Statistical methods in psychiatric research, with special reference on factor analysis

• Main Author: Miettunen, J. (Jouko)
• Format: Doctoral or Postdoctoral Thesis
• Language: English
• Published: University of Oulu 2004
• Subjects: alexithymia; bibliometrics; personality; validity; aleksitymia; bibliometriikka; temperamentti; validiteetti; Northern Finland
• Online Access: http://urn.fi/urn:isbn:9514273672

Abstract This interdisciplinary study describes in the first part the frequency with which various statistical research designs and methods are reported in psychiatric journals, and investigates how the use of these methods affect the visibility of the article in the form of received citations. In the second part focus is specifically on factor analysis, and the study presents two applications of this method. Original research articles (N = 448) from four general psychiatric journals in 1996 were reviewed. The journals were the American Journal of Psychiatry, the Archives of General Psychiatry, the British Journal of Psychiatry and the Nordic Journal of Psychiatry. There were differences in the utilisation of statistical procedures among the journals. The use of statistical methods was not strongly associated with the further utilisation of an article. However, extended description of statistical procedures had a positive effect on the received citations. Factor analysis is a statistical method based on correlations of the variables, which is often used when validity and structure of psychiatric instruments are studied. Exploratory factor analysis is designed to explore underlying latent factors, and in confirmatory factor analysis the aim is to verify the factor structure based on earlier findings in other data sets. Using data from the 31-year follow-up of the Northern Finland 1966 Birth Cohort Study this study aimed to demonstrate the validity and factor structure of scales measuring temperament (Tridimensional Personality Questionnaire, TPQ, and Temperament and Character Inventory, TCI) and alexithymia (20-item Toronto Alexithymia Scale, TAS-20). The results of exploratory factor analysis indicated good performance of the TCI and TPQ, though the results suggested that some developmental work is still needed. Of the two scales, the TCI worked psychometrically better than the TPQ. A confirmatory factor analysis showed that the three-factor model of TAS-20 was in agreement with the Finnish version of the scale. ...