Work Simulations: How to Assess Candidates Using Data

Published by:
Joe Caccavale
May 11, 2021
11
min read

Work Sample Cheatsheet

What are work simulations?

Work simulations (also known as work simulations) are a form of talent assessment designed to replicate parts of a role.

The basic idea behind a work simulation assessment is have candidates  perform tasks that they would encounter should they get the job.

Whilst traditional hiring methods rely on proxies to make assumptions about someones' ability, a work simulation test gets candidates to essentially do the job (or at least small parts of it) before actually getting it.

How you structure your job simulation will obviously depend on the job itself...

If you were hiring a bricklayer, you'd probably want to see how they build a wall.

If you were hiring a chef, you might want to try their food or see how they work in the kitchen.

For office-based roles, you can simply take the tasks or scenarios that would occur on-the-job and ask candidates how they would approach them.

Work sample example

Why use work simulations?

Traditional hiring methods are biased and outdated.

Candidates lacking specific experience or academic backgrounds are overlooked, perpetuating systemic inequality.

Someone’s experience can and often does make them a suitable candidate. However, it is no longer ethical or efficient to rely on this assumption.

Take a look at the results of the famous Schmidt-Hunter meta-analysis below…

Predictive validity of assessments (chart)


50+ years of research tells us that education and experience just aren’t very good at predicting someone’s ability.

But why is this?

Well... when we look at a CV, we're trying to tie someones education and experience to the skills needed for the job. If someone worked at a big company, we might assume that makes them competent. We might even assume that this makes them more competent than if they'd worked at a lesser known company.

These assumptions are sometimes correct. However, more often than not, they're wrong.

True, a certain combination of academic achievements and work experience might make someone the best possible hire...

But this could just as easily be a signifier of their socioeconomic background. And since this is difficult to separate from pure ability, we'd rather bypass these flawed proxies and test for skills directly - what could be more predictive than asking candidates to perform parts of the job itself?

‘Work samples’ are a type of job simulation and happen to be the most predictive form of assessment there is (you’ll notice they’re at the top of the chart above).

Work samples are similar to ‘tell me a time when…’ interview questions. The key difference is that they pose questions hypothetically, so that experience isn’t necessitated.

Instead of guessing at how well someone would perform in a job by looking at their background, work samples test this ability upfront.

This isn’t just a more empirical means of hiring, it’s also fairer.

Taking the emphasis off of education and experience means that every candidate gets a fair chance to show what they can do.

In a world where candidates with non-white sounding names need to send more than 70% more resumes to get the same number of callbacks as someone with a white-sounding name, how can we expect these people to have the sort of profile we’re looking for?

Extra CVs needed to receive 1 callback (chart)

What we might refer to as our 'gut instinct' is often just unconscious bias kicking in.

We subconsciously rely on mental shortcuts to draw patterns and make associations - often based on things like our past experiences, stereotypes and what we've seen in the media.

Whilst these shortcuts are extremely useful in our day-to-day lives, they can result in negative outcomes when it comes to people-based decisions.

By using a job simulation assessment to test someone's skills, you can reduce the opportunities for the bias to affect or judgment.

Since CVs are ineffective at identifying talent and trigger bias, we'd recommend using work samples anonymously at the screening stage.

Letting go of CVs completely can seem like a giant leap of faith. So, if you're not ready to take that step yet, try using work samples alongside CVs.

You'll find that those who you shortlist based on their work sample answers won't be the same people who would've made it via their CV alone.

Will your work simulation assessments put candidates off?

A common misconception about job simulations is that they deter candidates.

There will always be candidates who won't be happy that they can't rely on their credentials alone.

But the data suggest they are in the minority... and are these the people you really want to hire anyway?

Here at Applied, we have an 9/10 average candidate experience rating.
How candidates rate the Applied process (chart)

Why? Because candidates genuinely enjoy showing off what they can do, free from the assumptions that may arise from their background.


Turn2Us candidate quote

A work simulation test is undoubtedly more demanding of candidates than your typical CV/informal interview-based process.

And this may sometimes mean a smaller candidate pool.

However, this isn't a bad thing - you'll be fishing from a more concentrated pool of passionate, pre-qualified candidates.

And won't they cost more of my teams time too?

Using a process like work samples will require more time to be invested upfront...

Creating and scoring work samples takes more time than skimming CVs.

Although this is spent upfront will be regained later on - by using a more predictive assessment at the start of the hiring process you'll only need to interview candidates who've proven they have the skills needed for the job.

Work samples allow you to bring the value of the interview (where you get to see how candidates actually think and work) forward to the screening stage.

If you're dealing with hundreds of candidates for any single role, you can always use a multiple choice test to narrow the initial pool.

How to create work samples

Decide on skills

Work simulations are most effective when tied to skills.

You’ll need to decide on 6-8 core skills required for the role you’re hiring for.

These can be a mix of soft and technical skills.

By attributing 1-3 of these skills to each work simulation question, you’ll be able to build a clear skills profile for each candidate.

Hiring for a pre-determined skill set also makes it easier to be objective, since candidates are being tested against criteria, rather than using ‘gut instinct’ (which is usually just unconscious bias).

Use realistic scenarios 

The next step is to think of 3-5 scenarios or tasks that would test at least one of the skills you just outlined.

Work simulations of any kind work best when specific to your organization and the role you’re hiring for.

If the role is one that has been vacated, you can use tasks that have cropped up in the past.

And if it’s a brand new role, you can simply present candidates with tasks or challenges that they’ll need to tackle, should they get the job.

Your work simulation questions can consist of both everyday tasks and bigger projects/problems that’ll need to be tackled.

For these larger issues, you can ask candidates to explain how they’d approach them or to draw up a quarterly plan of attack.

Pose questions hypothetically

Once you’ve come up with your scenarios, you can turn them into work simulations by asking candidates what they would do.

For your screening questions, candidates will be limited to written answers, but this doesn't mean you have to test writing skills.

Be sure to communicate whether or not candidates need to worry about grammar, punctuation etc.

For most roles, these skills won't be relevant and so candidates could answer using bullet points.

For bigger, more strategic tasks, you can ask candidates to explain how they'd go about tackling it or to lay out their action plan.

When you get to the interview stage, you'll be able to role play scenarios (presentations, calls, meetings etc).


Here’s an example of a work sample we used for an Account Manager role:

Question

You've been given 50 accounts to manage, ranging from newly onboarded users to long term customers of Applied. They vary in terms of size, industry and knowledge of the Applied platform. Your manager asks you to come up with a plan for how you will focus your time to maximise the growth of your accounts over the next 6 months. What things would you consider in putting this plan together? How would you measure your success? Is there any other information you would need?

Skills tested

Accountability, Prioritisation, Data-driven

And for a Community Lead role:

Question:

We have 3000 people in our community. We have grown it by bi-monthly webinars, social content, a LinkedIn group and giving free resources out. The rate of growth using these methods is slowing. 

How do to speed up the rate of community growth to make sure it's primarily in the US? (imagine we are in a purely digital context)

Skills tested

Collaboration, Community Management Knowledge, Data-driven

Asking questions that are forward-looking means that candidates who haven’t encountered the given task before are given an equal chance.

Although experience can make someone the best candidate, it may actually be an outsider perspective that yields the best answer. 

The value of experience is that it forges skills.

And job simulations enable you to test for these skills without making any assumptions.

Traditional interview Qs to Work simulations

Create scoring criteria 

To data-proof your work simulations, you’ll need to jot down scoring criteria for each of your questions.

This ensures that candidates are being scored against skills, rather than a hirer’s preferences (and biases).

At Applied, we use a 1-5 star scale.


Below is the review guide we created for the Community Lead work sample above:

1 Star 

- Little or no effort

- Maybe 1 or 2 ideas that don't sound great or expounded with no depth or enthusiasm

3 Star 

- Comes up with a couple of original suggestions that sound interesting 

- Discussed methods to spread for our existing content which show a plausible understanding of different marketing channels

- Refer to how they might collaborate with the rest of the team

5 Star 

- Suggests an original idea not covered in the above examples, which sounds really compelling

- Suggests forms of growth that consider virality, not just our own hard graft or marketing spend

- Talks confidentially about different marketing channels

- Talks confidently about this in a US context 

- Shows awareness of who we want to be in this community, although not explicitly talking about our personas, at least hints that all community members are not equal.

- Has a strong sense of how other members of the team can be utilised for content and/or distribution, ideally across different functions


As you can see, we just added a few bullet points detailed what a good, mediocre and poor answer might include.

Here's what this looks like in the Applied platform...

Applied reviewer screen (screenshot)

How and when to use work simulations 

Job simulation assessments are generally used later on in the process, either before or after interviews.

However, this won’t bring dramatic results, since most talent is missed at the screening stage.

That’s why here at Applied, we use work samples as our screening method.

No CVs, no cover letters, no bias.

By using 3-5 work simulation questions to anonymously screen candidates, we’re able to identify who has the skills to actually do the job, without knowing anything about where candidates have come from.

Using this process, we’ve found that 60% of candidates hired would’ve been missed by a traditional CV screening - many of whom are from minority backgrounds.

To ensure bias is mitigated as much as possible, we have three team members review candidates’ answers.

If an individual reviewer has a bias towards or against a candidate, this will be averaged out by the scores of the other reviewers.

We’ve found three to be the ideal number - any more and you’ll see diminishing returns.

Once all answers are scored, we then average out candidates’ scores and bring the highest scorers through to the interview stage.

Do you need special software for pre-hire simulations?

The short answer here is no... you don't need to fork out for any specific software. Work simulations can be used completely independently of platforms like Applied.

Creating and scoring work samples can all be done manually.

What software like ours will do, however, is make this process much faster and more data-proofed.

If you're dealing with higher volumes of applications (anywhere from 20+), sending, scoring and scheduling will all take time.

Although there is nothing stopping you from adding candiate scores to a spreadsheet to make decisions and track thing like candiate quality and diversity, this would be a seriously laborious process to undertake manually.

This is where hiring platforms will be useful. Applied, for example, will automate and anonymize the assessment process, with a library of work samples read-to-use and all the data you need to track and report on hiring metrics.

Applied platform

Here’s what you can expect from using Applied:

  • Up to 4x attraction and selection of ethnically diverse candidates
  • 3x as many suitable candidates
  • 66% reduction in time spent hiring
  • 93% retention rate after one year
  • 9/10 average candidate experience rating

Applied is the essential platform for debiased hiring. Purpose-built to make hiring empirical and ethical, our platform uses anonymised applications and skill-based assessments to find talent that would otherwise have been overlooked.

Push back against conventional hiring wisdom with a smarter solution: book in a demo