A science recruitment revolution

How does a science employer go about getting their fair share of available applicants when recruiting? Is it all about salaries or does approach matter too? John Berry has some answers…

A recent ‘person specification’ for a senior consultant to be recruited to a scientific consulting practice read like it was describing a super-being. Competencies spanned the ability to present at international conferences, development and use of models for systems analysis and synthesis, use of Excel, development of software and application of geographical information systems.

This was before the addition of sector specific skills and knowledge, such as the ability to demonstrate fundamental knowledge in biological science and basic workplace and academic competencies like writing and personal organisation. You’d be right if you concluded that science is a pretty competency-rich environment.

But, like most things in life, the fewer degrees of freedom we’re prepared to allow, the fewer the opportunities – in this case, the fewer folk there out there who can excel in the job. And when there are fewer folk who can excel, the laws of economics apply – those who have the skills and the knowledge can expect inflated salaries. If the employer can’t meet the salary expectations (because perhaps the business simply can’t afford it) there is then pressure to relax the ‘person specification’, reducing those competencies in an effort to achieve a compromise.

Reports from the House of Lords Select Committee on Science and Technology make interesting reading. Government laments the numbers taking up STEM (science, technology, English and maths) subjects. Talking specifically about competency in maths, the Lords comment that ‘a high level of numeracy is of increasing importance to employers’. And in the same report, they go on to note that the UK is simply not producing enough science, technology and engineering technicians and graduates. If that’s true, our degrees of freedom look set to reduce further.

So, given this situation, how does a science employer go about getting their fair share of available applicants when recruiting? Indeed, if skills and knowledge are not abundant, how does the employer succeed, despite the tight employment market? Is it all about salaries or does approach matter too?

Traditional recruitment involves a three-step process; ask a recruitment agency to source CVs, select a shortlist from the CVs offered and interview. But how does one know that these are the only candidates available? In a tight market, only those actually unhappy in their work or those actually out of work will place their CVs on a jobs board to be found by a searching agency. Surely there must be a better way?

Traditionally, head-hunters succeeded where perhaps recruitment agents failed. They sourced by detective work, finding names and calling them in effort to encourage otherwise happy employees to look at new opportunities. And they charged over 25% of salary for their efforts.

Traditional selection involves a manager sifting the CVs and making a shortlist, perhaps seeking key attributes of applicant personal characteristics. Candidates from the shortlist are interviewed, often informally, and in meetings lasting around an hour. Research shows that few managers are trained in selection and, without some science in the process, the chances of a manager making a valid prediction of how well someone will do in a role are very small.

But things are changing and, arguably, they need to.

A recent report in Recruiter, the magazine of the recruitment industry, cited results of a survey showing that the leading jobs boards each only enjoy around 7% of successful placements. There’s a raft of lesser boards that enjoy significantly less. Towering high above them all, though, is LinkedIn at a colossal 21% of successful placements. Today everyone wants to be seen. Everyone wants to profess skills and knowledge to customers, future employers and colleagues alike. Only the recluse is below the radar.

It doesn’t take a detective to find people today. It just takes sound analysis of the requirements and a robust search. Finding staff is no longer a black art. Of course, it’s a different sort of person that’s found and managers need to be ready to unseat someone who perhaps wasn’t actually looking at that point in time.

Over the past 15 years or so, the science of selection has galloped ahead. Gut-feel is out and predictively valid selection instruments are in. Psychometrics help measure those numeracy skills. They assess those desirable workplace competencies, such as verbal and abstract reasoning. Managers know that it’s those who are pro-active and conscientious that make the best scientists: personality profiles measure these characteristics too. Managers know too that future staff must fit in the firm and person-environment fit can be assessed as well.

When it comes to interviews, the best include work sample tests. If the ‘person specification’ calls for competency in developing models to describe phenomena, ask the candidate to use Excel to build a model. And give them just 15 minutes to do it. The combination of test and pressure will show who will excel with Excel.

Those old interviews too can be modernised. Using competencies to sift CVs into a shortlist quantifies the task – the same competencies can be used to develop the interview questions and developing scored model answers makes decisions simpler.

The combination of psychometrics and improved testing and interviewing has raised the predictive validity from an R value of around 0.2 to around 0.6. For those statistically minded, R2, the percentage chance of successful recruitment from these selection instruments alone rises from 4% to 36%. While selection still isn’t perfect, it’s a lot better than the traditional ways.

In a tight market, where there are just too few folk to go round and where the competencies available fall short of that required, two huge changes are upon the science industry. The first is a revolution in the search activity: people can be found. And the second is an evolution in selection instruments: selection is now a science.

Technology has a habit of changing the goal-posts. There was a time when it was a huge skill to draw using tools like Harvard Graphics. Today, everyone can drive PowerPoint and the skill is in developing the message in the slide. With traditional recruitment, it was an art to head-hunt and interview. Today it’s a science and the skill is in describing the job right, developing valid queries, attracting candidates, building appropriate psychometrics and tests and making sure the selected candidate actually joins.

The old three-step process is now six: specify, search, recruit, select, appoint and induct. And the job’s not over until all six are complete.

Moving from art to science hasn’t changed the fundamentals though. Absence of future candidates entering education to eventually be in the labour market seeking scientific careers can’t be changed by changes to process. We’re still maybe after super-beings and there will still be a skills and knowledge shortage. But at least employers with vacancies can get a chance to illustrate what life would be like working for them. They’ll be able to persuade highly competent staff to move for increase in job-satisfaction and responsibility, even if the balance sheet won’t allow inflated salaries. If we can’t shake out more degrees of freedom, at least excellent employers can trade easier in the labour market.

Improvement in selection methods is in itself useful. Optimising predictive validity is a good thing. But there’s a spin-off. It’s not all one-sided when a candidate visits an employer. The candidate is evaluating the firm.

Splitting the recruitment phase from the selection phase serves two purposes. Firstly it allows the employer to focus on attracting the candidate at an initial meeting. An informal discussion can take place at which the bias is towards attracting rather than selecting. Secondly it allows a scientifically valid selection to take place at a second meeting.

Research shows that candidates who were put through a robust and valid selection process are more likely to accept a job offer, when compared with those who went through a traditional single interview.

So, while we can’t change the state of skills and knowledge in the UK in the short term, we can change the way employers approach the recruitment task. We can ensure that our firms are survivors and that we get to make choice from an adequate candidate pool using valid selection methods.