Do you agree that how much you can produce a day is more important than how long you can stay in front of the computer?

It depends on the mental sophistication of the task. For simple tasks such as data entry and management you could sit there all day plugging away. With writing, you only have a limited time over which your brain can function optimally. If you are mentally tired it is not likely you will produce anything of value. The same goes for revising and editing in the latter stages of the PhD.

I agree with Barramack - if it's a job like data entry or transcribing, you just need to put the hours in and get through it, but if you're writing or researching, it's probably better to work well for a few hours than work badly for 12 hours! I keep trying to 'train' myself to do bursts of productive work but it's not easy to move away from the idea that you should be sitting there all day.

The key thing I think is that it's output that counts, not hours done. As the others have said, for some tasks like data entry it can be very time-consuming with little obvious reward. But you have to put those hours in. Whereas writing may not take so much time, and can be done by some people in short spurts.

I had to complete my part-time PhD in no more than 5 hours total a week near the end, for the latter half really, including the writing stage. This was due to an MS-like illness, which also meant I couldn't work on it for more than an hour at a time, and then it would take me days to recover until I could do any more! I got very good at getting a lot of writing done in little time. No messing around, no surfing on the Internet. It was about producing output in concentrated bursts. And if you think along those lines I think it may help to be a more productive scholar.

Thanks guys. very helpful advice.=)