Page 305 - CyberSecurityEssentialsEnglish
P. 305

✕ Common misconception                                     Correction

                                                                   Match the question to the kind of
              Writing a broad or causal question and
                                                               knowledge it seeks. If the design cannot
               then choosing a method that cannot
                                                               exclude alternative explanations, do not
                              answer it.
                                                                     phrase the question causally.


           ? Question to consider: would a contrary result still matter?
           If the result came out against your expectation, would the question still be important and

           the method still acceptable? If not, the wording may have been designed to establish a
           position rather than to discover knowledge.

           ↻ Try it yourself: narrow one topic
           Write a broad topic, then turn it into a two-line problem, a demonstrable gap, one

           question, and two operational definitions. Ask the tool for three alternative explanations,
           then revise the question if needed.

           ● Remember: Section summary
           • Key idea: A research question is the result of a problem, a gap, and clear definitions, not
           merely an interrogative sentence.
           • Practical decision: Check the question against the type of knowledge, the five criteria,

           measurability, and ethics.
           • Mistake to avoid: Framing a broad or causal question without a method that allows it to
           be answered.

         7.3 Searching for Evidence: From Keywords to a Repeatable Strategy
           ◆ Key idea: searching is a repeatable strategy.
           Good searching is not one session in a search engine. It is a documented strategy that
           links the concepts of the question to synonyms, databases, and time and language limits.
           AI can widen the vocabulary and turn a question into queries, but the completeness and
           quality of the search need human review and testing against known studies.


         The search tool differs with the question. Web engines are useful for reaching reports, sites,
         and recent data, but they do not give systematic coverage of the literature. Specialized

         databases  index  journals,  fields,  and  subject  descriptors,  but  they  differ  in  scope  and
         language. One index does not represent all knowledge, so a search may need more than one
         database, along with grey literature such as government reports, theses, protocols, and

         standards.

         AI-supported search tools sit between several worlds: an academic database that surfaces
         indexed  records,  a  general  search engine  that  finds  recent  pages  and  reports,  a paper-
         summarizing tool that rephrases what you supply it, and a reference manager that stores



        304
   300   301   302   303   304   305   306   307   308   309   310