Disruption and Critical Transitions

Accompanying routines and software for:e

Loet Leydesdorfff & Lutz Bornmann, Disruption Indices and their Calculation using Web-of-Science Data: Indicators of Historical Developments or Evolutionary Dynamics? Jounal of Informetrics 15, 101219.

See also: Leydesdorff, L., Tekles, A., & Bornmann, L. (2021). A proposal to revise the disruption index. El Profesional de la información (EPI), 30(1), e300121;.

1: Computation of DI, DI*, and DI₅ using downloads from WoS

DI is based on WoS data (or similar data from Scopus). The routines di1.exe and di13.exe, at http://www.leydesdorff.net/software/di/di1.exe and http://www.leydesdorff.net/software/di/di13.exe, respectively, enable the user to compute values for DI, DI*, and DI_n on the basis of data samples downloaded from WoS.

The order of operations contains two cycles: DI.exe first organizes the DOIs which are used as input in the second cycle for the indication of DI, DI*, and DI₅.

1. Download the FPs in the “plain text” format including the cited references; rename the resulting file into data.txt. This is a required file name in the next step. If the download is composed of more than a single file, copy and paste these files in one single file “data.txt.”

2. Input this file to isi.exe in order to organize the data in relevant databases; the two files (data.txt and isi.exe) should be in the same folder.

3. Run di1.exe in the same folder; output is the file “doi.txt” with a search-string that can be used as input to WoS for searching with cited DOIs in a so-called Cited‑Reference Search. Note that one has to work with batches of not more than 1000 lines because of systems limitations in WoS.

4. In the next screen (with the results of the search), tick “select all” and “finish search” if necessary in batches of 1000.

5. Add the resulting search results to “Marked list” and download in plain-text format including the cited references. Create a new file “data.txt” after renaming the old one.

6. If necessary, collect downloads into a single file “data.txt” after storing the previous file under a different name. (The old files will be overwritten.)

7. For the same reason, rename core.dbf and core.dbt into fp_core.dbf and fp_core.dbt, respectively; rename cr.dbf into fp_cr.dbf.

8. Run isi.exe again; four files should now be available: core.dbf ; cr.dbf containing the cited references in the citing papers; fp_core.dbf and fp_cr.dbf (containing the cited references of the cited papers). Run di13.exe. The resulting file di13.dbf contains the values of DI, N_F. N_B, and N_R. The file di_full.dbf contains detailed information.

2. Critical revisions measured in WoS

1. For running this routine, one collects three sets in WoS: cited, citing, and focal records. Download in plain text format including cited references and process the three sets of data via isi.exe. Each set will generate a file cr.dbf with the respective cited references. Rename the files cr.dbf, respectively, into crante.dbf (cited records), fp_cr.dbf (focal records), and crpost.dbf (citing records). Files with these names will be further processed.

2. Run jj_cc.exe in the same folder as the three dbf-files and run the routine.

3. Output is written to screen and into the file jj_info.txt. It is also stored in jj_info.dbf. This file is generated if not present; otherwise, a record is added at the time of each run. After the values for the revision, the values of the revision in the reverse order are written (see section 7.1 above).

For example, for the case of six papers as follows (Table 4):

improvement -0.102 bits

I(post: ante) 0.962

I(post: between) 1.064

I(between: ante) 0.278

number of terms reverse the order between ante and between (Table 5)

improvement 0.545 bits

I(post: ante) 0.768

I(post: between) 0.222

I(between: ante) 0.997

number of terms 50

Loet Leydesdorff
Amsterdam, 23 July 2021