Habit lag: when "automatization" is dysfunctional

La Fave's habit lag construct, which specifies conditions under which previously automatized motor responses become disruptive of subsequent performance, was tested. Performance under stress was also examined as a possible factor in the occurrence of habit lag. Following a visual discrimination task, 48 women performed motor responses simultaneously: (a) repeating an invariant lever movement and (b) pushing one of two keys. The habit lag construct was supported, as Ss who had automatized responses produced more errors than nonautomatized Ss. Performance under stress, by calling attention to the risk of habit lag, reduced errors. Habit lag accounts for an intrusive type of error often found in motor performance where negative transfer fails to do so.