in recruitment.

The tool can be used to aid decision-making, the understanding of recruitment methods, and the dynamics of different recruitment scenarios. The math and theory behind the model are explained in detail at the bottom of the page.

How to use it?

Fill in information about the recruitment scenario you want to use for the analysis. You have full control over the recruitment scenario. If you are a recruiter you can choose the recruiting your company did last year, the last five years, or it can be a purely hypothetical scenario. It is possible to combine different job positions, as long as they have a similar job complexity level. E.g., positions in marketing and positions in communication can be combined but not trainees and managers.

You can decide to include a structured interview alongside the GMA-test a strategy deemed optimal by Schmidt, Oh, and Shaffer (2016). At last you choose which recruitment method you want to use as a baseline for comparison. Click the (?) if you need additional information at any given step.

NB. If you have problems with the visualizations refresh the page. The application might be a bit slow on some computers if the number of applicants and number hired is high.

1. Recruitment scenario

€50,000

2 yrs

Low complexity:
semiskilled jobs, unskilled jobs,
industrial workers
Medium complexity:
skilled crafts, technician jobs,
lower level adminstrative jobs High complexity:
managerial/specialist,
complex technical set-up

Recruitment scenario summary:

xx positions filled

xx applicants per position on average

xx job level

xx tenure on average

xx in average salary

The expected benefits from using GMA for recruiting instead of xx recruiting, in the recruitment scenario stated above:

The expected total financial benefit
- due to the higher job performance by those hired in their duration at the organization:
xx

The financial benefit can be explained by the expected
xx higher job performance by the average person hired by GMA-recruiting.

xx

=

xx

You have to hire xx with xx to get the same amount of job performance as the xx hired with GMA-recruiting.

Interested in learning more about the use of GMA-test in real life?

Disclaimer: Master™ is under no circumstances liable for any incidental, punitive, consequential, special, or exemplary loss or damage directly or indirectly in connection with the interpretation and/or use of the tool and/or results provided through the tool. All use of the Recruitment Decision Tool is at own risk, and Master™ is under no circumstances liable for errors, mistakes, miscalculation, and/or inaccuracies of the tool. ...

Master™ is the proprietary owner of the Recruitment Decision Tool and shall retain all intellectual property rights and other rights, title, and interest in the tool and any improvements, design contributions, or derivative works thereto. Recruitment Decision Tool may not be reused, republished, reprinted or in any way redistributed without written approval of Master™.

The use of the Recruitment Decision Tool is intended as a sample calculation tool and cannot be considered as substitute for professional advice. The tool is developed to generate generalized figures based on assumptions and calculations administered by Master™. The tool, calculations, and/or results are not to be considered conclusive, reliable, and/or accurate and should therefore not be used professionally to analyse a specific candidate or a specific person.

Mandatory Danish product liability rules in force from time to time shall apply to disputes regarding product liability. Unless otherwise required by mandatory legislation, Master™ shall not be liable to the user for product liability.

The results above are all based on theoretical calculations showing the expected benefits. Real life recruitment however involves chance and you can be both unlucky and lucky with your recruiting regardless of how good your recruitment method is.

Statistics allows us to model and simulate the uncertainty of recruiting. Using a simulation method called Monte Carlo we have simulated the results of the recruitment scenario you chose xx times and included the randomness of real life recruiting in the simulations.

The distribution of the simulation results can be seen below. Hover over each bar to see how many percent of the simulations that fell in each range.

The mean of the theoretical calculations:xx

The mean of the xx simulations:xx

95 % of the simulated estimates are in the range: x - x

Simulation estimates above certain cutoff:

Value	% Above
xx	xx
xx	xx
xx	xx
xx	xx
xx	xx
xx	xx
xx	xx
xx	xx
xx	xx

The Recruitment Decision Tool also offers an applicant analysis. A complex analysis that model and simulate the expected financial difference of increasing or decreasing the number of applicants in a recruitment scenario where the recruitment method is known.

Our applicant analysis is powered by the recruitment scenario that you have decided. Unlike the other analysis in this tool it is not a comparison analysis it is solely based on recruitment by GMA test and if selected - a structured interview.

The benefit of adding applicants is not identical across all recruitment scenarios. In some scenarios the benefit of adding just one extra applicant per hire can be huge and in others in can be negligible. Knowing these benefits are important and can be the foundation of a cost-benefit analysis of whether to invest more money in attracting more applicants.

The output of the analysis is two graphs giving a visual overview of the applicant-curve and a numerical output showing the benefits of adding 1,3,5, and 10 more applicants per hire.

Applicant analysis

Yellow circle = Your applicants per hire ratio

Applicant analysis - a closer look

Applicant analysis - in numbers

Our simulation show that the expected benefits from getting more applicants-per-hire is:

Applicants	Financial benefit total	Financial benefit per hire	Increased job performance
1 additional applicant-per-hire:	xx	xx	xx
3 additional applicants-per-hire:	xx	xx	xx
5 additional applicants-per-hire:	xx	xx	xx
10 additional applicants-per-hire:	xx	xx	xx

The model is based on utility analysis. Utility analysis is an umbrella term for a group of mathematical models that is used to describe the practical value (utility) of different decision options – like the practical value of different recruitment methods. The utility model we use is called Brogden-Cronbach-Gleser and has its origin in the late 40’s (Brogden, 1949) it is the most used utility model in HR-psychology. All our outputs are based on this model with small tweaks.

In rough terms and in the most used way the model works by first calculating the average score of those hired with the employment procedures, in a z-score form. The math behind this calculation is a bit complicated, it is the ordinate in a normal distribution at the point of the selection rate, divided by the selection rate. If this confuse you, don’t worry, the most important thing to know is that it relies on the selection rate – the proportion of applicants who are hired. If you must hire all who apply for the job, then no hiring procedure will have any practical value. However, if you have the privilege to only hire top 1% of the people who apply then the practical value will be large (Schmidt & Hunter, 1998).

This number is then multiplied with the incremental validity the difference in the criterion validity of the two methods that are compared. If you compare integrity test recruiting with recruiting based on reference checks, then the incremental validity is 0.20 (0.46 – 0.26) (Schmidt, Oh & Shaffer, 2016). The incremental validity is directly linked to the practical economic value of recruitment methods (Schmidt & Hunter, 1998).

After multiplying with the incremental validity, we have the average difference in job performance by the people hired in a z-score format. To transform the difference in job performance from a z- score into a practical metric we multiply it with a measure of variability of job performance. To get the financial benefit, we use a dollar value of a standard deviation in job performance SDy. To get the percentage increase in job performance we use a percentage of mean output of a standard deviation in job performance SDp. If you are interested in more details about the calculations click on the tabs below.

$$ \Delta U = N_{s} * T_{s} * SD_{Y} * r_{XR} * \frac{\phi }{p} - N_{a}C $$

$$ \Delta U $$

The main result, the total incremental utility (also called financial benefit) from the new recruitment method compared to the baseline method.

$$ N_{s} $$

Number of people selected.

$$ T_{s} $$

Expected tenure for a person in this position.

$$ SD_{Y} $$

The added monetary value of a standard deviation differences in the predictor variable. In this algorithm we use 40% of the salary which probably is a lower bound for a longer discussion about this see (Schmidt & Hunter, 1983) and (Hunter, Schmidt & Judiesch, 1990). We have chosen not to let it be affected by complexity to ensure the robustness and stability of the model.

$$ r_{XR}$$

The incremental validity. The validity of the GMA-recruiting minus the validity of the comparison method. All validity estimates are from (Schmidt, Oh & Shaffer, 2016).

$$ \frac{\phi }{p}$$

The mean standardized predictor score of selected applicants. The ordinate in N(0,1) at the point of the selection rate, divided by the selection rate.

$$ N_{a}C $$

The cost per test multiplied with the number of applicants.