Meta-analyses traditionally compare the difference in means between groups for one or more outcomes of interest. However, they do not compare the spread of data (variability), which could mean that important effects and/or subgroups are missed. To address this, methods to compare variability meta-analytically have recently been developed, making it timely to review them and consider their strengths, weaknesses, and implementation. Using published data from trials in major depression, we demonstrate how the spread of data can impact both overall effect size and the frequency of extreme observations within studies, with potentially important implications for conclusions of meta-analyses, such as the clinical significance of findings. We then describe two methods for assessing group differences in variability meta-analytically: the variance ratio (VR) and coefficient of variation ratio (CVR). We consider the reporting and interpretation of these measures and how they differ from the assessment of heterogeneity between studies. We propose general benchmarks as a guideline for interpreting VR and CVR effects as small, medium, or large. Finally, we discuss some important limitations and practical considerations of VR and CVR and consider the value of integrating variability measures into meta-analyses.

Many seed quality tests are conducted by first randomly assigning seeds into replicates of a given size. The replicate results are then used to check whether or not any problems occur in the realization of the test. The two main tools developed for this verification are the ratio of the observed variance of the replicate results to a theoretical variance and the tolerance for the range of the results. In this paper, we derive the theoretical distribution and its related properties of the sequence of numbers of seeds with a given quality attribute present in the replicates. From these theoretical results, we revisit the two quality checking tools widely used for the germination test. We show a precaution to be taken when relying on the variance ratio to check for under- or over-dispersion of the replicate results. This has led to the development of tables providing credible intervals of the variance ratio. The International Seed Testing Association tolerance tables for the range of the results are also compared with tolerances computed from the exact theoretical distribution of the range, leading us to recommend a revision of these tables.

The number of days of data required to accurately estimate usual nutrient intake of children is not well established. This study aims to calculate the variability and the number of days required to estimate usual nutrient intake in children aged 13–32 months. This cross-sectional study, which is part of the BRISA Project in São Luís, Maranhão, Brazil, involved 231 children from April 2011 to January 2013. Socio-economic and demographic data were collected using a questionnaire, and 3 non-consecutive days of food consumption were collected using a 24-h dietary recall (24HDR) survey. Intrapersonal and interpersonal variability and variance ratio (VR) were obtained for each nutrient using the Multiple Source Method® program (version 1.0.1). The number of days (d) needed was calculated using the formula proposed by Black et al. for different correlation coefficients (r) (i.e. 0·7, 0·8 or 0·9). For the vast majority of nutrients, intrapersonal and interpersonal variability values of <1 were observed, with even smaller intrapersonal variabilities, resulting in low VR (<1). More days were needed to estimate intakes of soluble fibre (12), insoluble fibre (11), total fibre (10), vitamin C (9) and PUFA (7), while fewer days were needed for energy, carbohydrate, SFA, Ca, Fe, P and Zn (all had 2 d for r 0·9). However, most nutrients required one, two or three 24HDR for r 0·7, 0·8 or 0·9.