This is an excerpt from Attention and Motor Skill Learning by Gabriele Wulf.
As we discussed in chapter 1, some scientists believe that it is important not to think about the movement while executing it (e.g., Masters, 1992; Maxwell, Masters, & Eves, 2000; Singer, 1985, 1988). For example, Singer suggested that the performer go though the motion mentally before actually performing the movement (step 2 of his Five-Step Approach). Subsequently, the performer is supposed to direct attention to an external cue in order to block out other thoughts; and during movement execution, the performer is instructed not to think about anything. Masters (1992) contends that conscious thought processes should be avoided altogether. Therefore, learners should be given as little instruction as possible. Otherwise, they may try to apply this knowledge and adopt a conscious mode of control, which is assumed to interfere with the automatic execution of the movement. For this reason, in their studies Masters and colleagues often have learners perform secondary tasks, such as random letter generation or tone counting, while practicing a motor skill (e.g., Masters, 1992; Maxwell, Masters, & Eves, 2000). This way, rules or explicit knowledge regarding the skill—which learners might develop if they were not distracted—is supposed to be reduced to a minimum.
Is it possible that the main advantage of adopting an external focus is that attention is not directed to the actual movement? In other words, is the only function of focusing on the movement effect that it blocks out other thoughts, or that it prevents individuals from gaining explicit knowledge about the movement? If this were the case, preventing learners from focusing on their movements by having them perform an attention-demanding secondary task might be just as effective as instructing them to focus on the movement effect. Yet, if there is an advantage to focusing on the movement effect, relative to focusing on another task, the former condition should lead to more effective learning.
Nancy McNevin and I examined this question (Wulf & McNevin, 2003) in a study in which we had participants learn the stabilometer task. One of the practice conditions was designed to prevent learners from focusing on their actions. In this condition, participants were required to continuously shadow, that is, repeat out loud, a story presented to them via a tape recorder while they were balancing on the stabilometer. Shadowing is a rather demanding task, so participants were not able to devote a great deal of attention, if any, to the balance task. We compared the effectiveness of the shadowing condition to that of external and internal focus conditions similar to those used in our previous studies. In the external focus group, participants were instructed to focus on keeping markers on the stabilometer platform horizontal, whereas in the internal focus group they were instructed to focus on keeping their feet horizontal. In addition, we included a control group without any attentional focus instructions.
The results are shown in figure 4.1. A couple of things are noteworthy here. First, the external focus condition was more effective than the shadowing condition. This was the case not only for performance during practice, when one might expect the secondary task requirement to degrade balance performance. Rather, the shadowing group was also less effective than the external focus group during retention, when the shadowing task was removed. This suggests that there were no “hidden” learning advantages of having to perform a distracting task. These results indicate that there is, in fact, an advantage to focusing on the movement effect. Simply “distracting” learners was not enough. Second, the external focus advantage was seen not only in relation to the shadowing group, but also in relation to the internal focus and control groups. As you can see from figure 4.1, there was no difference between the internal focus and control groups in the retention test. This finding is similar to the results of the ski simulator experiment discussed earlier (Wulf, Hö, & Prinz, 1998, Experiment 2). Those findings showed that instructions referring to the actual body movements are rather ineffective—in fact, no more effective than no instructions at all. The important point is that distracting learners from the motor task to be learned was not very effective. Even though it was not detrimental, it clearly didn't benefit learning. In contrast, having learners focus on the movement effect did.