How transparent is our study?
While I agree that our meta-analysis has limited coverage, I do not
accept the criticisms relating to publication bias and, in particular,
transparency at the peer-review stage. The conclusion is also based on
all alternative assessments showing no evidence for publication bias,
and both the raw data and code are submitted for revision. Owing to such
transparency, most issues raised in the comment have been carefully
evaluated by reviewers, who (i) repeated the literature search stepwise,
(ii) evaluated the influence of missing words, (iii) suggested to dig
into the data for a diet effect, and (iv) assessed the effects of
combining fully factorial with non-fully factorial experiments. They
have identified the limitations, but are enthusiastic about the detailed
picture provided and the opportunity to weigh empirical evidence with
theories (please refer to the point-to-point reply of Yin et al. (2019)
in the supplementary files).
With many thanks to Sánchez-Tójar
et al., I hope these discussions will promote open science practices,
methodological development and maturation of meta-analysis research,
which bridges empirical evidence with theories and enlightens future
explorations.