Psychology by Machine? Not for a While.

Psychology button on computer where "Enter" key should be

Technology can fly planes, drive cars; heck, virtually perform remote surgery (pun, not intended). Some believe that literally all jobs will eventually be performed by technology. For them, if a “machine” isn’t already doing it, just wait. (Note: This is an extreme view).

Technology is changing the world faster than ever. If you agree with Moore’s law, it will only continue to increase its impact even faster over time.

Will technology take my job?

Probably so. But don’t quit yet! If you’ve been around a few years, like I have, it’s likely that technology has already “taken” all or much of the job you had 10 years ago. You’ve simply changed to stay in front of the technological evolution.

What does science say?

A recent study looked at the rise of technology in relation to the probability of it overtaking more than 700 jobs catalogued in O*Net, a public database of jobs and the various knowledge, skills and abilities required for their performance. The researchers (Frye and Osborne, 2013) reasoned that the probability of technology overtaking a given job is closely related to the time it will take for this to occur. As such, they created a list rank ordering the probability that these 700 jobs will be overtaken by technology in 20 years.

The study is now a few years old, but seems to have already made some accurate predictions. For example, you’ve probably received a “robocall”, a task once was performed by a person.

The crux of the study is in the researchers’ identification of three key job characteristics they refer to as “bottlenecks to computerization.” The degree to which a job encompasses one or more of these “bottlenecks” predicts the probability (and time) required for technology to be able to perform that job. These three bottlenecks include: 1) Fine Perception and Manipulation, 2) Creative Intelligence and 3) Social Intelligence.

The three bottlenecks were further broken down into seven more discreet tasks. Of these seven tasks, Social Intelligence encompasses a majority of four.

The practical implication is that if your job requires you to “read” people or influence them, particularly in emotional ways, you’re likely safe from seeing a robot at your desk one morning anytime soon.

Specifically, the study predicts that social workers, therapists and teachers should have relatively long careers as far as “automation threat” is concerned. Psychologist, is also in the top 20 of the 700 jobs ranked according to the difficulty of automation.

Although this research is new, the issue isn’t. Psychological assessment has long been a topic of technological debate: Can a personality assessment alone more accurately predict behavior than an expert in psychological assessment?

Continue reading “Psychology by Machine? Not for a While.”

Why are you hiring “can do” when “will do” is what you’ll get?

Superhero or Superzero? True performance is less obvious to predict than you think.

Research confirms what most already know: When it comes to job performance, there’s a real difference between “can do” and “will do.” But how much does "can do vs will do" really influence the way you are hiring? Which person do you think shows up for your interview? Is it possible that we have a significant problem of hiring for “can do” when what we’ll eventually get is “will do?”

In scientific literature this is referred to as “Maximum” vs “Typical” performance. One of the most clever studies published illustrating the difference between Maximum and Typical performance came about partly by accident. The researchers, led by Paul Sackett, were interested in designing a valid way to assess and hire cashiers in a retail grocery store. For simplicity’s sake, I take the liberty of referencing the actual study from a story-teller’s perspective instead of a researcher’s. (Researchers, please cut me a little slack)

A given number of cashiers (say, 10) were instructed to do their best in terms of speed and accuracy when “checking out” the exact same lot of groceries. They knew they were being assessed; heck, their “supervisor” stood over them with a stopwatch! After all 10 of the cashiers had completed checking out the same items, these ‘subjects’ were thanked, business went about as usual and the study was concluded.

Or so they thought.

Later, unbeknownst to these cashiers, they were measured again for their speed and accuracy under identical conditions to the original assessment, except this time they did not know they were being assessed.

Once completed, the results from the first measurement (Maximum Performance) were compared with the results of the second measurement (Typical Performance). What would you guess the correlation to be between the two?

Almost zero!

The study found that the correlation, or relationship, between what people can do vs what they will do was a mere .14! Interestingly, this is about the same as the correlation between the typical selection interview and job performance. Coincidence?

Stop for a moment and consider the implications of this simple, but profound study. Essentially, there is very little in common between what a person “can do” versus what they typically “will do.”

So I ask again: Which person do you think shows up for your interview? And next, what are the implications for your carefully screened team?

Uh, Houston, ….

Experts in the science of psychology at work are well aware of this challenge. As for myself, I tell my clients that I assess for what people “will do” in addition to what they “can do.” Obviously, this requires special expertise since almost all job applicants want to make the best impression they can. It’s not easy to see through the “Sunday Best” they can maintain for a full day of interviewing or assessment.

However: It is possible for a trained and experienced expert in psychological assessment at work to distinguish the difference between can do and will do. But this is still something that many have a hard time doing – even experts.

So, for your next hire, ask yourself: “Am I really measuring this person for what they can do, or what they will do?” You might just find that an expert in psychological assessment can help.

Psychology at work: It really makes a difference.