A CROSS-VALIDATION OF THE FACTOR STRUCTURE OF CHRONOTYPOLOGICAL QUESTIONNAIRES

Dmitry A, Putilov

Novosibirsk State University, Psychology Faculty, Novosibirsk, Russia

Introduction. The majority of chronobiological questionnaires were developed without conducting conventional psychometric analyses. They were criticized for poor or lack of statistical analyses that could provide the basis for their psychometric evaluation (e.g. Smith et al., 1989). The general aim of our studies was to evaluate psychometrically a questionnaire battery for assessment of the individual traits of the sleep-wake cycle. In the present research, the factor structure of this battery was cross-validated with matched samples of university students living in Alaska and Novosibirsk.

Method. Two English language questionnaires with 4 scales/subscales (Folkard et al., 1979; Horne, Ostberg, 1976) and two Russian language questionnaires with 6 scales/subscales (Putilov, 1990, 1997) were included in the list of 92 questions/statements. After translation and minor rewording, the English version of the battery was given to 97 students of the University of Alaska Anchorage, and the Russian version was given to 117 students of the Novosibirsk State University.

Results. It was found that both versions are almost identical on 1) Cronbach alpha for any scale/subscale, 2) magnitude of the difference between age and gender groups on the mean scores, 2) correlation patterns between the scales/subscales'scores, 3) a factor structure of the whole battery, and 4) a factor structure of the 40-item Sleep-Wake Pattern Assessment Questionnaire (SWPAQ; Putilov, 1990). Besides, the factor structure of SWPAQ was also in agreement with the structure obtained earlier (Putilov, 1993) for a general population sample of 2000 males and 2126 females aged 9 to 79 yrs.

Discussion. The agreement between the psychometric results obtained with Russian and English versions of the questionnaires should be considered as confirming the validity of the main battery scales. Earlier (Cherepanova, Putilov, 1993; Putilov, 1993, 2000; Melnikov et al., 1999) we provided several other facts suggesting validity of SWPAQ and other chronobiological questionnaires. Our questionnaire results suggest that, in addition to circadian regulation, the contribution of another chronophysiological mechanism - homeostatic regulation of the sleep-wake cycle - needs to be accounted to explain the difference between people with low and high scores on two - morning and evening - lateness scales.

Conclusion. The results suggest that the subjectively assessed individual differences in the habitual features of the sleep-wake cycle reflect the underlying differences in the parameters of chronophysiological regulatory mechanisms. The questionnaire battery might be applied for chronotypological studies, i.e. for evaluation of: 1) habits related to sleep-wake cycle in different cultures, 2) chronobiological and personality correlates, 3) diurnal and seasonal aspects of personality, 4) the chronobiological risks of night and shift work, etc.