In peer review, you are asked for your professional opinion. This is distinct from your personal opinion.
The editor would not like personal opinions like "the authors are foolish idiots who must have failed any basic statistics training". The editor would like professional opinions like "the authors did not use a statistical model that accounts for repeated measures in the same subjects, which could increase the risk of false-positive errors; I recommend they use repeated measures ANOVA or mixed effects modeling instead".
The editor would not like personal opinions like "the authors clearly have no idea what is going on in this field". The editor would like professional opinions like "There is recent work by Baggins (ref, ref) and Bombadil (ref) that would help explain the result in Figure 3. The manuscript would be improved if the authors discuss their results in the context of this prior work."
These professional opinions are objective in that they are grounded in your training in the field you work in. They are not based on your hunches and feelings, but based on your expert assessment of the work that is done. They are not based on your personal feelings towards the authors who may have once been rude to you at a conference or forgot to cite your paper or published a paper that supports a theory you oppose. They are based on the work you are reviewing and would be useful feedback to anyone involved: for the editor to make decisions about acceptance, for the authors to make revisions, and perhaps to any other reader.