Page 301 - CyberSecurityEssentialsEnglish
P. 301
TABLE 7-2: WHAT CAN BE DELEGATED AND WHAT CANNOT
What remains the researcher’s
Stage Appropriate help from the tool
responsibility
Framing the An initial map of concepts and Defining the problem, its
problem stakeholders. importance, and its limits.
Comparing the characteristics of Choosing the design and judging
Method
different designs. its fit to the question.
Discovering Proposing synonyms and query Approving the search strategy
evidence formulations. and the inclusion criteria.
Screening and Assisted first-pass sorting or field The inclusion decision and the
appraisal marking. final judgment of quality.
Structured summarizing or Interpreting the results and tying
Analysis
inspectable calculations. them to the context.
Improving structure, language, Responsibility for the claims and
Writing
and clarity. the citations.
Documentation Preparing an organized usage Disclosure and compliance with
and publication record. institutional and journal policy.
✕ Common misconception Correction
Fluency and a plausible reference list are
Treating a fluent, well-referenced-looking formal properties. A finding requires a
output as a research finding. source that exists, matches, and supports
the sentence attributed to it.
✓ Practical tip: write the sentence of use
Before using any tool, write in one line: “I will use it at step X to produce Y, and I will
verify by Z.” If you cannot complete the line, the limits of the use are not yet clear.
● Remember: Section summary
• Key idea: AI is an assistant across the stages of research, not an automatic source of
evidence.
• Practical decision: Classify the use, define the impact of error, and set the level of
verification before you begin.
• Mistake to avoid: Confusing a convincing output with inspectable evidence.
300

