I scrolled through the entries in my RSS reader (I love Google reader- Thanks for the recommendation, Anoop) this morning and noticed an interesting post.
David Zinger, a blogger I follow regularly, posted some data indicating that engagement levels for employed American adults had grown from August 2008 to February 2009. I was mildly surprised at the news, given the rough economic conditions and the measures companies have had to undertake to survive.
Why might we see this increase, asks Zinger. It is an interesting question and my skeptical nature compelled me to reply (check out the comment I left on his board). As I hit “submit”, a few more ideas came to mind, and I thought I should share in the fun of trying to explain this phenomenon.
The obvious answer is that employee engagement scores are increasing because employee engagement levels are increasing. I am not saying that this is unlikely; it is very possible given the amount of very bright people out there helping organizations do just that. However, I am a psychologist and am trained in the business of theory busting. Somehow, I feel better after identifying all the different ways numbers can mislead us; remember Mark Twain’s words, “There are three types of lies: Lies, damned lies and statistics!”
So here are some possibilities to look out for:
1. Sampling: These increases may be specific to the organizations sampled in Zinger’s database. This is always a danger, though we try hard to avoid it. I don’t know the particulars of this database. It is most likely proprietary, much like our own Normative Data, so the details are pretty closely guarded.
Maybe the observed trends are attributable to larger organizations with both better representation in the database as well as larger budgets for employee engagement initiatives.
Perhaps the database features several new organizations at time 2 and these happen to have higher engagement scores.
Both of these are relatively easy to control and I suspect they have been… still, I thought they would be good to mention for any budding researchers out there.
2. Compensatory Factors. I know I have mentioned this before, but it bears repeating: People engage in behaviors that are rewarded. In my time with The Foresight Group, I have assisted a number of organizations incorporate engagement survey results into executive compensation plans. I love this idea on paper, but there is a real danger here. It is way too easy for managers to focus on the engagement survey numbers, rather than the engagement concept.
I think back to those increasingly frequent exchanges with retail clerks, promoting me for a specific response in customer satisfaction surveys. I frequently get asked for a “5;” Given my line of work, I instantly understand they are asking for the most favorable response option. For the most part I don’t mind, except that I occasionally have a poor customer service experience and then I am asked for a favorable score. People should not focus on the score itself, they should focus on executing those behaviors that lead to higher survey scores. Tying compensation to engagement survey results, rather than engagement outcomes can lead to artifactual increases in engagement.
3. Gratitude. The past few weeks have been pretty rough in my side of the world (anyone notice the relative silence on this blog?). I am working on a challenging data analysis project for a cherished client. I love the work and I really like the people on the project team. We occasionally run into a spot of trouble and and I think of all the other stuff I would rather be doing than recording my 60th hour of work this week. Inevitably, I think to myself, “It could be worse. How much does unemployment pay?” That’s usually good for a chuckle, and an “Indeed,” from Paul.
Could this gratitude lead to higher engagement scores? You betcha! In fact, I would be surprised if it did not. I am not suggesting that the entire increase is due to gratitude; things are seldom that easy in psychological measurement. But, I would be willing to wager my egg-laying, red-headed donkey that gratitude is in there somewhere.
So where does that leave us? Is this increase in employee engagement scores a result of increasingly engaged employees, or some less-optimistic cause? My opinions are aptly summarized by Abe Simpson, “A little from column A, a little from column B.”
Lukas sounds like a psychologist!
These are some interesting points. You provide some insightful alternatives to understanding these data. Thanks a bunch!
I’d be skeptical about these data. There’s the fact that we know next to nothing about sampling, collection or analysis methods employed. And, as you mention, there’s the chance that the changing composition of full-time employees has selected out the ‘disengaged’ from the sample.
But beyond that, it sounds from the firm’s press release like most of these changes are actually n.s. The press release mentions twice how the change in their discretionary-effort measure was statistically significant, but never mentions whether any of the other changes were statistically significant. Given how much they crow about their one sig. result, I would be surprised if the change in the other 5 measures wasn’t n.s.
So, if that’s the case — that the only one of these ‘engagement’ measures to have really changed was employees’ willingness to go above and beyond — there’s a different story to tell. It’s not that employees are broadly more engaged (thinking of engagement as basically an amalgam of job attitudes and discretionary effort), but rather that people are simply working harder because they perceive more uncertainty about their job prospects in the companies they work for.