We base our core employee surveys around an outcome measure known as Employee Engagement (EE) because this concept has brought together popularized, academic and pragmatic approaches so well. We wanted to build a new platform for people analytics but why not build on the research already available? There is a lot of research and data out there on this! There is plenty of research linking EE type metrics to important outcomes. In summary, although there are a range of different ways to measure EE, in its various forms it has been shown to have links with business outcomes such as profit, customer satisfaction and employee retention rates. If you want to read about some of the research check out this great set of resources (also see our article on 20 employee engagement questions every company should ask). Although there are a lot of definitions of EE we have based our approach on a few key points of convergence in psychological methods, academic literature and what we perceived to be good practice amongst organizational psychologists. Here are the guiding principles that lead us to our approach and our specific surveys:

There is convergence in the academic literature and academic reviews that EE represents an important psychological outcome or state measure. By outcome measure we mean to distinguish EE as a psychological state that is the outcome of a range of other influencing factors rather than a measure that includes the potential influences. We believe this is important because in our research we are aiming to understand what has the biggest impact on how employees feel and act overall. To do the best analysis of this we need to keep the overall outcome measure as separate as possible from the factors or things we think will impact that outcome. If we want to assess how happy employees are at work we would not create a measure that includes and mixes in how happy employees are with pay, leadership, resources, management etc. because what we're trying to understand is how each of those factors affects overall levels of happiness. We'll discuss further below how EE is different from just being happy at work but the example illustrates the general point.

This is quite an accepted principle in psychological measurement and actually quite easy to explain too. If we are trying to measure something that is a little complex and hard to measure one of the best things we can do is take multiple measurements and average them out. This reduces the error in our measurement by balancing out any particularly wrong measurements with other measurements. Additionally, if we think that the thing we're measuring is complex then we should make our measure reflect some of that complexity so we don't limit our measure unduly. We can also see quite intuitively how a multiple question measure will help differentiate people better too. Using one question on a 1-5 scale will give us a score from 1-5 - only 5 possible scores. If we combine five similar questions into an overall Employee Engagement Index score we get scores ranging from 5-25 - a much finer grained measure.

The principle behind using a multiple question index is that you are combining different measures of the same underlying psychological state. That means that each of the questions is tapping into a slightly different part of the same thing and once you have a few measures of that you are making diminishing returns with each additional question. We began by looking at what researchers have identified as some of the key components agreed upon across multiple sources (i.e., for an excellent review of the engagement construct see here). Based on this we then asked ourselves and clients what were the key five outcomes we wanted to start with as part of EE - we picked 5 because we wanted to have a reliable index but also keep our survey short. Here are the five components we arrived at (click here for more detail on the items associated with each):

We are not making a claim that there are not other good EE questions that could be included. The important thing is that we have a good group of representative questions because this means that additional questions are likely to be highly related anyway. Adding highly related questions won't change things much from a statistical or practical point of view. We are model agnostic so we're not selling you our model but giving you a great place to start - you'll also have access to benchmarks for our standard questions. And, if you'd like to understand why we use a five point agreement scale you can read about our rationale here.

Over the past four years we have used our EE questions with millions of respondents and each year we pull all our data together for benchmark research as well as conducting case studies. Our EE Index typically has Cronbach Alpha statistics ranging between .80 to .90 and factor analyses have repeatedly shown the index to reflect a strong core construct underlying the questions (with the retention questions occasionally reflecting a very highly correlated sub factor). Our EE Index questions have also shown validating relationships with other external metrics such as Glassdoor ratings and Mattermark Growth scores (research soon to be released). The external validity of our EE questions and driver (or impact) questions (see below for more details on these) has also been demonstrated in numerous case studies with our clients via linkage analyses.

With our EE Index in place we are in a good position to measure overall levels of this in a company. However, although knowing that you have high, low or average levels of EE might be interesting, the important thing is to understand more about what you can do to improve your workplace right? This is traditionally where survey length can get out of hand - people can start to feel that they need to ask about everything they can think of in painstaking detail. We've developed some strategies for dealing with that tendency that we've highlighted here. Our core Employee Engagement survey consists of an additional 45 questions covering well known factors that impact EE as well as a smattering of factors that may be unique drivers in some companies. We also ask a few questions on Company Performance perceptions as a way to crowd source on the ground People Intelligence. Research has shown that these 'performance-centric' type questions are often linked to other business metrics such as customer satisfaction and loyalty.

💬 Need help? Just reply with "Ask a Person" in a support conversation to speak with a Product Support Specialist.