There was no Internet in the 1980s outside the laboratory, but many of the ideas and processes being used in Internet learning and assessment today were performed on other kinds of computers in the 1980s. 
This useful book “A handbook of Computer Based Training” describes the state of the art in Computer Based Training in the early 1980s. It was written by Christopher Dean and Quentin Whitlock and published in the UK by Kogan Page and in the US by Nichols Publishing Company. This brief description will hopefully give you enough information to see if it’s worth tracking down a copy to verify prior art.
There are several editions of the book – it was updated into the 1990s, the description below is from the 1984 version (re-printed with some revisions from the 1983 edition). Google books have a searchable copy of the 1983 original edition here.
Much of the book describes technology which has been superseded, but there are two parts which might be interesting from a prior art perspective.
Chapters 1 through 4 of the book are on the design of learning sequences. It describes how one identifies training needs and analyses and breaks down tasks and then develops training objectives – with a common technique being to set up a module with an objective and a post-test. Modules are then combined in a learning plan, which can be individualized or adapted for different learners. There is discussion of pre-tests and entry requirements for a course, and giving people remedial or equalizing training and modular scheduling and branching – with learning adapting to the performance of individual learners.
Chapter 16 covers computer management of instruction, including:
- Use of a network
- Registering courses and students
- Testing and recordkeeping
- Directing the student through the course (based on topic scores)
- Course maintenance
- Reporting (with lots of example reports)
- Running a CML system
You can search through the book on Google here.
