Why You Should Allow Job Candidates To Use Google In Screening Tests
Posted on January 22, 2020 by P.K. Maric
Our customers often ask if we allow candidates to use Google (or other resources) while taking a test. The reason they ask is because they want to prevent candidates from cheating. But, disallowing the use of online resources is not a good way to do that. There are cheating protection methods that don’t restrict candidates from using Google. For example, webcam proctoring, copy/paste protection, email verification, and others. But the best way to prevent cheating is to use non-Googleable work-sample questions. This way candidates can still use any online resource they need to, and it’s important to let them do so.
Memorizing Vs Looking Up
The goal of work-sample testing is to mimic a real work environment as much as possible. Which means that the questions we use target pieces of knowledge that the job requires. We also allow candidates to use any online resource, as a real job does.
Programmers, for example, use Google very often. In fact it’s so common to use Google for programming tasks that it’s considered “an essential part of their software development toolkit”. How can you effectively test the skills of a software developer if you don’t allow them to use an essential part of their skill set?
This testing method is similar to open-book exams used in many universities. If you’re unfamiliar with this concept, it may seem like it defeats the whole purpose of an exam because it allows students to cheat by looking up answers. But that’s not the case at all. The point of open-book exams is to evaluate knowledge, not simple memorization. You can’t answer the questions by simply copying material. You need to use higher-order thinking skills to solve a problem or analyze information. These skills are essential for almost every job the requires specialized knowledge.
We had a whole blog post about questions that target higher-order thinking skills. Such questions are far better at evaluating knowledge than trivia questions that rely only on memorization, as those are easy to cheat.
This means that you need to ask questions that are not so simple that a quick web search can answer them. Questions should allow candidates to find the information they need to solve a problem.
Let’s say that you’re interviewing a programmer with some JavaScript knowledge. Asking them a trivia question about a programming concept, such as “What are Closures?”, is not very good. It takes 2 seconds to Google the answer. It’s much better to give them a programming task that requires understanding closures to solve a specific problem, like this one:
You can play with this question here.
If a candidate knows nothing about closures, looking up a definition won’t help. They already need to know how closures work to fix the bugs.
Some candidates might know what they need to do, but get stuck because they haven’t used JavaScript in a while. In this case, they can find what they need to online so that they can complete the task. If they were to work on a similar problem in an actual job, they would likely do the same thing, as most programmers do. To solve a programming problem, you need to know what to look for and how to apply what you find.
Here’s another example from the aviation industry: Pilots are required to go through a preflight checklist before takeoff. It doesn’t matter how experienced the pilots are, or that they’ve memorized the checklist after hundreds of flights. They still use it because it’s a safety issue. If you were hiring a pilot, would you test their skills without letting them use a checklist before takeoff to make sure they know what they’re doing? You shouldn’t, since not using a preflight checklist properly is a major contributor to aircraft accidents. But it’s not just pilots, engineers and other aircrew need to have access to information they need to do their jobs. On June 10, 1990 the windscreen of a British Airways Flight fell out, causing the cabin to decompress and the captain to be pulled halfway out of the plane. Luckily he survived. The problem was that the bolts on the windscreen were replaced with the wrong ones. The engineer doing the maintenance did not refer to the manuals to check which bolts were needed, but simply replaced the old bolts with similar looking ones.
Cheating Concerns
Because of the way questions are designed, candidates looking up answers online is not a major concern. But if you are still worried about cheating, there are ways to prevent it.
Even though we create questions so that you can’t find a complete answer online, all the question text on TestDome has copy/paste protection. This prevents candidates from asking for help on question and answer websites like Quora or StackExchange. Even if they could, it’s extremely unlikely that they’d get a correct reply before the question time expires.
Email verification is also an option, and can confirm that a candidate didn’t already receive an invitation to the same test. This way you can be sure that the candidate hasn’t seen the same questions.
You can use webcam proctoring to verify the identity of the candidate taking the test. The webcam will periodically take snapshots of the candidate and include them in the reports. This way you can ensure that nobody is helping the candidate with the test, or taking it for them.
Another way to prevent or discourage cheating is to have candidates take the test onsite rather than at home. People are far less likely to cheat when they are being directly monitored.
One thing we noticed is that cheating doesn’t happen nearly as often as people think it does. Candidates likely know that the skill test just one step of the interview process. So even if they were to cheat, it wouldn’t help them in the long run since they know they’d need to speak to an experienced person in the interview where it would be clear that their score on the test doesn’t reflect their knowledge. This is one of the reasons why we recommend testing candidates before the interview.
The best way to make cheating inconsequential is to follow up with a live interview at some point after the test. After all, the skill test is not the only part of the interview process, even if you have several testing rounds. If you’re using high validity screening methods and an optimal hiring process, then it’s nearly impossible for someone to cheat effectively without getting caught, and to do so repeatedly throughout the entire screening and interview process.
Conclusion
When it comes to hiring, the goal of a skill test is to predict a candidate’s work performance. To predict work performance as accurately as possible, you should allow candidates to use any tool they would be able to use on the actual job. You don’t need to worry about cheating if you use questions that target job-relevant tasks and you have a well-structured screening and interview process. But there are still cheating protection methods that don’t restrict a candidate’s access to any information they might need.
