Which model is typically used in the presence of low statistical heterogeneity?

The fixed effects model is commonly used in situations where there is low statistical heterogeneity among studies being analyzed. This model operates under the assumption that the effects observed in the studies are similar, allowing for the combination of data across studies to yield a single estimate of the effect size. When heterogeneity is low, it indicates that the studies are providing relatively consistent results, making the fixed effects model an appropriate choice, as it focuses on the average effect across these closely related studies without accounting for variability that might be present in a broader context.

In contrast, the random effects model is designed for scenarios where there is significant heterogeneity, acknowledging that the true effect may vary between studies. It incorporates both within-study and between-study variability, making it less suitable when the assumption of similar effects holds true. The meta-regression model and quality assessment model serve different purposes and contexts unrelated to low heterogeneity; the former analyzes the relationship between outcomes and study characteristics, while the latter is utilized to evaluate the methodological quality of the studies. Hence, for low statistical heterogeneity, the fixed effects model is the most fitting approach, as it simplifies the data integration process under the assumption of uniformity.