TY  - RPRT
TI  - Estimates of heterogeneity (I2) can be biased in small meta-analyses
AU  - Paul T. von Hippel
PY  - 2014
DO  - 10.1186/s12874-015-0024-z
UR  - https://arxiv.org/abs/1410.2296
ID  - 1410.2296
ER  -