I don't think so.

I don't believe, that different employees would understand the same think under the term "performance" and, even more important, I do not believe that they would undertand the same think as you or the firm.

I believe, that human tend to overestimate themselfes. For example I rember a survey which foudn that a huge majorety believes, that they are above average drivers. But how can be almost everyone above average?

I believe, employees tend to confuse performance with effort.

In my view there is only one way to measure performance and that is to look at the output of the employee, team, line establishment or what ever.

I agree completely, however it is difficult to obtain performance measurement from their output, especially if the output are not readily quantifiable.

I'm thinking what if the performance was rated by other source, for example the boss or the user, would it be applicable? I'm also thinking of performance measurement by customer in service setting, but the problem is to consolidate their measurement and the employee's. Any idea?

Good point. I was thinking about that idea as well. However, I believe this wouldn't work either.

For example, this would require the boss’s rating to be objective. I doubt that this is the case. There might be personal favorites or people they dislike, which might show up (subconsciously) in the performance rating. Bosses might also give upward biased ratings, maybe in order to avoid conflicts or because they believe they are leading the best department of the firm or they think they hired the best people (because each of them thinks that he/she is a better boss than the other). With the boss’s rating you just move the problems of the employee self appraisal to another level.

Customer satisfaction is better, especially if there is a broad and routinely conducted customer survey. However, in my view, the viable performance measure depends on the final goal of the firm. Customer satisfaction is not an end, it is means. In my - probably primitive - view the goal of firms is to compete successfully in the market place and earn profits. Profits, turnover or any measure as close as possible to it allow you to measure performance. Only in very good circumstances you will be able to break them down to single employees (salesmen and sportsmen are common examples). In most cases you have to go with the team, department or production line, and even this requires a close collaboration with the firm (an its controlling department). For many questions one might ask this is not a real problem. So the necessity to go to employee level and the availability of alternative approaches depends on your research question.

Metrics should be considered when evaluating one's performance. Again, the Job Description will be a very good tool in performance evaluation. This shall enumerate the bases specifically and clearly.

Thank you for all your responses.

I have found that employee performance can be measured by questionnaire, however:

1) The performance questionnaire should be filled by the direct supervisor, or

2) The performance questionnaire can be self-reported by the employee WITH validation from direct supervisor.

This is definitely harder than ordinary survey, but the procedure would minimize the self serving bias inherent in the nature of self report surveys.