Letter to the editor
Negative values of the intraclass correlation coefficient are not theoretically possible

https://doi.org/10.1016/0895-4356(96)00053-4Get rights and content

References (4)

There are more references available in the full text version of this article.

Cited by (66)

  • Reliability on multiband diffusion NODDI models: A test retest study on children and adults

    2021, NeuroImage
    Citation Excerpt :

    A MANOVA analysis was performed between data with different noise standard deviation estimates both on ICC and VF (see Tables 20,21 of Supplementary Material). Negative ICC values appear when the between-subject variation is relatively small compared to the within-subject variation and are theoretically difficult to interpret (Giraudeau, 1996; Rousson et al., 2002). Therefore, negative ICCs were set to zero (i.e. completely non-reliable), as suggested in other test–retest studies using the ICC (Braun et al., 2012; Duan et al., 2015).

  • Testing the reliability and effectiveness of a new tool for assessing urban blue spaces: The BlueHealth environmental assessment tool (BEAT)

    2021, Health and Place
    Citation Excerpt :

    Missing ICC values for attributes indicated too few cases were present or no cases were calculated. Any negative ICC values suggested poor or low agreement between raters due to the small sample size (n) (Giraudeau, 1996; Liljequist et al., 2019). The two-stage reliability testing revealed an improvement in scores from Stage-1 to Stage-2 for several attributes that required a more subjective interpretation of place quality and characteristics.

View all citing articles on Scopus
View full text