Effects of response sequence length on motor programming: a chronometric analysis.