Page 81 - UR221 Creative & Scientific Thinking_Neat
P. 81

6.17 Key Characteristics of Modern Scientific Methodology



               quantitative analysis.


               6.17       Key Characteristics of Modern Scientific Method-


                          ology



               6.17.1      Objectivity


                   Scientific investigations must be free from personal bias. Conclusions are based on
               data, not opinions. In computing:

                   ˆ   lgorithm performance is measured using benchmarks                                                       A

                   ˆ   esults are compared using objective metrics (time, space, accuracy)                                     R


               6.17.2      Systematic Approach


                   Problems are addressed through well-defined steps rather than random attempts.
               Example in computing:

                   ˆ   efining a problem                                                                                       D

                   ˆ   esigning an algorithm                                                                                   D

                   ˆ mplementing the solution                                                                                  I
                   ˆ   esting and evaluation                                                                                   T



               6.17.3      Empirical Evidence

                   Knowledge is built on observations and experimental results. In computer science:

                   ˆ   erformance testing                                                                                      P

                   ˆ   imulation results                                                                                       S

                   ˆ   xperimental evaluation of models                                                                        E


               6.17.4      Reproducibility

                   Experiments must be repeatable by other researchers under the same conditions. In

               computing:

                   ˆ   haring datasets                                                                                         S

                   ˆ   ublishing source code                                                                                   P
                   ˆ   ocumenting experimental settings                                                                        D






                                                     72
   76   77   78   79   80   81   82   83   84   85   86