[CONTINUED FROM PART 1]That depends on your circumstances. TheAppliedPsychologist has only used the Task Analysis Inventory variety of the Job Analysis Questionnaire. This format asks survey respondents to rate every task data collectors observed when they watched incumbents on duty. (You did have observers watch incumbents on the job before you created your survey, right? Excellent. We knew you wouldn't dare leave out such a critical step unless forced to at gunpoint.)
A Task Analysis Inventory will make empirical data analysis easier than most alternatives will. If used, it follows that your survey will have to incorporate job analysis rating scales into each item. We’ve all seen rating scales before:
“Please rate how often you perform this task on the job on a scale of 1 to 5, where 1 = ‘I would never perform this task even if I spent million years doing this job’ and 5 = ‘I await the sweet release of oblivion to save me from the soul-crushing, repetitive frequency of this task’.”
Finally, review the survey with your permanent project steering group before sending it out. But before you send it out...
7. Let the employees know what’s up – again. Decide (with the aid of your project steering committee) who in your organization has the most authority over the incumbent group. Get that individual’s office to spread the word to their employees about the pending job survey they will be asked to complete. This “person in charge” doesn’t have to write the notice, but they need to send it and stand behind it. This won’t guarantee full cooperation – nothing does. But it’ll help. The notification should assert that universal participation is *very important* to the success of the company. See how that works? ("I find your lack of participation...disturbing.")
8. Distribute your Job Analysis Survey, and carefully monitor the rate of return. Next comes the make-or-break stage: distribute your job analysis survey according to the detailed distribution plan and timetable – the one I’m sure you (or your consultants) included in your stupendous, bulletproof project plan.
"But wait," you ask. "Didn't I just spend a whole lot of time/money observing these people? Why do I have to spend even more time/money/effort to distribute a survey to them?"
For a few reasons, smartypants. First, unless the employee population is very small, there's no way you got to observe all of them on the job. You only observed some of them. But you'd like to get data from all of them. You made those observations in order to figure out what the job entailed. Now you're getting more specific: the incumbents can tell you how much time they spend doing each task, how important each task is, and what level the task must be performed at to be acceptable. If the survey is designed correctly, it will even permit a respondent to indicate that they do not perform an included task at all.
Employees tend to supply this information most honestly when they don't have someone looking over their shoulder that could be perceived as a threat. It's called anonymity. Depending on the circumstances, anonymous responses may either encourage participation (by encouraging honesty and decreasing fears of reprisal), or discourage participation (through the tragedy of the commons). Overall, the pros of anonymous data outweigh the cons…and hiring a competent third party for this phase will improve the odds that your confidentiality promises will be taken seriously. Using a third party also enables consultants to collect identifying data, but keep that info away from snooping employers.
We aren’t exaggerating: aside from the communications pieces, this is the single most important phase of the process. The data you fail to collect will make or break the entire process. Spare no effort to ensure that you get back as many completed surveys as possible from as many different types of people and locations as possible. Hell, offer a one-time meaningful incentive if you have to. But don’t expect a rate of return exceeding 15% to 20% when dealing with a large employee population…and you may only hit those numbers you’re lucky. Fortunately, the principles of sampling and statistics can overcome a low return rate as long as your sample is diverse enough.
9. Analyze survey responses and finalize the job analysis. There are many different strategies to analyze collected job analysis data, but we are going to focus on the principles behind the analysis. First of all, you should gather a team of subject matter experts (SMEs) to go over the results. 10 to 30 of them should suffice, depending on the size of the employee population being analyzed. Diversity in this group is as crucial as diversity amongst the recipients of the job analysis survey. This group (guided by you and/or your consultants) will go through the responses and distill the survey results down to the essence of the job. The two most important questions for this group are:
1. What tasks are the most important tasks out of all the included survey items? 2. Are these tasks accomplished successfully through on-the-job training, or are employees expected to walk in on Day 1 and perform them successfully?
This is an extremely treacherous phase of the project: the nexus where good data can move into a world of awful outcomes for your organization if you’re not careful. SMEs are usually chosen because they are (or were) good at their jobs. That means they are often opinionated and stubborn about how their job should be done, and have inflated ideas about the KSAs employees need to apply to the role as soon as they walk in the door. Some of these expectations are realistic, but others are either way overboard or no longer apply to the modern conditions of the job.
For instance: while an employee working as a cashier may have made complex calculations during each transaction 35 years ago, it is unlikely such a skill is required today given the advent and ubiquity of computers and calculators. And do your admins really need to know shorthand? How likely are you to find the person you end up profiling? And what is the likelihood that your idealized candidate will take the job offered by YOUR organization, rather than another job? Such questions must be taken into consideration during this process. One effective way to navigate these oceans is to tie each KSA to a specific job task. This practice will often provide a useful filter. You can get even more specific by asking the group to rate the degree to which a particular KSA will help an employee complete a listed task.
Gird your loins, for these conversations may get heated. It is your job to keep everyone focused on the goal and moving forward. Know your role, but don’t shut your mouth if you sense the group is getting bogged down or is inflating the job. At the same time, don’t be condescending: if you have an honest question, voice it without making it sound like you know more than they do. They are the experts, not you.
When you’re done, you should be left with the critical KSAs required to perform the job of _______, listed in order of importance. For entry-level positions, this list should contain far more “raw material” KSAs, rather than “specific knowledge/skill” KSAs…unless you don’t have a training program, or you are hiring a skilled craftsman (carpenter, mechanic, plumber, etc).
************************************************************************
If this sounds a lot more difficult and time-consuming than you initially thought, that’s because it is. Performing a successful job analysis involves just as much art as it does science. Experienced job analysts are a lot more likely to do a good job than inexperienced analysts are – and that’s how consultants stay in business. There’s no shame or waste in soliciting experts, but you should be able to tell whether or not they are doing things right.
Renegade Psychology believes that knowledge is power. Even if you aren’t in a position to perform all this work, knowing the principles should at least go a long way towards helping your organization avoid some of the most obvious job analysis pitfalls. And if you can convince your boss to read this, you might even be able to sway him or her to hire experts for this work. In the long run, it will save your team a ton of time…and time is money.