I am employed in a company where some workers are designed to enter data into system.
They receive the data as pdf files on email. The quality is not great therefore OCR is not really an option (it's actually received through fax system). Data is five pages with tables, usually only few cells are filled although sometimes there are tens of numbers. They have to find the corresponding cell and enter the number.
What they do is - print out everything and then type from papers into the software.
This amounts to about 400 pages of paper per month. The work is done by three people with minimal training. There's no formal quality control process, but we do get feedback when one of these orders is served incorrectly.
What we (resource department) suggest is to place email and the program (the window for a table is as small as textboxes here) side by side on the screen and NOT print hundreds of papers where most of the cells are empty. As data entry clerks they have sufficiently large screens (22 or 24 inches).
They object that it's more precise from paper. We think that side by side comparison should actually be better for precision. But is there any proof of this?
Could you provide any sources, some research done about data entry quality and speed comparing entry from paper and entry from the same screen or from screen to screen?