Lecture Notes
When we get to collecting job requirements information, there are multiple methods and sources that we can use and ideally would use a mix of methods and sources to try to get the best data. So you can review these in your textbook but I will just point out, when it comes to methods, prior information includes Onet. Onet is a great way to get the information without a lot of work. So you'll need to certainly use other sources and methods to get specific to the jobs in your organization after you start from Onets, more occupational level work analysis. I'd point out that with the task questioners this includes the PAQ, the position analysis questionnaire. Your book talks about this for a while and this has a long history and it's been used across many jobs in the United States so there's a lot of data and it provides a standardized way to conduct job analysis in your firm. So that can often be a good starting point. But again multiple methods and sources is kind of the key here. Part of this is because you measure different aspects of performance with different methods and different sources. Of course you have to keep in mind cost and time as you try to find the optimal balance between the methods and sources. I'll point out here as the book does that synthetic validation is something can be used as well in validating job requirements. Validating job requirements would mean seeing how they relate to performance. So the idea here is that experts would rate how different measures measure given job requirements. Okay so how different measures get at or measure the performance on given job requirements. So experts might find that a paper and pencil test can measure intelligence and certain parts of job performance that are closely linked intelligence. But they might find that you also need for certain jobs, a test of eye and hand coordination. And that the intelligence test doesn't get that so in order to figure out the job requirement of how much hand eye coordination one has, for example military job, it might require a physical test of some sort where you're able to ascertain an individual applicants' hand eye coordination. Different methods and synthetic validation though can help rate the expected validity that you're going to get on these things without having to go through a large study to actually collect the data. So the real data would always be better but sometimes you just ask experts who have been in the job, know what works well on the job for employees and these experts can rate and help us determine what tests and what methods would be the best way to get at and to measure certain job requirements.
