Great expectations: some suggestions for applying the balanced placebo design to nicotine and smoking

There are a number of methodological problems with the traditional placebo-controlled design used to investigate the effects of nicotine in human subjects. For example, the experimental instructions may create a 'guessing set' in the participants such that they search for possible clues to try to identify which condition they are in. If they guess correctly, the internal validity of the design may be threatened. Furthermore, by attempting to control for the effect of subjects' expectancies, the traditional placebo design precludes any estimate of expectancy effects. Continued use of this methodology in the smoking field is likely to lead to an overemphasis on pharmacological factors in smoking and a corresponding underemphasis on cognitive factors and nicotine-expectancy interactions. These problems can be overcome by the use of the balanced placebo design which orthogonally manipulates instructions (Told Nicotine versus Told Placebo) and nicotine (Received Nicotine versus Received Placebo). The balanced placebo design offers a powerful methodology for studying many aspects of smoking behaviour and nicotine effects, and three examples are given to illustrate its potential in this field: nicotine titration, nicotine and performance, and priming effects in smoking relapse.