* People trained mostly in camelCase (also, nonprogrammers) read camelCase more accurately, but snake_case faster. (Previous study which I didn't follow up.)
* People trained mostly in snake_case read both of them equally accurately, but snake_case faster. (But, grand total of fifteen participants.)
(I find this mildly surprising: I seem to read snake_case slightly slower than normal text, and camelCase slightly faster, just based on the pauses between words. But camelCase would be easier to lose myself in. I find hyphenated-words easiest to read, but not many languages permit them...)
thisIsAnInstanceOfMyClass
this_is_some_ruby_or_kernel_c_i_copied_from_stack_exchange
_thisIsAFunctionParameterHavingBeenPassedIn
THISISACONSTANTDONOTCHANGE
THIS_IS_A_FRIENDLIER_ENUM
_____how_many_spaces__in_this_var
EDIT: Aaaand then there is the JS approach: