Does anyone know of any research/studies that examine the cost to the business of not doing end user training?

There are three perspectives that interest me
1. Diminished effect on individual productivity arising from not training effectively to use technology & the cost of that - and also average time (& therefore money) spent trying to find out "how to"

2. The "Hey Joe" factor - ie if someone doesn't know how to do something how many people on average do they ask for help and what is the average total time involved (& cost)

3. Effect on calls to the helpdesk - ie if people are trained, hypothetically should reduce the advice "how do I"calls to service centre leaving then free to concentrate on the "it doesn't work" kind of calls, which may eventually enable each analyst to manage more users - again generating a cost saving

Beverley Ashton