IS GOOGLE SCHOLAR RELIABLE IN SCIENCE?
JOSÉ VICENTE SOLER
Professor Emeritus of Murcia University (SPAIN)
If you want to make a comment you must identify yourself as I do and send it to me jvs1@um.es
Updates 26/11/21; 30/11/21; 24/3/22; 30/1/22; 8/5/22; 25/6/22
This article is the result of my experience (scarce and useless) with Google Scholar (GS). I must say that my opinion is that of someone who has used it for scientific purposes. Perhaps in other areas its reliability would be different.
As I will have the opportunity to prove that the answer to the question posed by the title of this article is a clear NO, and I assume that GS has no interest in harming me specifically, I allow myself to assume that what I observe regarding my case can be generalized. This assumption is not gratuitous since GS has as a rule, according to my experience and to the information obtained from reliable sources, not to pay attention to any complaint posed by scientists about errors they observe in GS citations. On the other hand, the elimination of wrong assignments or duplications, and therefore the list of publications, is let in the hands of the authors themselves resulting in a high level of insecurity.
Since I do not like GS because of how it is implemented and because of its lack of reliability, I have to admit that I am not one of those who collaborate to improve its own profile. As a consequence, 348 publications appear in my GS page (https://bit.ly/3FykVm3) which represent 20% excess with respect to those appearing in a prestigious scientific database that collects exactly ALL my publications, neither one more nor one less: 288 (https://bit.ly/3BIZxYI). In other words, I should be happy with the information GS offers about me.
Since it is my will not to collaborate in the improvement of my GS profile, I may not be the right person to criticize it with respect to this point. However, the analysis in this article should serve to warn that, since the person who consults GS can not know the degree of commitment of the authors in the improvement of their profiles, the information obtained is of little or no value. In other words, anyone can claim that he is accountable for his data in GS but what he will never be able to do is to extend that confidence to all the articles collected from other authors and, therefore, that any general study using GS to assess the scientific production of those authors is not reliable.
A few of the most significant errors among the 59 excess publications in my GS are presented at the end of this page with the aim to give the reader an idea of what to expect when consulting GS and give it the value he/she deems appropriate. At the bottom of this page we present a list of some of these excess publications (you can easily find them by searching in this page ****).
An additional problem with GS is that it collects citations that come from Theses and Dissertations, written in languages not always understandable. If these works are later published in journals, it is foreseeable that the citations in both academic and editorial formats will coincide (same author and similar content) and, therefore, that the authors of the cited publications will have more citations in GS than if only the citations from the publications to which the Theses and Dissertations give rise were taken into account. If these works are not published, it is even more unfair to take them into account because it would mean they have been accepted in GS without the support of a scientific journal. Something similar occurs when considering citations in books.
GS, Clarivate (ISI WoS) and Scopus have a useful alert service that inform the user of the publications that cite him/her. However, there is an important difference: if an article cites two or more articles by an author, the GS alert indicates only one of them; the others report all of them. If the recipient of the GS alert has access to the citing journal, he/she should download the article in question to see if any more are cited and which ones. Good service!
33 of my publications appear in GS without citations while they have actually been cited. Consequently, my h-index is in GS lower (52) than that in ISI WoS (56), using only the most restrictive option: CSI-EXPANDED (https://bit.ly/3BIZxYI). I have found a case of an author in which the GS cites are 40% higher than in ISI WOS. It must be said that in ISI WoS authors can apply for, and get within a period not exceeding a few days, the justified correction of the number of citations of any publication, which is impossible in GS. Moreover, I have never observed in ISI WoS any error of the types I point out below and I am convinced that, if it were the case, the rectification would also be done with the same celerity.
The unreliable nature of GS makes some studies based on its data invalid. Very recently the paper "Ranking of researchers in Spain and Spaniards abroad" has been published (https://www.webometrics.info/en/GoogleScholar/Spain) based on h-indexes and number of citations. However, comparing the number of citations of researchers from very different scientific fields is a big mistake: in the areas of Biology the number of citations is higher than in Mathematics, for instance. And within Biology, researchers on Biochemistry have more citations than those on, for instance, Entomology (https://grupodih.info/). This situation makes the study even more inconsistent: the research areas in which the number of citations per publication (CxP) is higher will have a higher percentage of researchers from these areas in the top positions, which does not imply that they are better than those from areas with lower CXP.
Another aspect to consider is that in certain areas it is common for some publications (sometimes the majority) to have hundreds (if not thousands) of authors. Let us imagine one of these with 500 authors and that each author self-cites once (self-citations should occur to avoid self-plagiarism). Both the number of citations and the h-index of these authors of choral publications should not be compared with those who publish articles with, at most, for example, 20 authors. The problem is that the comparison is made and the result is obviously faker than a wooden euro. As evidence of what I say, the authors at the top of the ranking are usually from Corpuscular Physics or High Energy Physics who, due to the needs of their work, have to publish with a high number of collaborators. There is, therefore, no criticism on our part of this type of work but, on the contrary, that these authors are compared with the vast majority who do not publish articles of this nature.
To top off the inconsistencies, the work intends to make a ranking of Spanish researchers (residents in Spain and abroad). This is an impossible task since Spaniards are those who have a Spanish ID card and/or passport, and this information is not collected in GS or in any scientific database. The nationality of the authors cannot be established from the nationality of the institutions they work in or from their surnames since there are both, foreign researchers in Spanish institutions and researchers with Hispanic surnames in most countries over the world. It appears evident that the group of researchers GS says to be considered is undefined. The same applies to the research centers since only one is assigned to each researcher in spite of the fact that many of them specify several in their publications. An example of a researcher in 2021: Univ Autónoma Madrid & Hosp Princesa & Inst Salud Carlos III & Univ Calif San Diego. To which institution do they assign this author? This and other examples (even with more research centers invoked) make inconsistent the news that usually appear on rankings of these centers based on this and similar studies. According to the above, I find it very scandalous the support that the Spanish National Research Council (CSIC) gives to this study. I have no objection to the work of the authors who, as far as I can see, disagree with my opinion on the reliability of GS.
To conclude, serious databases publish the list of journals from which they draw their information taking their quality as a certain guarantee of the articles they include and the citations they provide. In other words, the information sources concerning the articles of an author or the citations they achieve are fully transparent, although we may disagree with the selection in some cases. GS does not publish the list of the publications, theses, dissertations, books, etc it considers. They just search in the Internet and include what they find, which obviously does not guarantee the quality of the sources. In my particular case I can boast of having been cited by an article published in the Tikrit Journal of Pure Science.
It is argued that GS is a free source of information, which is not the case with others. The truth is that this expense, in the case of ISI WoS or Scopus, is already made by our administration (with very good criteria) and, therefore, it is a waste to make use of a very deficient resource when there are much better ones available when it is not expensive to use. The reason may be that in GS it is the authors who control their publications and, provided that the list of their publications is correct (in my case it is not because of my decision not to collaborate with GS), this facilitates the work of those who use it for bibliometric purposes, while the use of ISI WoS or Scopus requires them to carry out a search that needs considerable effort. In other words, obtaining reliable data involves an effort that those who prefer to use GS are not willing to make.
Epilogue:
Four months after writing this paper I find on the Internet an article (signed by Emilio Delgado López-Cózar, Nicolás Robinson-García* EC3 Research Group: Evaluation of Science and Science Communication, Faculty of Communication and Documentation, University of Granada and Daniel Torres-Salinas EC3 Research Group: Evaluation of Science and Science Communication, Center for Applied Medical Research, University of Navarra, Pamplona) that anyone could think inspired me to write this paper. Its content is a wonderful complement to what has been written so far. I copy the part of the introduction that most directly relates to this blog. "In this paper we show the results of an experiment undertaken to analyze Google Scholar's capacity to detect citation counting manipulation. For this, six documents were uploaded to an institutional web domain authored by a false researcher and referencing all the publications of the members of the EC3 research group at the University of Granada. The detection of Google Scholar of these papers outburst the citations included in the Google Scholar Citations profiles of the authors. We discuss the effects of such outburst and how it could affect the future development of such products not only at individual level but also at journal level, especially if Google Scholar persists with its lack of transparency."
Added on 8/5/22
Research.com (Rc; https://research.com/) has just aborted a collection of rankings using GOOGLE SCHOLAR (Spanish researchers in chemistry can be seen at https://research.com/scientists-rankings/chemistry/es). To complete the destruction, those at Rc use the Microsoft Academic Graph (MAG) search system based on the use of AI, among other missions, they say, to distinguish the different ways in which authors' names appear. This addition is surprising if the point is to report on the h values that appear in GS. It seems that it is an attempt to raise the level of the work (AI is a not negligible plus), when the search for h values in GS requires no more than using in the basic Google search engine an expression such as "google academic Damià Barceló" to find the page of the one who leads the ranking of Spanish Chemists. There his h=157 appears, which coincides with the h=157 in the Rc study. It is, therefore, a mystery what has been the role of the AI. Of course, the use of MAG has not prevented the same author Avelino Corma (number 2 in the ranking of Spanish chemists and h=152) from appearing also as A. Corma with a different h (70) and occupying the 72nd position. It is clear that the AI not only does not avoid the botched GS studies, but also serves to create new ones, since I have not been able to find in GS the page of the supposed A. Corma.
Added on 8/7/22
I consult the Research.com ranking of Spanish chemists, among which I am not (because I am very very bad and critical), although there are many with h values lower than mine. There they appear (almost all of them) assuming that the h-values of authors from different areas of Chemistry are comparable, for example, CHEMISTRY APPLIED, CHEMISTRY MEDICINAL and CHEMISTRY PHYSICAL. Of course, for each author his h value and the institution he works for are given but not his area of research. Nothing more to add!
Conclusions (no offense intended): if, after knowing what is said here you believe, what GS publishes about you or what the rankings "manufactured" using GS reflect, you are one of those who are happy competing alone.
**********
The errors I mention in my GS research profile are of different types:
(A) Posts by my namesakes.
(B) Duplicate postings for no apparent reason.
(C) Duplicity due to different versions of the publication.
(D) "Triplicity" for its own sake
(E) Publication of indefinite nature
(A) Publications of my namesakes. Although my scientific areas are Organometallic Chemistry and Coordination within the more general Inorganic Chemistry, 12 articles are attributed to me in which the author J. Vicente who signs them is not me, such as,
A1) The Problem of Personal Identity in Analytic Philosophy
JÓB Vicente
Daimon Revista Internacional de Filosofia, 67-83, 2003
A2) Voltammetric determination of linuron at a carbon-paste electrode modified with sepiolite
P Hernandez, J Vicente, M Gonzalez, L Hernandez
Talanta 37 (8), 789-794, 1990
A3) Determination of clonazepam by flow injection analysis
C Latorre, MH Blanco, EL Abad, J Vicente, L Hernández
Analyst 113 (2), 317-319, 1988
A4) Isomeric Transformation Of 4-Aminobiphenyl by UV Radiation and its Influence on the Determination by Flow Injection Analysis with Amperometric and Spectrophotometric Detection.
MH Blanco, MC Quintana, P Hernandez, J Vicente, L Hernandez
Analytical letters 28 (15), 2683-2697, 1995
A5) Electrochemical Study and Direct Determination of Adenosine‐5′‐Monophosphate Coupled to 6‐Thioguanosine and a Glassy Carbon Modified Electrode with Gold Nanoparticles
L Mora, P Hernández, J Vicente, F Galán, L Hernández
Electroanalysis: An International Journal Devoted to Fundamental and …, 2008
A6) QUANTITATIVE SEPARATION OF COPPER WITH DOWEX A-1'RESINE
L Hernandez, T SEVILLA, J Vicente
ANALES DE QUIMICA SERIE B-QUIMICA INORGANICA Y QUIMICA ANALYTICA 81 (1), 82-87, 1985
A7) J=+ i Ph) C12 (title written by an ape)
J Vicente, MD Bermudez, MT Chicote, MJ Sanchez-Santano
J. CHEM. SOC., CHEM. COMMUN, 141, 1989
A8) Alkyltris (tertiary phosphinejzobalt compounds, cleavage of phosphine-metal bonds and re-arrangements in (R. Mohtachemi, G. Kan-nert, H. Schumann, S. Chocron, M. Mich-man),(310 …
AR Rotsch, C Rucker, ML Rudulier, F Sakai, MJ Sanchez-Santano, Journal of Organometallic Chemistry 310, 1986
A9) Determination of 2-amino-5-nitrothiazole by square-wave cathodic stripping voltammetry, P Hernández, MJG de la Huebra,
J Vicente, MH Blanco, L Hernández
Fresenius' journal of analytical chemistry 351 (7), 686-688, 1995
A10) Handbook of Mathematical Functions Handbook of Mathematical Functions, 1972
O ALEJOS, CDE FRANCISCO, P HERNANDEZ, M MUNOZ, J VICENTE,
A11) Inventory, dissemination and preservation of the geological heritage in urban areas–Lisbon City Case Study
C Pinto, J Vicente, M Pinto, GE Santo, M Muñoz, I Moitinho
GeoJournal of Tourism and Geosites 2 (8), 263-271
A12) JOURNAL OF THE ORGANOMETALLIC CHEMISTRY, VOL. 252, No. 2
NV Aiekseev, P Andrianary, A Arcas, W Beck, H Blau, MI Cl9 Bruce, ...
(B) Nine of my articles appear duplicated::
B1.1) o-N it rophenyl pal ladium (11) Complexes. Crystal and Molecular Structures of cis-[Pd {O-c6H4N (O) b},] and &-[Pd {oC~ H4N (o) o)(o-C6H4No t
J Vicente, MT Chicote, J Martin, M Artigao, X Solans, M Font-Altaba
J. CHEM. SOC. DALTON TRANS, 141, 1988 ¡El título está redactado a trompicones!
B1.2) o-Nitrophenylpalladium (II) complexes. Crystal and molecular structures of cis-[Pd {o-C 6 H 4 N (O) O} 2] and cis-[Pd {o-C 6 H 4 N (O) O}(o-C 6 H 4 NO 2)(py)] J Vicente, MT Chicote, J Martín, M Artigao, X Solans, M Font-Altaba, ... Journal of the Chemical Society, Dalton Transactions, 141-147, 1988
---------
B2.1) Maleonitriledithiolate complexes of Au (III), Au (I), Pd (II) and Pt (II) containing neutral or anionic ligands R Uson, J Vicente, J Oro Inorganica Chimica Acta 52, 29-34
B2.2) MALEONITRILEDITHIOLATE COMPLEXES OF GOLD (III), GOLD (I), PALLADIUM (II) AND PLATINUM (II) CONTAINING NEUTRAL OR ANIONIC LIGANDS R USON, J VICENTE, J ORO Chemischer Informationsdienst 13 (9)
-------------
B3.1) PREPARATION AND PROPERTIES OF STABLE SALTS CONTAINING MONO-(PENTAFLUOROPHENYL) AURATE (I) OR BIS-(PENTAFLUOROPHENYL) AURATE (I) AND MONO-(PENTAFLUOROPHENYL) AURATE (III), TRIS
R USON, A LAGUNA, J VICENTE
JOURNAL OF THE CHEMICAL SOCIETY-CHEMICAL COMMUNICATIONS, 353-354 1976
B3.2) Preparation and properties of stable salts containing mono-or bis-(pentafluorophenyl) aurate (I) and mono-, tris-, or tetrakis-(pentafluorophenyl) aurate (III) ions
R Usón, A Laguna, J Vicente
Journal of the Chemical Society, Chemical Communications, 353-354 51, 1976
-----------
B4.1) Preparation of organogold (III) complexes by oxidizing dichlordethanedigold (I), or bis (pentafluorophenyl)-mu-bis (diphenylphosphino) ethanedigold (I)
R Uson, A Laguna, J Vicente, J Garcia 2019
B4.2) Preparation of organogold (III) complexes by oxidizing dichloro-, or bis (pentafluorophenyl)-μ-bis (diphenylphosphino) ethanedigold (I)
R Uson, A Laguna, J Vicente, J Garcia
Journal of Organometallic Chemistry 104 (3), 401-406, 26, 1976
---------------
B5.1) REACTIVITY OF KETONYLGOLD (III) COMPLEXES-CRYSTAL AND MOLECULAR-STRUCTURE OF SP-4-4-[AU (2-C (6) H (4) N (2) PH)(CH2COC (6)-H4 (OME)(3)-3, 4, 5) CL (PPH (3))] AND SP-44-[AU (2 …
J VICENTE, MD BERMUDEZ, MP CARRILLO, PG JONES
JOURNAL OF ORGANOMETALLIC CHEMISTRY 464 (1), C38-C38 1994
B5.2) Reactivity of ketonylgold (III) complexes. Crystal and molecular structure of SP-4-4-[Au (2-C6H4N2Ph){CH2COC6H2 (OMe) 3-3, 4, 5} Cl (PPh3)] and SP-4-4-[Au (2-C6H4N2Ph)(CH2COMe …
J Vicente, MD Bermúdez, MP Carrillo, PG Jones
Journal of organometallic chemistry 456 (2), 305-312, 1993
-------------
B6.1) Stepwise Cyclopalladation of 2
MT Chicote, C Rubio, D Bautistab, J Vicente
B6.2) Stepwise cyclopalladation of 2-phenacylthiopyridine to give C, C, N-pincer complexes
MT Chicote, C Rubio, D Bautista, J Vicente
Dalton Transactions 43 (40), 15170-15182 6 2014
------------
B7.1) Synthesis and luminescence of poly(phenylacetylene)s with pendant iridium complexes and carbazole groups
J Vicente, J Gil‐Rubio, G Zhou, HJ Bolink, J Arias‐Pardilla
Journal of Polymer Science Part A: Polymer Chemistry 48 (17), 3744-3757 28, 2010
B7.2) Synthesis and luminescence of poly (phenylacetylene) s with pendant iridium complexes and carbazole groups Part A Polymer chemistry
J Vicente, J Gil-Rubio, G Zhou, HJ Bolink, J Arias-Pardilla 2010
-------------
B8.1) Synthesis of the First Hydroxomonoaryltin (IV) Complexes. Crystal and Molecular Structure of [{Sn [C, H,(N= NC6H,-M e-4')-2, Me-51 CI (p-0 H)} 2]
J Vicente, MT Chicote, MC Ramírez de Arellano, PG Jones 2019
B8.2) Synthesis of the first hydroxomonoaryltin (IV) complexes. Crystal and molecular structure of [{Sn [C 6 H 3 (N [double bond, length half m-dash] NC 6 H 4 Me-4′)-2, Me-5] Cl 2
J Vicente, MT Chicote, PG Jones
Journal of the Chemical Society, Dalton Transactions, 1839-1845 1992
---------------
B9.1) Synthesis of the First Trithiocarbonatogold complex:[N (PP~~) I~[AU~(~~=~~~, CS~)~ I. First Crystal Structure of a pz-+ Bridging Trithiocarbonato Complex
J Vicente, MT Chicote, P Gonzalez-Herreroa, PG Jones
J. CHEM. SOC., CHEM. COMMUN, 745, 1995
B9.2) Synthesis of the first trithiocarbonatogold complex:[N (PPh 3) 2] 2 [Au 2 (µ 2-η 2-CS 3) 2]. First crystal structure of a µ 2-η 2-bridging trithiocarbonato complex
J Vicente, MT Chicote, P González-Herrero, PG Jones
Journal of the Chemical Society, Chemical Communications, 745-746 28, 1995
(C) Duplicity is also caused by considering articles in the German and English versions published in Angew Chem as different (2 examples):
C1.1) Synthese zweier Diacetonyl‐TlIII‐Verbindungen und Röntgenstrukturanalyse von [Tl {CH2C (O) Me} 2 (μ‐CF3SO3)(bpy)] 2
J Vicente, JA Abad, G Cara, PG Jones. Angewandte Chemie 102 (10), 1184-1185, 1990
C1.2) Synthesis of Two TIIII Diacetonyl Compounds and Crystal Structure of [Tl{CH2C(0)Me}2(μ‐CF3S03)(bpy)]2
J Vicente, JA Abad, G Cara, PG Jones. Angewandte Chemie International Edition in English 29 (10), 1125-1126, 1990
----------
C2.1) New Motifs in Aurophilic Self‐Assembly: Synthesis and Structures of [Au(NHCMe2)2]CF3SO3 and [Au(CCSiMe3)(CNtBu)]
J Vicente, MT Chicote, MD Abrisqueta, R Guerrero, PG Jones
Angewandte Chemie International Edition in English 36 (11), 1203-1205, 1997
C2.2) Neue Muster bei der aurophilen Selbstorganisation: Synthese und Kristallstruktur von [Au (NHCMe2) 2] CF3SO3 und [Au(CCSiMe3)(CNtBu)]
J Vicente, MT Chicote, MD Abrisqueta, R Guerrero, PG Jones
Angewandte Chemie 109 (11), 1252-1254, 1997
(D) In one case, an article appearing in Angew Chem appears twice in the German and once in the English version:
D1.1) EIN MOGLICHER GOLDTRAGER IN HYDROTHERMALEN ERZLOSUNGEN: SYNTHESE UND STRUKTUR VON ERZLOSUNGEN: SYNTHESE UND STRUKTUR VON (PH3P) 2NAU (SH) 2, DEM ERSTEN HOMOLEPTISCHEN. VICENTE, MT CHICOTE, P GONZALEZ-HERRERO, PG JONES, Angewandte Chemie 106 (18), 1966-1967, 1994
D1.2) Ein möglicher Goldträger in hydrothermalen Erzlösungen: Synthese und Struktur von [(Ph3P) 2N][Au (SH) 2], dem ersten homoleptischen Hydrogensulfido‐Metallkomplex
J Vicente, MT Chicote, P González‐Herrero, PG Jones, B Ahrens
Angewandte Chemie 106 (18), 1966-1967, 1994
D1.3) A Postulated Carrier of Gold in Hydrothermal Ore Solutions: Synthesis and Crystal Structure of [(Ph3P)2N][Au(SH)2], the First Homoleptic Hydrogensulfido Complex. J Vicente, MT Chicote, P González‐Herrero, PG Jones, B Ahrens. Angewandte Chemie International Edition in English 33 (18), 1852-1853, 1994
E) There is an indefinite article missing (it seems to have to do with me, but I am not able to recognize it):
E1) www. rsc. org/chemcomm Volume 49| Number 73| 21 September 2013| Pages 7971–8118
I Saura-Llamas, J Vicente