Standard "out-of-the-box" personality profiles work great at helping build teams. Many do a tremendous job at describing a person and how they may relate to others.
But... are they effective at predicting how well a person will perform in a particular job? The answer is yes..... and no. It depends completely on whether the profile has been developed specifically with the job's needs in mind. Or, if the profile uses a set of questions that measure the same set of characteristics no matter what the position.
Over the 23 years that I've been working on developing tools to accurately predict performance of individuals, I've seen a lot of both types of personality profiles. Ones that purport to predict performance, and ones that actually do predict performance.
What's The Difference Between Predicting And Non-Predicting?
"Out-of-the-box" personality profiles simply describe the individual's personality. And that's usually where they excel - describing someone's personality. There are a lot of companies that use them for this purpose and they're very happy with their choice of profile. But that's because the profile is being used for what it’s designed to do.
Another thing that "out-of-the-box" profiles can do very well is help you make sure someone fits into your existing team. But even here you need to be careful.
Generally, these types of profiles are legally defensible for describing personality but not for predicting performance.
Performance predicting profiles on the other hand are job-specific. When they’re designed, the developers tap in to the knowledge of Subject Matter Experts, like franchisors in the case of franchisee performance, or QSR restaurant managers for QSR staff performance. Working with the SME's helps identify which characteristics and work behaviors are common in top performers.
From this, either personality “dimensions”, with their attendant personality questions, are developed, or, a new set of “dimensions” are developed along with non-personality type questions developed specifically for these “dimensions”.
Most predictive profiles contain a combination of personality and non-personality items and have been developed using a strict scientific set of procedures to confirm that the characteristics and job behaviors do predict performance.
Properly developed predictive assessments are legally defensible when used to predict performance but not for describing an individual’s personality.
Buyer Beware
Before considering any profiling tool, a decision must be made about whether you’re looking for a descriptive tool, or a predictive tool.
Here’s why:
Let's say you're hiring an outside salesperson. Do you want to hire someone that fits in to the team? Or do you want someone that will perform? It's very important to answer this question because high performing sales-people work best alone and not part a team. They're simply too entrepreneurial and want to "go it alone".
A similar situation exists when selecting retail franchisees. In this case you do want someone that will work well with head office, but that's only a small part of being successful as a franchisee.
If you want to predict the performance of a franchise candidate, you’ll need to consider additional job-specific behaviors and attitudes. Things like:
- Their work ethic,
- How comfortable with local store marketing and getting on the shop floor with their staff to sell,
- Their management style and how they treat employees,
- Whether they have a positive attitude,
- How they handle customer service issues and
- Will they get stressed dealing with a lot of people on a daily basis? Or will they welcome the opportunity?
That’s why, at the very beginning of your research into which profiles to consider, it’s so important to decide whether you prefer:
- Someone that needs to feel they fit into your team, or
- Someone that can go out and build their own team to suit their management style?
It’s just as important to make sure you don't use a profile developed to predict performance of franchisee candidates for any other purpose or position.
Using our business as an example, we develop our predictive assessment tools specifically for a particular position, whether it be franchisees, retail store managers, service staff, etc.. The only people and performance data in our database for the appropriate positions. Naturally, since our database does not contain any data on the general population, our assessment tools can only predict performance.
When a company is evaluating one of our performance predictors, like the FranchiZe Profile, sometimes they make the mistake of evaluating the effectiveness of it by having several head office managers complete the questionnaire. Then they wonder why the profile does not give a good “read” on the person. What they’re doing is comparing apples and oranges. They’re trying to evaluate the effectiveness of our predictions by looking at how well we describe their personalities, Only after they follow my recommendation to have some existing franchisees complete the questionnaire, do they find out how accurate our predictions are.
Choose Based On Your Needs
Many "out-of-the-box" profiles are unfortunately sold as "predictive" rather than at what they excel. Describing a personality. Instead, they’re sold as profiles “customized” for the position. The problem with that is that only the wording of the report has been slightly changed to make it appear to be "customized" when it’s really not. For instance, I've seen reports that simply substitute the word "franchisee" for "sales person" or "manager". All other wording in the report stays the same.
One of the reasons I’m suggesting caution so strongly about using "out-of-the-box" profiles for predictive purposes, is the fact that more and more researchers have been finding there are limitations to using them.
As an example, from PERSONNEL PSYCHOLOGY 2007, 60, pages 683-729, "Reconsidering The Use Of Personality Tests In Personnel Selection Contexts", when the panel was asked "What Recommendations Would You Give About the Use of Personality Tests in Selection Contexts?"
Neal Schmitt responded: "I would make three specific recommendations. First, avoid published personality measures in almost all instances. The only time I would use them is if they are directly linked in a face-valid way to some outcome. Second, I would construct my own measures that are linked directly to job tasks in a face-valid or relevant fashion. Now, you might say that is when you are actually going to get faking. Fine. If you think that is going to be the case, I would include a carelessness or faking scale that detects individuals who are careless or those who are obviously lying or faking."
A copy of this paper is available from our website at: http://www.dynamicperformancesystems.com/personality.
Meanwhile, a properly developed predictive selection tool can predict performance with great accuracy. I've been told by several different franchisors of their experiences of replacing poor performing franchisees with others identified by our FranchiZe Profile as superior performers. They found that simply by replacing the existing franchisee, the poor performing location was able to show a sales increase of over 40% almost immediately. Since similar increases were seen by different franchisors in different market segments, and even in different countries, it demonstrates how effective predictive selection tools can be.
Using a performance predicting assessment for selecting QSR staff should lead to much lower turnover rates, happier employees and managers that would gladly rehire the individual.
Evaluating Effectiveness
If you're given the opportunity to try out a profile, no matter if it’s a descriptive personality profile or a predictive profile, it’s very important to get a good sample to evaluate its effectiveness. Have several people in the relevant position complete the questionnaire. The more people you include, the better read you'll get on how effective the profile is.
In the case of a descriptive profile look for how well the profile will help you in building a team.
If you’re considering a predictive profile, choose a sample of good performers, average performers and some not-so-good performers. Then compare their results to their on-the-job performance.
That's why we offer an unlimited usage, no-charge or obligation, 30-day trial account. Before you make a decision as important as this one, the more information you can gather the better the more likely you'll find the best solution for your needs.
If You Want To Predict Performance
When you're shopping for a profiling tool that will help you predict the performance of a candidate, ask the salesperson the following:
- Does the vendor also sell profiles for other positions? Then,
- Ask to see the questions used for each position and
- Ask to see proof that the profile(s) have been independently validated which confirms that the profile does predict performance for the position.
The use of "independent third parties" (point 3 above) is very important for the follow-up studies as it eliminates the possibility of highly selective interpretations of scientific data. I've seen cases where a study didn't yield the results the developer wanted. So the person doing the analysis simply chose to look at the results in a far more "sales-friendly" way.
In addition, studies done by independent third parties will help identify if there are any changes needed and if the predictions can be improved upon.
If you'd like to see a sample of our methodology when developing a new performance predicting assessment, go to: http://www.dynamicperformancesystems.com/Development-Methodology and we'll email you a copy.
To sum up, a decision must be made about whether you're looking for a personality descriptor to help you in team-building, or if you want to predict the performance of candidates in particular jobs.
If you decide you want something that predicts performance, look for one that:
- Has been developed in consultations with subject matter experts,
- Uses questions developed specifically for the position in question, and
- Is scientifically validated to make sure it predicts performance.
And ask for proof on points 2 and 3. If you're familiar with the position for which you're considering using the profile, point 1 will be apparent from the questions on the questionnaire.
If you'd like to discuss this further, e-mail me at fred@dynamicperformancesystems.com.
Otherwise, we'll connect again in the next issue.
See you then!
Post new comment