Scientific Reports 6: Article number: 20552; published online: 03 February 2016; updated: 28 March 2017

This corrigendum aims to correct errors detected in the outputs of the Extended Data Tables from our article Scientific Reports 6, 20552 (2016). A mistake occurred due to a failure in the conversion of the variable “species” from integer to a categorical (factor) variable. We have updated the tables with ‘species’ as a factor (Extended Data Tables below). It is important to state that the essence of our results and conclusions do not change from those presented in the article.

Additionally, we took opportunity of this corrigendum to explore the variation in the group composition (party size) of humans that suffered large carnivore attack over time and across species. We simplified our response categorical variable ‘group composition’ into two categories, i.e., ‘victim alone’ and ‘victim in a party’ (binary coded variable). We conducted this analysis using a two-level categorical approach (binomial family in R) instead of using the previous multinomial approach. This change does not affect the results or conclusions of our analysis, as reported in the article, but is a simplification.

The correct Extended Data Tables 1, 2, 3 and 4 appear below as Table 1, 2, 3 and 4 respectively.

Table 1 Variation in the number of large carnivore attacks on humans over time and among species.
Full size table
Table 2 Variation of the age of victims in large carnivore attacks on humans in relation to time and species.
Full size table
Table 3 Variation of the group composition (party size) targeted in a large carnivore attack on humans over time and across species.
Full size table
Table 4 Relationship between the yearly number of large carnivore attacks and the number of recreation visitors in national parks in the US.
Full size table