Along with my co-authors, we would like to thank readers for their interesting comments and for their interest in our work. by reading all the comments, and in agreement with the editor, we realised the need to provide more details regarding some procedures taken in our studio. in particular:

1) Randomization order and sample size
we found a typo in the reported distribution of male and female participants in the original manuscript. the correct distribution was as
follows:
- male participants (44.2%): neutral condition: 22; violent game condition: 25; violent-sexist condition: 22.
- female participants, neutral condition: 29; violent game condition: 30; violent-sexist condition: 26.

Moreover, it is noteworthy that our research was conducted in a real context (i.e., high school) rather than in a lab context. thus, the practical limitations of this context did not allow us to assign participants to group conditions in a complete randomized way and the randomization process was dependent on participants' classes. crucially, participants' age, when entered as covariate, did not affect the pattern of our results.

2) Male role norm inventory measure
we think that a more detailed description on why and how we selected the items used in our study for measuring masculine beliefs is needed. it has to be noted that our research was conducted in a high school setting and mainly involved under-aged students. after a careful analysis, the high school committee in charge for the approval of our study, evaluated some of the original items proposed by levant and colleagues (2007) to be unsuitable for a sample of adolescents (i.e., "men should always take the initiative when it comes to sex"). thus, some of the items were excluded whereas the wording of some other items was adapted to our sample.

3) Identification with the game character scale
the identification with the game character represents a key-factor in the model described in our paper. thus, would like to better specify how it was operationalized. the original scale proposed by van looy et al., 2012 is composed of three different sub-scales: (1) wishful identification, (2) similarity identification, and (3) embodied presence.
to keep our study within the 1-hour time limit given to us by the high school, we dropped the "similarity identification" factor, which is defined as "the degree to which the player sees their avatar as similar to him/herself" (van looy et al., 2012, p.129). this aspect is typical of the mmorpg virtual environments (van looy et al., 2012), in which players are usually required to create their own avatar and not to play with a predefined character (as it was in the video games adopted in our procedure). furthermore, the reliability for the five-item wishful identification subscale was rather low (alfa = .67), thus we decided to drop it from our analyses and to consider only the embodied presence subscale that was much more reliable (alfa =.92).
moreover, a conditional process model (process model 11) in which the type of video game was entered as the predictor, gender and the identification with the game character (computed considering all the available set of items) as the moderators, masculine beliefs as the mediator, and empathy toward female violence victims as the outcome variable confirmed the reported results. as in the original model, participant age, video game violence rating, and frequency of video game play were also included as covariates. analyses confirmed the predicted 3-way interaction between type of video game, participant gender, and identification with the game character on masculine beliefs (b = 0.29, se = 0.11, t(140) = 2.58, p = .010).

4) Revision of the online public dataset
to give to a broader number of researchers the opportunity to reanalyze our original dataset (available on the open science framework public database: https://osf.io/f62sa/) we have now published a revised version of the original database with more clear item labels.