Reference map vs batch correction vs both

[MarieWangBioinformatics]

Dear Seurat users, I want to compare publicly available scRNAseq datasets. If I use the same reference map to identify different cell types, do I still need to do batch corrections beforehand? My understanding is that the reference map is a way to 'correct' the batch effect. So I am not sure if the batch correction is still needed. Thank you for your input.