“…In cases where temporal stability or inter-rater reliability estimates have been reported, high variability across studies has been observed (Schaefer-McDaniel et al, 2009). More importantly, few tools are validated prior to being applied to large etiologic studies (Weich et al, 2001, Pikora et al, 2002, Gauvin et al, 2005, Zenk et al, 2007, Millington et al, 2009, Fuller and Muhajarine, 2010, McDonell and Waters, 2010, Paquet et al, 2010 although validation studies should be performed as formative steps to substantiate theoretical underpinnings of measurement and improve the measures themselves before being formally used in etiologic studies. As well, reliability is a necessary but insufficient component of validity: valid associations and interpretations cannot be derived from unreliable measures (Cook and Beckman, 2006).…”