To the Editor.—
In a recent article (Archives 1983;40:220-222) Matuzas and Glass reviewed two studies concerned with the treatment of phobic states published in the same issue of the Archives. The reviewers highlighted many important methodologic points that might have contributed to the difference in conclusions reached by the two research groups; however, two aspects of the planning of the studies were not mentioned and since future improvement of experimental design and hence more reliable findings must take account of these, I wish to draw attention to them. The points concern the duration of the phobic disorder and previous treatments that the patient may have received before the study commenced. In the study by Zitrin et al (Archives 1983;40:125-138,139-145) the mean duration of disorder for the agoraphobic group was more than ten years, and 70% of their patients had received previous treatment. In the study by Marks et al (Archives