Review of APA Statistical Guidelines: "Predictive Value of Alexithymia: A Prospective Study in Somatizing Patients"
The August 1999 article in the American Psychologist discusses proper statistical methods and how they should be utilized in journal articles. Using some of the guidelines put forth in the article, I will attempt to show the extent to which Bach & Bach (1995) follow these principles.
The article in question investigates the role of alexithymia as a predictor of persistent somatization. The study was conducted over a two-year period with the consent of thirty patients exhibiting somatoform and anxiety disorders. The personality trait of alexithymia is a disorder in which a person shows an inability to express any emotion that they might be feeling. They also show a limited ability to express participation in fantasy experiences.
It is not made clear exactly what kind of study the researchers are presenting. It is easy to assume, however, that the study is a test of a hypothesis. This made known by the statement of purpose at the end of the introduction. It is said that the purpose is to evaluate the differences in alexithymia scores between different patients over a two-year period and to determine if those scores predict outcome independent from other disorders. So the hypothesis being tested is that high alexithymia scores can predict long term chronicity of a somatoform disorder.
Is the population that the researchers are concerned with made clear? No, not really, the population must be assumed by the subject matter. We are to assume that the study is concerned with people who have a somatoform disorder or anxiety.
The sample is definitely outlined well in the article. The authors make it known to us that the participants for the study were volunteers from a previous study about somatoform disorders. They clearly state the age ranges, sex, and education of all the patients.
For the purposes of this study, nonrandom assignment was used. This was not stated but rather an obvious point. The procedure called for the participants to all receive the same treatment throughout the study. A control group was not used so suspicions must be raised as to the validity of the research.
In terms of variables, the study uses a simple approach to tackle this task. Each patient was grouped by whether or not they exhibited a certain diagnostic disorder at the time of the testing. They were either a yes or a no for five different types of disorders.
As far as questionnaires, these are the only procedural devices used in the study. The authors make it very clear as to what psychometric tests were used. They...