I usually found it slightly easier to teach VOR intercepts 1st. It seems to me to involve less mental integration than the ADF since the VOR provides a solution in terms of 'Where I am in relation to the aid', whereas the ADF is 'Where the aid is in relation to me'.
Once VOR skills are developing then go on to ADF Intercepts.
However, for both these devices, I found it easiest to develop a situational awareness (I know, an overused term, but...) about where they were, where the desired track/radial/bearing is, most expeditious way to get there etc., followed by rote rules that could confirm whether the decision making was correct.
I know this sound long winded, but I think it's a combination of all of the above. I know that when I'm using the instrument then I use a variety of methods depending on the situation.
At first I teach along the lines of:
1. the RMI (DI if RMI is not available) is akin to a god's eye view from above (insert preferred simily here) of the aircraft flying over the earth's surface. The outer edge of the instrument represents the furthest scale of the intercept problem eg currently located at 10 nm would like to intercept outbound at 20 nm --> then the circumference equates to 20 nm.
2. In this sense, the aircraft will ALWAYS be on the tail of the RMI needle, flying UP the instrument panel. If no RMI then use a transposed needle onto the DI. A pencil to represent the ADF needle can help here.
The aid will always be at the centre of the RMI / DI.
ergo, the tail of the needle is aligned with the brg FROM the station, the head of the needle the brg TO the station.
Using a finger (or whatever) point to the current position.
3. Point (on the RMI / DI) to the what track / BRG / radial is required (including whether towards or away from the station). This graphically shows where in relation the current position vs. desired position is.
4. Turn the shortest distance of arc to get there.
Steps 1 to 4 are good for orientation but how to confirm once turned that the solution is correct?
I see that this is where 'rule based' methods are advantagous. eg "the head always falls", "on an inbound intercept the head must fall (on the OPPOSITE side of the ADF instrument to the DI) to the difference between the current heading & the desired track.
Yeah, I know this one in particular SOUNDS long winded - blame English grammar

. It's rather more easier to show
Similar rules are available for other circumstances.
These are there as confirmirmations (or more importantly, as error traps) for the orientation & decision making process.
This all seems rathery lengthy

, but it's difficult to accurately describe what is essentially a demonstrative & pictorial process.
My $0.000002 worth (how DO I make a UK pounds sign using a US keyboard?

)