Tuesday, February 15, 2005

Searching, searching, ever searching

I was off-line yesterday, serving on a selection committee for an arts-related position in our local community (won't say which). This interview team was the fourth such search and screen process I've been part of in as many months (for positions ranging from public library director to campus arts coordinator to administrative staff), and I was struck again by the strange processes we use to identify, screen, and select administrative leaders in the arts.

A few of these interview processes were close to the norm these days, especially for public institutions: a standard list of questions, read one-by-one by a panel of interviewers -- with varying degrees of ability to follow-up or converse on the topic. The goal here is equity, for each candidate to have the same questions in the same order, and to focus the interview team on the skills they determined were important for the job.

In another process, candidates moved through different 'stations,' intended to simulate actual job functions and assess their performance in those situations -- arguing a budget before a city council, presenting to a Rotary club, responding to an irate e-mail, leading a staff meeting, explaining a policy choice to a board. Each station would have small teams of interviewers pretending to be the audience, and rating the candidate on their actual performance.

The first process -- the series of questions -- is falling out of favor with large corporate recruiters, since it turns out to be a poor predictor of actual performance on the job. The second process is an extreme version of what seems to be taking the place of the traditional interview: the performance-based or behavior-based interview.

According to this summary article:

The premise behind behavioral interviewing is that the most accurate predictor of future performance is past performance in similar situations. Behavioral interviewing, in fact, is said to be 55 percent predictive of future on-the-job behavior, while traditional interviewing is only 10 percent predictive.

In a traditional, skills-based interview process, you're likely to favor candidates that interview well -- which may or may not translate into actual fit with the job. In a behavior-based interview, the goal is to encourage story-telling rather than skills listings. Within those stories lies all sorts of depth about how they think, how they adapt, how they measure success, and how they really work. Of course, you could be favoring good story-tellers, but in the arts, that may not be a bad thing.

There are lots of resources on behavior-based interviews (here's one quick article on how to run one). The process, however, does not resolve one of the key problems of hiring management -- actually defining clearly what it is they are expected to do. Without this clear and cogent description, no process will find you the right person, except by complete accident. According to this article on the process:

The key to success in using behavior-based interviewing is preparing beforehand a list of what skills and behaviors are relevant to the position, she says. If a position does not require teamwork, for example, there's no point in asking a question about the applicant's experiences with teamwork. Another important piece to the technique's success: the qualities identified as necessary for the job must be included in the advertising for the position to attract the right candidates.

According to several articles, those job descriptions should focus on three general types of abilities, skills, or aptitudes:

• Technical: comprising job skills and related knowledge
• Functional: skills that are transferable across jobs, including managing and organizing people or information
• Adaptive: personal characteristics, such as dependability, flexibility, or a strong work ethic

As I learned first-hand from my recent search and selection experiences, the hiring process is fertile ground for all the cognitive biases of any individual decision process, with the added fun of a group of people, mixing their biases into a glorious mess.

Your best bet for reigning in that chaos is to structure the process in the most productive way. The list of skills questions turns out to be several steps better than the other common alternative: the ''lets bring 'em in and talk to 'em about stuff'' approach. But the behavior-based interview process seems to be one interesting attempt to go a few steps further.

posted on Tuesday, February 15, 2005 | permalink