Reconsidering the encoding of data in physics education research