Even if each reviewer has certain biases they can’t detect, as long as those biases are not shared by the other reviewers, then they will be removed.

If (B) is not true—all scientists share the same biases—then no scientists will be able to detect and remove them. As a result, all scientific papers would share these biases, and the conclusion would fall apart.

The argument is unconcerned with the scientific value of these papers. We care about whether or not these papers are biased, not if any present bias affects their value.

The premises list data interpretation as one potential area for bias, but they also tell us that any bias overall can be removed by the reviewers, provided they don’t share it. Even if some other aspect of the paper is biased, it could still be detected, so this answer choice is not necessary.

The premises give us review as one method of detecting bias, but it doesn’t have to be the only method. In fact, negating (E) to say, “biased data can be removed by means other than review by scientists who do not share the same biases,” supports the conclusion that scientific papers are unbiased, so it cannot be our answer.