Transfer of baseline reject control to transitivity trials and its effect on equivalence class formation