There is much talk about whether employers google candidates, whether they should google candidates and whether it is legal to google candidates.
As a work psychologist, whose original speciality was personnel psychology, I know heaps about selection. So I ask the question differently. What are we hoping to gain?
Yes, before I meet someone, I look them up on the internet. Before I go to a job interview, I know more about the interviewers than they realize. But if I were the interviewer, would I look up the candidate? I am afraid not, and this is why.
The art of selection is
- to identify variables on which we vary reliably
- and to identify which of those, correlate with difference in job performance.
To find anything that varies with job performance, we must be first find out what variations there are in job performance that are themselves reliable and meaningful.
Collecting information on people ad nauseum just clutters the process. If you want to use Google and Facebook in selection, you need to show me that what you are looking at is reliable and relevant.
Otherwise you are gossiping. Harsh words, I know, but go back to the numbers.
We deal with weak effects. Generally we deal with effect sizes of around 0.2 or 0.3 and we account for 4-9% of variance on the job. The grandest claims are 25%. Muddy that prediction and you are left with nothing but randomness. Moreover, you are likely to create adverse impact (select on a like-me basis and open your company up to valid charges of discrimination).
In selection, we stick to variables that we know are relevant to job performance and that we can measure reliably. If it cannot be done, then it cannot be done. That is the professional and ethical position.
So, what are we really worried about?
People aren’t daft. So, what is going on when we try to select minutely?
Organizations take people and make them live in closer proximity than if they were married. Anxiety goes up. Is this going to be heaven, or is it going to be hell? And if I am the manager, will I be held accountable for the outcome?
My answer is not to make the selection process more complicated. There is nothing to gain. If something is not predictable, then it is not predictable.
Rather put we should put our energy into managing the relationship.
- Improve the working conditions.
- Improve the job design to set clear boundaries.
- Set up communication systems
- And coach ‘on demand’ (I mean it – be on call).
Above all, attend to why the manager is so anxious. Why do they believe they will be blamed? Most likely because there is no common ethos on what performance it is reasonable to expect. It would be better to work on the collective understanding of what is reasonable and to lower tensions all round.
- We need HR people who understand job performance and what variations are manageable.
- We need HR people who can be close to the work team and help them with the ‘pressure cooker’ existence of living in far too close physical proximity.
- We need HR people who can grow the understanding of what is manageable and what needs to be worked through.
- We need HR people who are credible because they focus our attention successfully on what can be done.
Will candidates look us up on the internet? I do hope so. Will they have Facebook profiles? I do hope so. Will they have a life outside work? I really do hope so.
But that is nothing to do with selection. In selection we deal with what is predictable. If it is not predictable, then exclude it from the selection process!
A place for everything and everything in its place. There is a lot more to HR, management and leadership than selection.
UPDATE: There are two other considerations to make when we use ambient data to make decisions about people. First, remember issues of privacy. If you look, you will record. What did you record and do you have the person’s permission? Second, employment is about a relationship. Build one! Use professionals to do the cold work of making probabilistic predictions on reliable factors. You get going on building a warm, normal and trusting relationship.