Time taken to receive the first Citation by single and multiple authored Papers: an Analysis

by Dr. I.K. Ravichandra Rao
Professor and Head, Documentation Research and Training Centre,
Indian Statistical Institute, 8th Mile Mysore Road, Bangalore - 560059.
Email: ikr@isibang.ac.in

Abstract

[Data pertaining to single & multiple authorship and the data on the first citation received by those articles published in seventeen core journals in economics, library and information science, mathematics and statistics were analyzed. The purpose of the analysis is to determine the relation between the type of authorship (single or multiple) and time taken to get the first citation in the same journals. For each of the 17 journals, data were collected for about 17 to 20 years. The study suggests that there is no differences in the

• proportion of single authored and multiple authored papers, in general
• average number of single authored and multiple authored papers
• average time (in months) to get the first citation (in the same journal) by the single authored and multiple authored papers.

Similar analyses were carried out to each of the journals and also for each of the subjects. The results are discussed in the paper.]

1. Introduction

Collaboration in science is increasing almost every year and thus it is expected that the number of papers by multiple authors is increasing every year. It is further believed that papers by the multiple authors are likely to receive many citations and they get quick citations, since such papers are likely to have more visibility than the papers by the single authors.

The main objectives of the paper are to study

1. whether there is any difference in the average number of papers published by the single authors and multiple authors, in recent times
2. whether the papers by single authors receive first citation much earlier than the papers by multiple authors.

For practical reasons (-- due to the constrains of time & finance), the study is limited only to a few core journals in four subjects. Only those citations received in the same journals where the articles were published, are taken into account.

1.1 Related Work

Gupta and Karisiddappa (2) in their study on "Collaboration in theoretical population genetics" analyzed the growth of funded and collaborative research publications and authors in the area of genetics from 56-60 to 76-80. Their study indicates that the number of funded and collaborated research publications has not proportionally increasing along with the growth of total research publications. They also studied the extent which multi-authored publications are increasing. They have observed "increase in multi-authored publications". It has been increased from 23.28% during 61-65 to 28.75% during 78-80; three authored publications increased from 3.81% to 11.08%; four authored papers increased from 3.81% to 11.08%; four authored papers increased from 0.40% to 4.29% for the said period. O'Neill (3) in his study on "Authorship in theory based versus research based journals" observed that majority of the authors in "Educational Theory" and Journal of Educational Thought" were single authors. These findings challenge predictions that multiple authorship would eventually outnumber single authorship. Glanzel and others (1) in their article on a "bibliometric analysis of international co-operation of the European union (1985-95), based on the citation studies, observed international collaboration in particularly advantages for less advanced countries, but highly industrialized countries also benefit from it .

2. Data Collection

Seventeen journals (-- four core journals each in economics, mathematics and statistics; five journals in library and information science) were examined for a period of 17 to 20 years to collect the following data about:

• single authored papers
• multiple authored papers
1. multiple authors who are all within an institution
2. multiple authors who are within a country
3. multiple authors who are between countries

There were 14,835 articles; further references in each of these articles were examined to identify the articles which were received the first citation in the succeeding issues. The data on first citations are further grouped as:

• citations to single authors
• citations to multiple authors
1. within an institution
2. within a country
3. between countries

Among the 14,835 articles, 7,669 (51.7%) articles are single authored papers. Among the 7,166 multiple authored papers:

• 2,784 articles are by those who are within an institution (18.77%)
• 2,830 articles are by those who are within a country (19.08%)
• 1,552 articles are by those who are from different countries (16.46%)

Except statistics, in other three subjects the number of single authored papers is more than the number of multiple authored papers. The number of papers by single authors is less than fifty percent in following journals:

• Economic Journal (e3)
  
• Bulletin of Medical Library Association (l4)
  
• Journal of Royal Society (Series B) (m2)
  
• Communications on Pure and Applied Mathematics (m4)
  
• Sankhya: the Indian Journal of Statistics Series B) (s2)
  
• Journal of Statistical Planning and Inference (s4)

Further, it is observed that in

• Quarterly Journal of Economics (e1)
• Econometrica (e4)
• Journal of Documentation (e3)
• Journal of Royal Society (Series A and B) (m1 and m2)
• Annals of Mathematics (m3)
• Sankhya: the Indian Journals of Statistics (Series A and B) (s1 and s2)

the number of single authored papers is decreasing over the years and in all other journals, it is difficult to predict such a trend. The average values for different variables related to the authors and for first citations are given in Tables 1 and 2.

3. Data Analysis

3.1 Number of Papers by the Single and Multiple Authors

In all the seventeen journals together (in all the four subjects), it was observed that:

• there is no difference between the average number of papers published by the single authors (x₁ = 23.6) and multiple authors (x₂ = 22.05); the t- test indicates that P(T<=t) for two tail test is 0.1794; t-statistic is 1.3442 and the critical value is 1.9638.
• there is no difference between the average number of papers by multiple authors within the Institute (x₃ = 8.57 ) and within the country (x₄ = 8.71); the t-test indicates that P(T<=t) is 0.7834 and t-statistic is -0.27501.
• there exists a difference in the average number of papers between co-authors who are within the Institution (x₃ = 8.57) and who are from different countries (x₅ = 5.06); the t-test indicates that P(T <= t) for two-tail test is 4.99E-12; the t-statistic is 7.0416 and t-critical value is 1.9638. Further, one sided t-test indicates that the average number of papers by the co-authors who are within the Institution is more than the average number of papers by the co-authors from different countries.
• there exists a difference in the average number of papers between the co-authors who are within a country (x₄ = 8.71) and who are from different countries (x₅ = 5.06); the t-test indicates that P(T <= t) for two-tail test is1.08E-12; the t-statistic is 7.2686 and t-critical value is 1.9638. Further, one sided t-test indicates that the average number of papers by the co-authors who are within a country is more than the average number of papers by the co-authors from different countries.

By subject-wise, except in library and information science, it has been observed that there is no difference between the average numbers of papers by single and multiple authors. In library and information science, one sided test indicates that the average number of papers by single authors is more than the average number of papers by multiple authors; P(T <= t) is 0.003645. The t-statistic is 2.712259 and the critical value is 1.9724. Further, it was observed that for the following journals:

• Journal of Economic Literature (e2)
• Journal of Documentation (l3)
• Journal of College and Research Library (l5)
• Annals of Mathematics (m3)
• Sankhya : the Indian Journals of Statistics (Series A and B) (s1 and s2)

there is a difference between the average numbers of papers by single and multiple authors. Further, one sided t-tests indicate that except for Sankhya (Series B) the average number of papers by single authors are greater than the average number of papers by multiple authors. There may be many reasons for more number of papers by single authors; it may be even due to the editorial policies of the journals. I understand from one of my colleagues (Dr. Surekha Rao) that the journal of Economic Literature generally requests authors of repute to write survey like type articles in certain topics; it is perhaps because of this editorial policy expected that most of the papers are written by single authors. For the following journals:

• Quarterly Journal of Economics (e1)
• Economic Journal (e3)
• Econometrica (e4)
• Scientometrics (l1)
• JASIS (l2)
• Bulletin of the Medical Library Association (l4)
• Journal of Royal Society (Series A and B) (m1 and m2)
• Communication Pure and Applied Mathematics (m4)
• Journal of Statistical Planning and Inference (s3)
• The Annals of Probability (s4)

it has been observed that there is no difference between the average number of papers by single and multiple authors.

3.2 Average Time taken to receive the first Citation

For all the seventeen journals together, it has been observed that there is no difference in average time to receive the first citations by the single (on an average 5.79 months) and multiple authors (on an average 7.49 months). P(T<= t) is 0.0919; the t-statistic is -1.69113 and the critical value is 1.9685. In library and information science and statistics, it has been observed that there is no difference in average time to receive the first citations by the single authors and multiple authors .

However, this hypothesis was rejected in mathematics indicating that average time to receive the first citations by the multiple authors is greater than the average time to receive the first citations by the single authors. In economics, however, results indicate that the average time to receive the first citations by the single authors is greater than the average time to receive the first citations by the multiple authors. Thus, in mathematics and economics there is no impact of co-authorship in receiving early citations in the journals wherein the cited articles were published.

For all journals except

• Journal of College and Research Library, and
• The Annals of Probability,

it was observed that there is no difference in average time to receive the first citations by the single authors and multiple authors. There exists a difference in average time to receive the first citations by the single authors and multiple authors in the journal of College & Research Library and in the Annals of Probability. Further, the t-test indicates that the single authored papers receive the first citation much earlier than the multiple authored papers.

Data analysis for all the four subjects together based on the seventeen core journals that suggests that there is no difference between the average time to receive the first citation between the co-authors who are

• Within an institution and within a country
• Within an institution and between countries
• Within a country and between countries

4. Conclusion

In this study, it has been observed, in general,

• proportion of single authored papers is as high as the proportion of multiple authored papers.
• only in six core journals, the number of single authored papers is decreasing
• there is no difference between average numbers of the papers by single and multiple authors
• there is no impact of co-authors receiving early citations in mathematics and economics.

It is planned to carry out the study to determine the results for a large set of data based on SCI database.

Table 1: Averages, percentages and number of papers by single and multiple authors.

Table 1

Table 2: Time taken to receive the first citation

5. Bibliographical References

i) Glanzel, W. and others. "A Bibliometric Analysis of International Scientific Co-operation of the European." Scientometrics, 45 (2), June 1999, p.185-202.

ii) Gupta , P.M. and Karisiddappa, C.R. "Collaboration in Theoretical Population Genetics Specialty." Scientometrics, 42 (3), July-Aug. 1998, p.349-76.

iii) O'Neill, G.P. "Authorship patterns in theory based versus research based journals." Scientometrics , 41 (3), March - April 1998, p.291-98.

Back to Papers List
Top