Page 343 - CyberSecurityEssentialsEnglish

address it. Avoid asking an author to run an entirely new study merely to broaden interest,
and avoid imposing unnecessary citations.

When receiving reviews, a tool can help you classify the comments and build a response
table, but it should not write an automatic defensive reply. Read the comment, decide

whether it is correct or the product of ambiguity in your text, and then state the change or
the reason for not changing, with evidence. A response is not a victory; it is an improvement

of the record.

In institutional review of policies or reports, the same logic applies: an independent
reviewer, a list of criteria, documented comments, clear responses, and no concealment of
amendments. The presence of a tool does not remove the value of a human second pair of

eyes, especially when the report affects people, a budget, or a reputation.

The most serious problem in automated review is overconfidence. A tool may criticize the
absence of an element that is in fact present, request a test unsuited to the data, or generate
references to justify a comment. Every suggestion needs verification. More than one

perspective can be used, a methodological reviewer, a field reviewer, and a clarity reviewer,
with the decision remaining with the researcher or the editor.

TABLE 7-33: RELATIVELY SAFE AND HIGH-RISK USES IN REVIEW

Relatively safe uses High-risk uses

Building a checklist of appraisal questions Uploading a confidential manuscript to a

for a design. public tool.

Rephrasing your own comment more Delegating the accept or reject decision to
clearly and respectfully. a model.

Checking internal consistency in your own Accepting a generated criticism without
draft. verifying it.

Organizing received comments into a Letting a tool generate references to

response table. justify a comment.

Summarizing the structure of your own Sharing reviewer identities or confidential
paper before submission. correspondence.

342