File(s) not publicly available

Reviewing the psychometric properties of contemporary circadian typology measures

journal contribution
posted on 06.12.2017, 00:00 by Vitale Di Milia, A Adan, V Natale, C Randler
The accurate measurement of circadian typology (CT) is critical because the construct has implications for a number of health disorders. In this review, we focus on the evidence to support the reliability and validity of the more commonly used CT scales: the Morningness-Eveningness Questionnaire (MEQ), reduced Morningness-Eveningness Questionnaire (rMEQ), the Composite Scale of Morningness (CSM), and the Preferences Scale (PS). In addition, we also consider the Munich ChronoType Questionnaire (MCTQ). In terms of reliability, the MEQ, CSM, and PS consistently report high levels of reliability (40.80), whereas the reliability of the rMEQ is satisfactory. The stability of these scales is sound at follow-up periods up to 13 mos. The MCTQ is not a scale; therefore, its reliability cannot be assessed. Although it is possible to determine the stability of the MCTQ, these data are yet to be reported. Validity must be given equal weight in assessing the measurement properties of CT instruments. Most commonly reported is convergent and construct validity. The MEQ, rMEQ, and CSM are highly correlated and this is to be expected, given that these scales share common items. The level of agreement between the MCTQ and the MEQ is satisfactory, but the correlation between these two constructs decreases in line with the number of ‘‘corrections’’ applied to the MCTQ. The interesting question is whether CT is best represented by a psychological preference for behavior or by using a biomarker such as sleep midpoint. Good-quality subjective and objective data suggest adequate construct validity for each of the CT instruments, but a major limitation of this literature is studies that assess the predictive validity of these instruments. We make a number of recommendations with the aim of advancing science. Future studies need to (1) focus on collecting data from representative samples that consider a number of environmental factors; (2) employ longitudinal designs to allow the predictive validity of CT measures to be assessed and preferably make use of objective data; (3) employ contemporary statistical approaches, including structural equation modeling and item-response models; and (4) provide better information concerning sample selection and a rationale for choosing cutoff points.

History

Volume

30

Issue

10

Start Page

1261

End Page

1271

Number of Pages

11

eISSN

1525-6073

ISSN

0742-0528

Location

UK

Publisher

Taylor & Francis

Language

en-aus

Peer Reviewed

Yes

Open Access

No

External Author Affiliations

Institute for Health and Social Science Research (IHSSR); School of Business and Law (2013- ); Universitat de Barcelona; University of Education (Heidelberg, Germany); Università di Bologna;

Era Eligible

Yes

Journal

Chronobiology International