Menu depth vs. breadth

My paper Workload Analysis of Menu Layering in Unmanned Aircraft Systems Control Stations looks at how deeply buried menus add to operator workload. It cites the Department of Defense design standard's guidance that a menu should not offer more than about ten options and that a deep hierarchy should not take more than four steps. Those two limits pull against each other: fewer options per menu means more levels. Try it.

Every command or display the control station offers.
How many choices each menu shows at once.
Steps to reach a function
Options per menu
Estimated selection time
per function, practiced operator

How the numbers are worked out

Steps = the number of menu levels needed to give every function its own path (breadth raised to the depth must reach the number of functions). Selection time uses the Hick-Hyman law, a standard human-factors estimate: each level costs about 0.2 s + 0.15 s × log2(options + 1) to decide, plus 0.4 s to move and select. These are textbook teaching values, not measurements from the paper; real stations add reading, scrolling and errors.

The point of the paper stands without the arithmetic: put the functions an operator needs under time pressure at the top, keep the rest shallow, and test the layout with real operators.