About Impact Factor
PWRD enjoys a decent impact factor (5.512 in 2019) among similar electrical engineering journals. However, PWRD editorial board does not believe impact factor (IF) is a good indicator for the quality of a journal. This is because many non-quality related factors can affect IF, as explained below:
1. Legacy topics: A paper on legacy topics tends to draw fewer citations even though it can be useful to industry and academia. If a journal covers many legacy topics, its IF will be lower. One evidence is PWRS versus SG. Although both attract similar papers, PWRS has lower IF since it covers legacy topics such as power system reliability. Another evidence is that many newer journals can build up IF fairly quickly as they don’t cover legacy topics or few authors of legacy topics submit to such journals.
2. Researcher population: Each journal covers a few trendy topics. However, the researcher (i.e. author) population in each topic can be different. So the number of citations in each topic can be quite different. One example is the HVDC topic covered by PWRD versus microgrid topic covered by SG. Both are trendy topics. But HVDC has smaller research population. The highest first year citations for HVDC paper is about 25 (regardless of where it is published). On the other hand, that for microgrid is about 35 (2017 data).
3. Experimental versus computing based research: Experimental research takes longer time to complete and submit. Then the cited references can fall out the 2 year window of IF calculation. (The effort required to do experimental research also affects the researcher population). If a journal covers more experimental related topics, its citation will be lower in comparison with a journal covering computing/algorithm related research, even though everyone knows experimental research results can be more convincing.
4. Industry application papers: Industry oriented papers can be a strength for engineering journals, as the ultimate goal for engineering journals is to bringing value to the society by publishing solutions to real-world problems. The same applies to research that is related to industry standards. However, such papers tend to have lower citations even though industry people read and use them more than academic or theoretical papers.
In summary, IF can be affected significantly by the nature of topics covered by a journal rather than the quality of research conducted on a given topic. Table 1 lists 5 IEEE Transactions that have the highest IFs in 2018. One can easily see that all of them are related to computing/algorithm research.
Table 2 lists 5 IEEE transactions that have the highest IF among the transactions that cover legacy topics, apparatus/hardware research, and experimental components (2018 data, two transactions are not included due to historical reasons). It can be seen all have a considerably lower IFs in comparison with computing/algorithm oriented transactions.
Furthermore, IF indicator can be counterproductive to engineering research. Science research is about discovery. A useful discovery means more follow up discoveries (i.e. high citations). Engineering research is about solving problems. A good engineering paper that fully solves a problem tends to draw less citations since there is little follow up work needed. (But industry will read and apply it). Citation rewards engineering papers that cannot solve a problem fully. In fact, creating a new (artificial) problem may lead to more citations.
Pursuing IF as the main goal for an engineering journal can have many negative consequences, such as discouraging experimental research that takes more time, discouraging research on classical topics even though the contributions can be significant, encouraging superficial research can be easily followed by other researchers.
Additional Information:
San Francisco Declaration on Research Assessment (DORA);