Selfcite

Introducing Selfcite 2.0 - career enhancing software

Nick Craddock, Michael C O'Donovan, Michael J Owen

Citation rate is becoming increasingly important as an index of success within medical academia.(1) Self citation (referencing one's own earlier publications in a new publication) is a useful method for increasing citation rate.(2) Although some academics are undoubted masters of this approach, maximising the benefit from self citation can be tedious, so we have developed a user friendly software package, Selfcite, to help in this task. In this paper we introduce the theory of self citation and give a short description of the Selfcite program.

The benefits of self citation: theoretical considerations
To achieve maximum benefit from self citation, each new publication should cite all earlier works.(3) Thus, in one's second paper, the first is cited. In one's third, the first and second are cited, making a total of three citations. In the fourth, the first, second, and third are cited, making a total of six citations. The mathematically inclined reader will immediately recognise that the number of possible citations is the sum of a series of triangular numbers, and that the total number of citations possible on publication of the nth paper is given by n(n-1)/2.(4) Thus, the tenth paper produces 45 references and the 100th produces 4950. The expected accumulated reward is shown graphically in fig 1.

Total potential number of self citations as a function of the number of publications

Clearly this is the ideal situation and it will rarely be possible to realise this goal: despite a commitment to the principle of self citation, the content of manuscripts may prevent later citation. In this regard it is worth remembering that early publications should be kept very general with a view to later citation in progressively more specialised works.(5)

Recognising the burden on time and ingenuity of incorporating citations of one's earlier work into new manuscripts we were motivated to develop Selfcite.

The Selfcite 2.0 package
Selfcite 2.0 is an integrated suite of programs that runs on IBM compatible personal computers under Windows and is compatible with the major word processing packages, reference managers, and online reference databases. A database within Selfcite stores details (including title, key words, and abstract) of each of the user's publications. The program can be used in interactive mode or non-interactive mode. In interactive mode the user composes the text of a new manuscript and Selfcite makes context sensitive suggestions for citations of one's own work. The user can then select or reject suggestions. In the non-interactive mode Selfcite can be used to scan the final text of a manuscript and automatically insert self citations. The algorithm used to decide on inclusion of a citation involves assigning a score to each of the user's previous publications based on the similarity between words and phrases in the nascent manuscript and those used in the database record for previous publications. The user sets a threshold score that determines whether or not a citation will be inserted automatically in scan mode.

The package has several important features:

Maxicite--this option maximises the number of self citations and can usually incorporate over 80% of the user's earlier work. |ay Megacite--this option ensures inclusion of all one's previous papers and if necessary inserts them randomly into the text. We recommend that this option should be used only sparingly as reviewers and editors may notice that references do not have a crucial bearing on the text.

Minicite--This option allows the user to choose the desired number of self citations for a given manuscript and the program selects those most appropriate.

Modesty--This overrides the program and prevents self citations (not recommended for general use).

Multicite--This allows multiple users to maximise the number of joint self citations. This is particularly useful for research groups.

|ay Shaft--A popular option introduced in version 2.0. It allows the user to specify one or more other researchers who should not be cited. A warning message flashes on the screen if one of the researchers in the shaft list is coauthor of a paper that the user is inadvertently trying to cite. The closest self citation is suggested as an alternative.

Paracite--An extremely useful feature. The program deliberately miscites one's own earlier publication in the hope that another worker will write a letter to the journal pointing this out, thus allowing the user to publish a letter in reply, thereby providing further self citation opportunities. (The default setting is paracite on but only once per manuscript.) |ay Citers-block--The user decides on the references to be cited and the program offers suggestions for plausible passages of text to support their citation.

Stats--This set of options provides statistics about the overall efficiency of the self citation process (total number of self citations divided by total number possible) and will present the data in graphical form. The algorithms monitor potential self citation classics and automatically increase their citation priority in the main part of the program.

Conclusion
We believe that self citation is currently handled inefficiently by most (but by no means all) researchers and that its correct use is beneficial to an academic's career. We have developed a user friendly package of programs that minimises the arduousness and maximises the benefit of the self citation process. We are confident that it will prove useful to researchers as well as to clinicians in training. The current manuscript has been prepared using Selfcite 2.0 with multicite = on, minicite = 5, and shaft = global.

Programs are available from the authors at: http://www.onan.Selfcite.

Funding: NC is a Wellcome trust senior research fellow in clinical sciences.
Conflict of interest: None.

References:
1 Craddock N, Owen MJ. Modern molecular genetic approaches to psychiatric disease. Br Med Bull 1996;52:434-52.

2 O'Donovan MC, Guy C, Craddock N, Murphy KC, Cardno AG, Jones LA, et al. Expanded CAG repeats in schizophrenia and bipolar disorder. Nature Genetics 1995;10:380-1.

3 Craddock N, Asherson P, Owen MJ, Williams J, McGuffin P, Farmer AK. Concurrent validity of the OPCRIT diagnostic system: comparison of OPCRIT diagnoses with consensus best-estimate lifetime diagnoses. Br J Psychiatry 1996;169:58-63.

4 Craddock N, Khodel V, Van Eerdewegh P, Reich T. Mathematical limits of multilocus models: the genetic transmission of bipolar disorder. Am J Hum Genet 1995;57:690-702.

5 O'Donovan MC, Guy C, Craddock N, Bowen T, McKeon P, Macedo A, et al. Confirmation of association between expanded CAG/CTG repeats and both schizophrenia and bipolar disorder. Psychol Med (in press).