Scoring AA and BB trials

Gavin-Perry / OdorChoice

Pico Odor Choice for Camillo & Ale at WUSTL

0 stars 0 forks source link

Scoring AA and BB trials #46

Closed Gavin-Perry closed 3 weeks ago

Gavin-Perry commented 4 weeks ago

For AA or BB the larger must be chosen for a reward to be given. Is it stronger odor (OdorL > OdorR) or more drops (RewA < RewB) that is the deciding factor. I realize that in order to make sense they should be the same thing. If a weaker odor gave more drops that would be very confusing; but the software needs to know which to use to make the choice.

alessli commented 4 weeks ago

Agree. Quantity and concentration are monotonically related, therefore within AA or BB trials, the highest concentration is a proxy for the correct response.

Gavin-Perry commented 3 weeks ago

Related problem for the graph

For AA trial there is no no B choice so that column should never score as B choice or does the correct bigger A count as a B on the graph For BB trial smaller B is wrong so scores as NOT B on graph?

alessli commented 3 weeks ago

For AA trials, lower A scores as B. For BB trials, lower B scores as A.

Gavin-Perry commented 3 weeks ago

Lower AA scores as B, so he should pick the smaller amount for it to be B? In other word A is the preferred choice not B as B would be a fail right? That's weird even if it makes some theoretical sense.

alessli commented 3 weeks ago

Scoring is good in v31. For reference:

in trials AA and AB0, B in the graph is a proxy for mistakes.
in trials BB and A0B, A in the graph is a proxy for mistakes.

Gavin-Perry commented 2 weeks ago

Got it. It makes sense now