Extending the reach of mousetracking in numerical cognition: a comment on Fischer and Hartmann (2014)