Guest Post from Prof. Helena Dodziuk on Excessive Science
Today we have a guest post from Prof. Helena Dodziuk on ‘Excessive Science.’ Prof. Helena Dodziuk studied Physics at Warsaw University before completing her Masters in Physical Chemistry and her PhD in Organic Chemistry. She then went on to work in chemical institutes throughout her career. Helena Dodziuk was presented her ‘Professor’ title from the President of Poland in 2002. More about this can be read in an interview with Prof. Helena Dodziuk on ChemistryViews. More details about Prof. Helena Dodziuk scientific activity can also be viewed on her webpage.
She has authored/edited the following books; ‘Modern Conformational Analysis’, ‘Cyclodextrins and Their Complexes’, ‘Strained Hydrocarbons’, ‘Introduction of Supramolecular Chemistry’ and its Polish updated version. At present, Prof. Helena Dodziuk submits articles in Polish to the site of physicists from the Toruń University.
In addition to Chemistry and Popularization, Prof. Helena Dodziuk loves singing, traveling and sightseeing.
New or old but always hot
First of all I would like to specify what I am going to write about. The only criteria will be my fascination by phenomena, discoveries or simple results. Their actuality would not be decisive, today one forgets about important or simply appealing yesterday’s findings. In addition to new ones, I would like to remind you what caught my attention earlier. Today “Excessive science” will be presented. With it, I would like to start a serious discussion of inflated and overspecialized science, in which we all suffer of too much information. Since I studied Physics and carried research in Chemistry, my texts will be often related to these domains. However, more general topics like that discussed today as well as those related to health will also be covered. To stay not only in a very serious mood, in the next part I am going to discuss nanoputians, molecules with formulae resembling humans that have been synthesized without any practical aim, simply to have fun.
A colleague told me that the new version of the Gaussian program suite contains more than 300 functionals1 while a special website “Basis set exchange,” provides, as of 19 June 2013, 490 of them. According to the latter there are 250 published basis sets for the hydrogen and 310 for the carbon atom. Too much to reasonably deal with, even for someone involved only in the DFT quantum calculations who should also be able to analyze reliability of experimental data to compare with and to know specific properties of the molecules under study. I am interested in molecular properties determined by a combined application of experimental and theoretical methods2-4 and this immensity of the computational methods is unbearable for me. I also believe it is for many others. But this is only one facet of the crisis situation experienced by science today. Some specific manifestations of the crisis are discussed in journals and the Web. Proposals for the cure of the problems are given however no general discussion on the whole has been published. As it will be shown below in one example, the curing procedures themselves can also be problematic.
Most of the problems can be summarized by saying too much: too many experimental and computational methods, too much data and too many papers they create.
The abundance of experimental and computational methods and narrow specializations is typical for all areas of science, in particular chemistry. Note that this abundance is accompanied by the interdisciplinary character of most studies today. In addition, new methods replace the old ones. In 1990’s Prof. Kenneth Wiberg from Yale University told me that almost no one carried out calorimetric measurements at that time. Certainly, this was an exaggeration, but some basic information seems to be increasingly difficult to find even if it exists. To a great extent this is due to too much data which is difficult to handle.5 For instance, the flood of publications (analyzed by Bauerlein et al.6) according to Michael Mabe (as cited in ref. 6), the number of publications grows at a rate of 3.26% per year doubling in about 20 years. As presented in7a, doctors/practitioners who 20 years ago refused to accept any clinical guidelines from the British National Institute of Health and Clinical Excellence concerning the drugs and procedures to be used, today have changed their minds; unable to face 18 – 20 papers that they should read every day seven days a week to cope with the data.
The growing number of scientists competing for grants and the pressure to publish, especially in high impact journals, is to blame for the surfeit of scientific information. A significant percentage of the published articles in science are not only not cited but also not read at all. According to Bauerlein et al.6, “the amount of redundant, inconsequential, and outright poor research has swelled in recent decades,” raising the cost of publishing, reviewing, and simple reading of scientific literature.
On the other hand, a dangerous opinion that “anything more than a few years old is obsolete,” becomes more and more accepted. Thus, there is neglect of older papers independently of their real value.
Decreasing the number of published papers and limiting them to only high-quality and short ones, e.g. “stopping the over publication,” and taking into account impact factors, IF, of only few articles for a position is presented by Bauerlein and collaborators as a wonder drug which will heal science.6 In the same way as favoring the candidates for the academic positions “with high citation scores, not bulky publications,” in our opinion, will not help significantly.
A concern for high citations is they are prone to manipulation8, to fight it down, the owner of Institute of Scientific Information, Thomson Reuters, every year bans from its listing journals for self-citations9. Not only single journals but also “citation cartels” of journals and individuals do exist10. On the other hand, even in the same scientific domain, if you produce outstanding works in an important but very specialized area, you will never reach high scores.
Another burden difficult to fight with is publishing of results that cannot be reproduced. (Of course, we are not talking about the Journal of Irreproducible Results11) Some of them deliberately produced frauds, others contain simple errors or misconceptions. The most famous from the former category was the case of a rising star aspiring to the Nobel Prize in Physics, Jan Hendrik Schön (whose numerous articles were withdrawn by highly respected journals after the publication5) and Hwang Woo-suk who claimed to succeed in cloning human embryonic stem cells12. Wikipedia even gives a list of experimental errors and frauds in physics5. In 2009 an article13 with the intriguig title “How many scientists fabricate and falsify research? A systematic review and meta-analysis of survey data” appeared.
Closer to applications, Begley and Ellis14 analyzed preclinical cancer data published in top-tier journals, 90% of which could not be reproduced even by the investigators themselves. As a result, Nature recently published a special comment by Begley14, presenting criteria allowing one to specify papers reporting preclinical cancer research “that don’t stand up to scrutiny.”
In an effort to fight frauds, the Journal of Cell Biology adopted a special screening procedure to detect image manipulations10. As the result, 1% of the publications accepted by the journal are revoked after the positive referee reviews. Moreover, 25% of all accepted articles were found to contain at least one figure that should be corrected because of “inappropriate” manipulation. Similarly, the Organic Letters journal has just hired a data analyst to inspect the submitted data (in articles as well as in the Supporting Information) for evidence of manipulation15. The scientific misconduct is a great problem but it is only the tip of the iceberg of severe problems the science faces today and, in my opinion, there is no simple method to solve the problems.
Let us know your views by writing a comment in the ‘Leave a reply’ box below
(1) http://www.gaussian.com/g_tech/g_ur/k_dft.htm functionals in Gaussian; 2013.
(2) Dodziuk, H. In Modern Conformational Analysis. Elucidating Novel Exciting Molecular Structures; Wiley-VCH: Weinheim, 1995; pp 157-211.
(3) Bernatowicz, P., Ejchart, A., Ruszczynska-Bartnik, K., Dodziuk, H., Kaczorowska-Molchanow, E., Ueda, H. J. Phys. Chem. B 2010, 114, 59-65.
(4) Dodziuk, H., Korona, T., Lomba, E., Bores, E. J. Chem. Theor. Comput. 2012, 8, 4546-4555.
(5) http://en.wikipedia.org/wiki/List_of_experimental_errors_and_frauds_in_physics Errors and frauds in physics; 2013.
(6) Bauerlein, M., Gad-el-Hak, M., Grody, W., McKelvey, B., Trimble, S. W., http://chronicle.com/article/We-Must-Stop-the-Avalanche-of/65890/ stop the overpublishing; 2013.
(7) http://www.nature.com/news/assessing-the-value-of-health-treatment-1.12701#auth-1 Assessing the value of health treatment; 2013.
(8) Harzing, A.-W., http://www.harzing.com/esi_highcite.htm How to become an author of ESI Highly Cited Papers?; 2012.
(9) van Noorden, R., http://blogs.nature.com/news/2012/06/record-number-of-journals-banned-for-boosting-impact-factor-with-self-citations.html, Record number of journals banned for boosting impact factor with self-citations; 2012.
(10) http://jcb.rupress.org/site/misc/about.xhtml manipulated figures; 2013.
(11) http://www.jir.com/ Journal of Irreproducible Results.
(12) https://en.wikipedia.org/wiki/Hwang_Woo-suk Hwang Woo-suk scandal.
(13) Fanelli, D. PLOS one 2009, 4, e5738.
(14) Begley, C. G., Ellis, L. M. Nature 2012, 483, 531-533.
(15) Smith, A. B., III, Org. Lett. 2013, 15, 2893-2894.