Hellow.
I find the idea of H index problematic as it becomes a motivation by itself to ensure it will increase. it is problematic for another reason, I Believe that it is an indirect reflector (to some extent) of the of magnitude and the depth of networking and collaborations of a researcher.
In other words, researchers can raise this index simply by writing papers with peers, small groups or large groups, without actually making a discovery or being truly innovative. It is not filtering quotes from peers of the same institutions as an example.
I would like to replace it (for physics only) with an index which is raising only based on a single criteria, How many testable predictions a physicist made, and how many of them turned out to be roughly correct after margin ? It's the only thing that counts, the only measure that a TP impact should be measured upon, and it should also be binary, zero or an integer.
Physicist who made wrong predictions should have minus the number of wrong predictions they made. overall I think it can help further clarify who is at the top of the field rather than the current method of Hirsch.
The index you proposed seems to be applicable only to a rather narrow part of science.
For criticism of the h index, of impact factors etc., see also these discussions:
https://www.researchgate.net/post/Petition_Stop_Tracking_Science
https://www.researchgate.net/post/Your_opinion_about_the_following_link
https://www.researchgate.net/post/What_do_you_think_about_scientific_production_evaluation_indices_such_as_those_of_Hirsch_h_or_Google_Scholar_i10_or_others
https://www.researchgate.net/post/Research_Publications_Impact_factor_H-Index_I-Index_Is_these_words_becoming_root_cause_for_stress_nowadays_rather_than_Passionate
https://www.researchgate.net/post/Why_do_bibliometric_studies_get_so_much_attention_in_todays_world_of_academia
https://www.researchgate.net/post/What_are_impacts_researchers_can_create_beyond_citations_and_how_to_create_them
https://www.researchgate.net/post/Death_of_impact_factor_system
https://www.researchgate.net/post/Is_Impact_factor_to_be_avoided_from_journal_metrices_soon
https://www.researchgate.net/post/Impact_Factor_why_did_it_take_so_long_to_bury_this_stupid_variable
https://www.researchgate.net/post/Is-there-something-wrong-with-the-impact-factor
See also this interesting paper with criticism of the current misuse of citation analysis: How scientometry is killing science, https://www.geosociety.org/gsatoday/archive/24/12/pdf/i1052-5173-24-12-44.pdf
Wolfgang R. Dick Hellow.
"seems to be applicable only to a rather narrow part of science." well, Theoretical Physics is a narrow part of science.
Thanks for commenting.
I agree with your view, Dr Ohad Manor , that h-index is problematic and drives people a certain way - and that is generally incorrectly. I wonder, as suggested by Dr Wolfgang R. Dick , how it can be applied to areas other than theoretical physics.
Hellow Dr Joseph C Lee and thank you for your comment. It's fantastic hearing from you again. several questions:
1. In what ways you find it problematic ? I assume you have additional reasons than what I mentioned.
2. It is a very hard questions in non-exact fields which can't be falsified.
What would you think about an index which is filtering internal connections based upon certain criteria, as an example, people of the same institute, people who collaborated more than once and so on. The idea is to take into account only the quotes from unrelated people so it become less biased.
2.1 Perhaps it will be possible to use AI in order to evaluate creativity and uniqueness in non -exact fields, and take them into account rather than only the raw number of quotes.
Best regards
Ohad.
Hello, again, Dr Ohad Manor
I have had to think many times about how to answer your questions!
The easiest is perhaps your first question. Basically I think it is problematic that some people think that it is easy to judge another person's academic ability/achievements and their suitability for employment, just by the h-index.
This may lead to the answer to your second question. How can we make it better. It has to be both acceptable and reliable. As well, it has to be practical. Thus, the answer to your question is very hard and it may lead to more and more complicated questions! I agree that internal connections and collaborating with people you have worked with before should be less rewarded.
i don't think your concept would work, not even in physics. how would you define what counts as a prediction? (what if i predict that tomorrow will be a sunny day and my prediction comes true?) how would you define "roughly correct"? (any definition would necessarily depend so much on the nature of the prediction, that i can't imagine it not being more or less arbitrary.) who would decide about correctness? (in a narrow field, people from the researcher's "network" would be involved, so your concept would not eliminate "networking" bias.) why should only predictions count, but not inventions or discoveries? and why should your metric be binary? (there are more or less important predictions or discoveries.) and finally: your new metric would capture something very different from what the h-index is about, so why would you want it to replace rather than supplement the latter? networking is not bad per se, it enables collaborations and is therefore an prerequisite for advances in science.
Thomas Grischott
1. "how would you define what counts as a prediction?"
Its simple, A clear numerical result, which can be tested by modern means of measure within a specific Standard error.
2. "who would decide about correctness?"
If it is within the Standard error than Correct.
"people from the researcher's "network" would be involved, so your concept would not eliminate "networking" bias.)"
In my field people co-operate in massive groups, solely for raising their index. one of many examples, SM report, non technical involved above 35 people. Not a single equation was presented. The people who were in that group are in the "elite" Universities.
4. "why should only predictions count, but not inventions or discoveries?"
Because there is an over-inflation of "inventions" in TP, and progress can only be made if there exist a possibility to falsify, which is a direct result of a prediction. most papers do not predict and therefore do not count in one's view.
5. "so why would you want it to replace rather than supplement the latter?"
because the sole number of quotes not a real measure to impact. not In Theoretical physics.
Best regards.
- Badges
- Science topic
- Similar topics
- Indexes