.loin@published by Elsevier Science Ltd, Oxford and Akad~miai Kiadr, Budapest

Scientometrics, Vol. 42, No. 3 (1998) 349-376

COLLABORATION IN THEORETICAL POPULATION GENETICS SPECIALITY B. M. GUPTA,* C. R. KARISIDDAPPA**

*National lnstitute of Science, Technology, and Development Studies, Dr. K. S. Krishnan Marg, New Delhi 110 012 (India) **Department of Library and Information Science, Karnatak University, Dharwadl 580 003, Karnataka (lndia) (Received May 18, 1998) Analyses the growth of funded and collaborative research publications and authors as reflected in selected theoretical population genetics literature from 1956-60 to 1976-80. Indicates that the number of funded and collaborated publications has not proportionally increased along with the growth of total research publications and authors with time, but however, there is a strong correlation between the two. Indicates the extent o f multi-authored research publications in different countries, and studies the growth of multi-authored publications from 1956-60 to 1976-80. Studies the impact of funding and collaboration on the productivity of authors over a period of time. Concludes that the authors who are more productive are generally found to be more collaborative and funded. The average productivity per author is observed to be larger in funded and collaborated authors subset and smaller in non-funded and non-collaborated authors subset, than the average productivity per author in the total authors subset in all the five block years studied. There is a systematic increase with time in the average productivity per author in the funded and collaborated authors subset. Studies the nature and type of collaborated research from 1956-60 to 1976-80, and the role of funding. Highlights the research priorities of few important countries in collaborative research. Indicates the collaboration linkages among various countries in transnational collaborative research. Concludes that with time, the focus of research is slowly shifting from internal collaboration to domestic and international collaboration, supported by increasing funding from government agencies in theoretical population genetics research.

Introduction Collaboration has existed in science since its beginning, however, the first collaborative paper was published in 1655. Since then the number of collaborative publications have increased, initially slowly, and then dramatically faster after the

0138-9130/97/US $15.00 Copyright 9 1998 Akadkmiai Kiadr, Budapest All rights reserved

13. M. GUPTA, C. R. KAKISIDDAPPA: C O L L A B O R A T I O N IN THEORETICAL POPULATION GENETICS

middle of the eighteen century. The extent of collaboration further increased towards the end of this century. Collaboration came forward in response to professionalisation and increased knowledge according to Beaver and Rosen 1-3 from their analysis of publication data from 17th century to 20th century. Their study of the first professional scientific community, that of Napoleonic France, suggested the collaboration as a typical research style was associated with professionalisation. When France was the only country with a professional community uptill early 19th century, its scientists produced the bulk of the collaborative research publications. By the end of 19th century, Germany had become the world's scientific leader, and professionalisation of science had proceeded quiet far in England. Consequently the French share of collaborative research publications decreased over time, as the English and German share increased. Overall statistics for that period show that by that time too, collaboration activity was eventually divided among the major research producing countries. According to Pao 4 professionalisation refers to a complex set of dynamic processes involving organisations and scientists, individually and collectively, with respect to both the scientific group and community at large. It also sets the criteria for the new entrants, and formalises rules of behaviour among members, and promotes interaction between the specific groups and outsiders. Some indicators of professional maturation of an individual or research field can take the form of initiating professional societies, specialised journals, innovative applications from research programmes, and institutionalization of prizes and awards. The society, in turn also recognises the contribution made by an individual scholar or profession either taking professional advice from the scientists in the field of policy-making or in advisory capacities. Beaver and Rosen maintained that the collaboration functions as a social regulator within the profession, though its influence on recognition and subsequent access on resources, thereby increasing the visibility and productivity of authors. It was during the twentieth century that the professionalisation of science had its greatest impact on the members of the scientific community. And because o f this, there has been an increasing trend towards collaboration in almost all fields of science and technology with time. However, the extent of collaboration and their growth rate is observed to vary from one subject to another, one branch to another of the same subject, and from one country to another. Changes in the growth rate of collaboration is probably related to both change in the acknowledgment of teamwork, and as a actual increase in its practice. The critical change in the growth rate of collaboration occurred simultaneously with critical changes in the support o f science.

350

Scientometrics 42 (1998)

B. M. GUPTA, C. R. KARISIDDAPPA: COLLABORATION 1N THEORETICAL POPULATION GENETICS

A number of studies indicate that high productivity (in terms of published output) is indeed correlated with high level of collaboration and funding. Beaver and Rosen 1-3 while tracing the history of research collaboration, from 17th century onwards, indicates that scientific collaboration was the key element in the advancement of knowledge and the productivity of researchers in the field and enhances their mobility and visibility. In one of their study, Price and Beaver 5 found that the productivity per researcher increased as the number of collaborators increased. They noticed the existence of a small core number of active researchers, surrounded by a large floating population of scholars, who collaborated with leaders in only one or two projects, and then disappeared. In this study, they found that out of 555 authors in 1966, high producers with 14 or more publications were those who had 12 or more collaborators. Those who worked alone or with one collaborator were unable to produce not more than four publications in the same period. Zuckerman, 6 while studying 41 Noble laureates, showed a high degree of correlation between collaboration and productivity. According to her, laureates generally published more, that is why were more apt to collaborate, than the matched sample of scientists. Pao 7 investigated the relationship between collaboration and productivity in musicology, a subfield of humanities. Only 15% of the total literature was the result of collaborative activity involving 38% of the total authors in this field. Applying a normalised diversity measure, she found a high degree of correlation between collaboration and productivity. Pravdic and OluicVukovic 8 analysed collaborative patterns in chemistry, at both individual and group level. They found that scientific output as measured by publications is closely dependent on the frequency of collaboration among authors. The nature of the effect on. productivity depends upon the type of links. While collaboration, on one hand, with high productive scientists tends to increase personal productivity, and on the other hand, collaboration with low-productive scientists generally decreases it. Further the most prolific authors seems to collaborate more frequently and authors at all levels of productivity tend to collaborate with high productive authors than with low productive authors. Many scholars have also studied the impact of funding on collaboration and productivity. Price 9 suggested that the extent of collaboration in a field was directly linked with the size of funding support available. Hafner 1~ investigated the relationship between collaboration and financial support in four fields, namely political science, psychology, biological science, and chemistry. In all the four fields, financial support in research was associated with an increase in the total number of researchers (including co-authors and sub-authors) involved in the production of knowledge per research publication, the association and correlation was particularly strong in the field of

Scientometrics 42 (1998)

351

B~ M. GUPTA, C. R. KAR1SIDDAPPA: COLLABORATION IN THEORETICAL POPULATION GENETICS

biology and chemistry. Another study by Subramanyam 11 in the field of biochemistry and biochemical engineering, endorses Hafner's views. In both the fields, higher proportion of multi-authored or collaborative publications were supported by grants, than that of single-author publications. Pao 4 studied the schistosmiosis research, found a strong interrelationship between co-authorship, funding, and productivity with two different types of co-authors: (i) highly productive "global researchers" who receive research funds themselves or collaborate in a variety of funded groups or research projects, and (ii) locals, who are less often associated with funded research collaborate with only one group, and are less prolific. Gupta12 studied theoretical population genetics research speciality through a case study. He studied literature in this field appearing in a core journal "Genetics" from 1916-1980, and found a strong relationship between collaboration, funding, and productivity. He found that average productivity per author in collaborative research was 1.99 (n = 291), as against 1.84 (n = 173) in non-collaborative research, using normal count method. He also revealed that the average productivity per author was 2.15 in funded research as against 1.67 in non-funded research. Collaboration in research can take variety of paths, but necessarily takes the form either of cooperation between two researchers or organisations. Cooperation actually takes place between two researchers, but normally in practice we look at this collaboration at other levels - between research groups within a department, between departments within same institution, between institutions, between sectors, and between geographical regions and countries. Subramanyam 13 in a review publication has identified the following types of collaboration: (i) teacher-pupil collaboration, (ii) collaboration among colleagues, (iii) supervisor-assistant collaboration, (iv) researcher-consultant collaboration, (v) collaboration between organisations, and (vi) international collaboration. Qin 14 studied the research publications of The Philosophical Transactions of the Royal Society of London from 190I-I991, and further simplified the types o f collaboration as: (i) collaboration in a department within an institution, (ii) collaboration between two or more departments within an institution, (iii) collaboration between two or more institutions within a country, and (iv) international collaboration. Gupta12 studied the types of collaborative research on the pattern of classification suggested by Qin, by taking a sample of publications in theoretical popnlational genetics from an intemational journal "Genetics" from 1916-80. He noticed a shift in the nature and type of collaborative research over the years.

352

Scientometrics 42 (1998)

B. M. GUPTA, C. R. KARISIDDAPPA: COLLABORATION IN THEORETICAL POPULATION GENETICS

Table 1 Different levels of collaboration and distinction between inter and intra forms Intra

Inter

Individual

-

Between individuals

Group

Between individuals in the same research group Between individuals or groups in the same department Between individuals or departments in the same institution Between institutions in the same sector Between institutions in the same country

Between groups (e.g. in the same department) Betweem departments (in the same institution) Between institutions

Department Institution Sector Nations

Between institutions in different sectors Between institutions in different countries

K a t z and M a r t i n 15 points put out that besides distinguishing several different levels o f collaboration, we also need to recognise that collaboration can occur either between or within different levels. For simplicity they use the prefixed inter and intra respectively. According to them, the various levels o f collaboration, o f both the inter and intra forms, are summarised in Table 1. Collaboration can also be either homogeneous (i.e. unambiguously either the inter or the intra form o f collaboration) or heterogeneous (i.e. a mixture o f the inter and intra forms o f collaboration). A m o n g the types o f collaboration explained, international collaboration has been studied most extensively.

Aims and objectives The main objectives o f this study are: (i) to study the trends in funded and collaborated research in theoretical population genetics research, and (ii) to study the impact o f funding and collaboration on the productivity o f authors in the area o f theoretical population genetics research; and (iii) to study the trends and shifts in nature and type o f collaborative research over the years.

Database and methodology The data for this study has been taken from the research publications appearing in 11 core journals in theoretical population genetics. These journals are A m e r i c a n

Scientometrics 42 (1998)

353

B. M. GUPTA, C. R. KARISIDDAPPA: COLLABORATION IN THEORETICAL POPULATION GENETICS

Journal of Human Genetics, American Naturalist, Annals of Human Genetics, Biometrics, Evolution, Heredity, Genetical Research, Genetics, Journal of Theoretical Biology, and Theoretical and Applied Genetics. These joumals were identified as core journals on the basis of their contribution in the field and for their contribution in the Bibliography of Theoretical Population Genetics, 16 compiled by Felsenstein in 1981. The duration of the study was from 1956 to 1980. Since the sample was small, therefore, the publications in five years block were cumulated beginning in 1956 were taken as sample. There are now five block of publications pertaining to periods: 1956q50, 1961-65, 1966-70, 1971-75, and 1976-80. In each block, the total number of publications and authors involved were recorded according to the kind of collaboration and the availability of funding. Each publication has been classified under the following types of classification: (a) Intra-departmental collaboration: collaboration within a department (in an organisation or institution); (b) Inter-departmental collaboration: collaboration between two or more departments (in an organisation or institution); (c) Inter-institutional collaboration: collaboration between two or more organisations or institutions (located in different cities or in the same city within a country); and (d) International collaboration: collaboration between two or more organisations or institutions (located in more than one country). For discussion, we have further simplified the above classification of the types of collaboration, and will refer now to generally three types of collaborative relationship: internal collaboration (type a and b taken together); domestic collaboration (type c), and international collaboration (type d).

Analysis In all, 2502 research publications were published in these 11 core journals of theoretical population genetics, of which 1474 research publications (468 of which are funded) are single-authored, and 1028 research publications (667 of which are funded) are multi-authored. A chronological breakdown of these publications indicate that 34.61 per cent pertain to 1976-80. For previous periods, the percentage contribution in the total sample were: 29.05 per cent for 1971-75, 20.58 per cent for 1966-70, 10.47 per cent for 1961-65, and 5.27 per cent for 1956-60. In terms of geographical distribution, as shown in Table 5, the largest number of research publications, 1325 came from United States, contributing 52.95 per cent to the total sample. The next

354

Scientometrics 42 (1998)

B. M. GUPTA, C. R. KARISIDDAPPA: COLLABORATION 1N THEORETICAL POPULATION GENETICS

important group of countries in terms of contribution are: United Kingdom, Australia, Japan, Canada, Israel, Germany, and France contributing 20.78 per cent, 8.19 per cent, 3.91 per cent, 2.83 per cent, 1.83 per cent, 1.56 per cent, and 1.48 per cent, each respectively in the total sample. Rest of the contributions have come from 35 other countries. A frequency distribution of the number of publications and the number of countries producing them is shown in Table 2. For simplification in the analysis, we have divided the total sample in two groups: A and B. In Group A, we have included United States (USA), as it is a major contributor in the total output. In Group B, we have included all other 42 countries, which together have contributed 47.04 per cent to the total output. Table 2 Number of research publications contributed by various countries Number of publications

Number of countries

Total publications

1325 507 205 96 72 46 39 37 26 22 20 12 8 7 6 5 4 3 2 1

1 1 1 1 1 1 1 1 1 1 1 1 1 2 1 2 4 5 10 6

1325 507 2O5 96 72 46 39 37 26 22 20 12 8 14 6 10 16 15 20 6

Total

43

2502

Growth in funded of research Table 3 provides data on the growth of total research publications and funded research publications from block years 1956-60 to 1976-80. In the total sample which

Scientometrics 42 (1998)

355

B. M. GUPTA, C. R. KARISIDDAPPA: COLLABORATION IN THEORETICAL POPULATION GENETICS

comprise of 2502 publications, 41.88 per cent publications were observed to be funded. In ease of United States, United Kingdom, Australia, Japan, Canada, France, Germany, Israel, and India, the percentage of funded publications were 61.05 per cent, 21.15 per cent, 16.58 per cent, 56.12 per cent, 45.07 per cent, 28.94 per cent, 12.82 per cent, 52.17 per cent, and 15.00 per cent, each respectively. The number of funded research publications have not systematically increased with the growth of total research publications, but however, there exists a strong correlation between the two, the correlation coefficient being 0.9987. The percentage of funded research publications were 12.66 per cent during 1956-60, and it has increased to 51.50 per cent during 1976-80, the mean average percentage increase being 38.91 per cent. In Group A (United States), the number of funded publications was observed to be much higher than in Group B (Other Countries). The percentage of funded publications, on one hand, have increased from 22.41 per cent during 1956-60 to 67.21 per cent during 1976-80 in Group A, while in Group B, it has increased from 12.16 per cent during 1956-60 to 30.10 per cent during 1976-80. The mean average percentage of funded publications in Group A and Group B are 53.16 per cent and 23.47 per cent, respectively. There is a positive correlation between the growth of research publications and funded research publications in both Group A and Group B, the correlation coefficient being 0.9985 and 0.9851. Table 3 Growth of total research publications and funded research publications Period

1956-60 1960 65 1966-70 1971 75 1976 80

Total

Group A

Group B

NP

NFP

NP

NFP

NP

NFP

132 262 515 727 866

22 87 237 343 446

58 131 263 379 494

13 72 I54 236 334

74 131 252 348 372

9 15 83 107 112

NP = Number of publications. NFP = Number of funded publications.

Table 4 provides data on the growth of total authors and funded authors from block years 1956-60 to 1976-80. The number of funded authors has systematically increased with the total authors. The percentage of funded authors in total authors has increased from 33.58 per cent during 1956-60 to 65.17 per cent during 1976-80, the mean averge percentage increase being 52.96 per cent.

356

Scientometrics 42 (1998)

B. M. GUPTA, C. R. KARISIDDAPPA: COLLABORATION IN THEORETICAL POPULATION GENETICS

Table 4 Growth of total authors and funded authors Period

TA

TFA

PFTA

1956-60 1961-65 1966-70 1971-75 1976-80

134 197 382 504 689

45 94 233 289 449

33.58 47.71 60.99 57.34 65.t7

TA = Total authors; TFA = Total number of funded authors; PFTA = Percentage of funded authors in total authors.

Growth in collaborated research

The growth in the proportion of collaborated (co-authored) publications in the total publications, as shown in Table 6, provides an index of growing professionalisation in the research speciality. Of the total 2502 research publications in the sample, 41.08 per cent (1028 research publications) are observed to be co-authored. Studying the extent of co-authored publications in different countries as shown in Table 5, we notice that Table 5 Number of single-authored and multi-authored publications Country

USA UK Australia Japan Canada Israel Federal Republic of Germany France N e w Zealand Italy India

TP

SAP

MAP TP

POT

1325 507 205 96 72 46

702 344 157 58 41 23

623 163 48 38 31 23

47.02 32.15 23.41 39.58 43.05 50.00

39 37 26 22 20

29 25 14 9 7

10 12 12 13 13

25.64 32.43 46.15 59.05 65.00

TP = Total publications; SAP = Single-author publications; MAP = Multi-authored publications; POT = Percentage of total publications.

Scientometrics 42 (1998)

357

B. M. GUPTA, C. R. KAR1S1DDAPPA: COLLABORATION IN THEORETICAL POPULATION GENETICS

the percentage of co-authored publications in United Kingdom, Australia, Japan, Canada, France, Germany, Israel, and India were 47.02 per cent, 32.15 per cent, 23.41 per cent, 39.58 per cent, 43.05 per cent, 32.43 per cent, 25.64 per cent, 50.00 per cent, and 65.00 per cent, each respectively. The percentage of co-authored publications in the total sample has increased from 36.36 per cent during 1956-60 to 45.15 per cent during 1976-80, the mean average percentage increase being 8.31 per cent. The percentage of co-authored publications in Group A has increased from 44.82 per cent during 1956-60 to 51.22 per cent during 1976-80, the mean average percentage increase being 4.46 per cent. The percentage of co-authored publications in Group B has increased from 29.72 per cent during 1956-60 to 37.09 per cent during 1976-80, the mean average percentage increase being 16.15 per cent. A positive correlation exists between the growth of total publications and collaborated publications in the total sample, as well as in both Group A and Group B. Table 6 Growth of total publications and collaborated publications Period

1956-60 1961-65 1966-70 1971-75 I976-80

Total

Group A

Group B

NP

NCP

NP

NCP

NP

NCP

132 262 515 727 866

48 72 204 313 391

58 131 263 379 494

26 48 116 180 253

74 131 252 348 372

22 24 88 133 138

NP = Number of publications. NCP = Number of collaborated publications.

Table 7 Growth of total authors and collaborated authors Period

Total authors

Total collaborated authors

PCTA

1956-60 1961-65 1966-70 1971-75 1976-80

134 197 382 504 689

83 111 276 363 523

61.94 56.34 72.25

72.02 75.91

PCTA = Percentage of collaborated authors in total authors.

358

Scientometrics 42 (1998)

B. M. GUPTA, C. R. KARISIDDAPPA:COLLABORATIONIN THEORETICAL POPULATION GENETICS

Table 7 provides data on the growth of total authors and percentage of total authors involved in collaborated research in different block years. The percentage of authors involved in collaborated research has found to have increased from 61.94 per cent during 1956-60 to 75.91 per cent during 1976-80, the mean average per cent increase being 6.22 per cent.

Extent of co-authorship Out of the total sample of 2502 publications, 1474 publications (58.91%) are observed to be single-authored, 752 publications (30.05%) two-authored, 205 publications (8.19%) three-authored, 49 publications (1.95%) four-authored, and 22 publications (0.88%) having more than four authors per publication. Table 8 shows data of the extent o f single-authored and multi-authored publications in the total sample from block year 1956--60 to 1976-80. The share of single-authored publications have decreased from 72.51 per cent during 1961-65 to 55.88% during 1976-80. In contrast, the share of multi-authored publications have increased. The percentage share of: twoauthored publications have increased from 23.28% during 1961-65 to 28.75% during 1976-80, three-authored publications from 3.81% during 1961-65 to 11.08% during 1976-80, and four-authored publications from 0.40% during 1961-65 to 4.29% during 1976-80. The distribution of number of authors per publication is observed to b e erratic during 1956-60, compared to other block years. Table 8 Trends in growth of multi-authored publications NAPP

1 2 3 More

Share of publications (in percentages) 1956-60

1961-65

1966-70

1971-75

1976-80

64.39 26.51 6.81 2.27

72,51 23.28 3.81 0.40

60.58 30.87 6.79 1.76

60.24 31.36 6.19 2.21

55.88 28.75 11.08 4.29

NAPP = Number of authors per publication.

Impact of collaboration and funding on productivity To what extent the productivity o f scientists are enhanced through collaboration and availability o f funding is studied in this section. There are three'methods of measuring the productivity of scientists: normal count method, where every author or scientist

Scientometrics 42 (1998)

359

B. M. GUPTA, C. R. KARISIDDAPPA: COLLABORATION IN THEORETICAL POPULATION GENETICS

participating in writing a joint publication is assigned an equal weightage one; straight count method, where only first author in a joint authored research publication gets a weightage one; and fractional counting method, where every author is assigned 1/n weightage in a n-authored publication. In the analysis, we have used normal count i methodTor measuring the productivity of scientists. Table 9 Growth in the number of publications and authors Period

Number of publications

Number of authors

1956-60 1961-65 1966 70 1971-75 1976-80

132 262 515 727 866

136 197 382 504 689

Table 9 presents data on the number of publications and the number of authors involved in producing these publications from block years 1956-60 to 1976-80. It is observed that with the increase in number of research publications, there is not a proportional increase in the number of authors involved with time. However, there is a pos!tiye correlation between the number of publications and authors over the years. Table 10 presents data on the productivity distribution of all authors in each of the five block years. With the increase in the number of publications, there is not a proportional increase in the number of new authors. As a result, we find that simultaneously with the entry of new authors in each block year, the productivity of existing authors increases and get strengthened with time. As a consequence, the average productivity per author is observed to have increased from 1.75 during 1956-60 to 2.19 during 1976-80. :In order to study the effect of funding on productivity of authors, we have subdivided the total authors in each of the five block years into two further subgroups: funded authors and non-funded authors, and the distribution of their productivity is presented in Table 11, and the calculated average productivity per author in Table 12. From the results as presented in Table 11, it is clear that with the increase in the number of funded and non-funded publications, the number of authors involved in funded research publications has proportionately increased much faster than the number of authors involved in non-funded research publications. We also observe that average productivity per author is greater in funded authors subset and smaller in non-funded authors subset, than the average productivity per author for total authors subset for all

360

Scientometrics 42 (1998)

B. M. GUPTA, C. R. KARISIDDAPPA: COLLABORATION 1N THEORETICAL POPULATION GENETICS

the five block years. The averge productivity, on one hand, has systematically increased with time from block years 1956-60 to 1976-80 in case o f funded authors subset, while in case o f non-funded authors subset, it has declined, with no clear trend for the same period. Table 10 Distribution o f productivity o f total authors from 1956-60 to 1976-80

NP 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 22 23 24 26 30

A 88 25 11 3 1 3 1

N u m b e r o f authors (NA) Group B C D 128 35 14 10 4 2 1 1 1 1 2

231 75 22 14 10 10 5 3 5 2 1 2 I

E

325 79 33 25 9 8 3 4 3 5 2 3 1

1

448 110 48 15 14 13 4 5 5 7 3 3 2 2 2 2

1 2 1 1 1 1 1

T

136

197

382

504

689

AP

1.75

1.76

2.09

2.16

2.19

NP = N u m b e r o f publications; T = total; AP = Averge productivity per author; N A = Number of authors. Group A = 1956-60; Group 13 = 1961-65; Group C = 1966-70; Group D = 1971-75; and Group E - 1976-80.

Scientometrics 42 (1998)

361

B. M. GUPTA, C. R. KAR1SIDDAPPA: COLLABORATION IN THEORETICAL POPULATION GENETICS

Table 11 Distribution o f productivity o f funded and non-funded authors Funded authors Group NP

A

Non-funded authors Group

B

C

D

E

A

B

C

D

E

1

27

51

134

172

240

61

77

97

153

208

2

8

21

45

45

89

17

14

30

34

21

3

6

9

14

18

46

5

5

8

15

2

4 5

1

7 2

8 8

18 6

12 12

3 3

3 2

6 2

7 3

3 2

6 7 8

1 1 1

8 3 3

7 3 4

11 4 5

1 1

1

2 1

1

2 1

2 1

4 2

3

4 7

1

1

1

4

9 10 11

1

12

1

13 14

1

15 16

3

1

2 2 2

3 1

1

2 2

1

17

1

1 1 1

2

18 19

1 2

22

1

23 24

1 1

26

1

30

1

T

45

94

233

289

449

91

103

149

215

240

AP

1.95

2.02

2.36

2.57

2.67

1.64

1.52

1.75

1.57

1.30

NP = Number of papers; T = Total; AP = Averge productivity per author. Group A = 1956-60; Group B 1961-65; Group C = 1966-70; Group D = 1971-75; and Group E = 1976-80.

362

Scientometrics 42 (1998)

B. M. GUPTA, C. R. KARISIDDAPPA:COLLABORATIONIN THEORETICALPOPULATIONGENETICS

Table 12 Average productivity per author of total authors, funded and non-funded authors Average productivity per author Total Funded Non -funded authors authors authors

Period

1956-60

1.75 (n=136) 1.76 (n=197) 2.09 (n=382) 2.16 (n=504) 2.19 (n=689)

1961-65 1966-70 1971-75 1976-80

1.95 (n=45) 2.02 (n=94) 2.36 (n=233) 2.57 (n=289) 2.67 (n=449)

1.64 (n=91) 1.52 (n-103) 1.75 (n=149) 1.57 (n=215) 1.30 (n=240)

Table 13 Productivity distribution of collaborated and non-collaborated authors Collaborating authors

NP

Group C

Non-collaborating authors

B

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 22 23 24 26 30

56 11 7 3 2 1 2 1

69 24 7 6 2

T

83

111

276

363

523

53

86

106

141

166

AP

1.79

1.80

2.26

2.34

2.42

1.67

1.70

1.57

1.65

1.46

1 1

160 51 17 12 8 9 5 3 4 2 1

D

E

223 57 29 17 8 7 3 3 3

324 84 39 13 12 12 4 4 4 7 3 2 3 2 2 2 2

4

1 I

2 2 1

1

A

Group C

A

B

32 14 4 2 1

59 11 7 4 2 2 1

71 24 5 2 2 1 1

1

1

D

E

102 22 4 8 1 1 1

124 26 9 2 2 1 1 1

1 2 1 1 1 1 1

NP = Number of papers; T = Total; AP = Averge productivity per author. Group A=1956-60; Group B = 1961-65; Group C = 1966-70; Group D = 1971-75; and Group E = 1976-80.

Scientometrics 42 (1998)

363

13. M. GUPTA, C. R. KARISIDDAPPA: COLLABORATION IN THEORETICAL POPULATION GENETICS

Table 14 Averageproductivityper author of total authors, collaborated, and non-collaboratedauthors Period

1956-60 1961-65 1966-70 1971-75 1976-80

Averageproductivityper author Total Collaborated Non-collaborated authors authors authors 1.75 (n=136) 1.76 (n=197) 2.09 (n=382) 2.16 (n=504) 2.19 (n=689)

1.79 (n=83) 1.80 (n=l 11) 2.26 (n=276) 2.34 (n=363) 2.42 (n=523)

1.67 (n=53) 1.70 (n=86) 1.57 (n=106) 1.65 (n=141) 1.46 (n= 166)

In order to study the effect of collaboration on productivity, we have subdivided the total authors in each of the cumulative five block years in two subgroups: collaborated authors and non-collaborated authors, and their distribution of productivity is presented in Table 13, and calculated average productivity per author in Table 14. With the increase in the number of collaborated and non-collaborated publications from 1956-60 to 1976-80, the number of authors involved in collaborated research has proportionally increased much faster, than the number of authors involved in non-collaborated research. The average productivity per author is again observed to be larger in the collaborated authors subset, and smaller in the non-collaborated authors subset, than the average productivity per author for the total authors subset. The average productivity per author, on one hand, has systematically increased in collaborated authors subset from block years 1956-60 to 1976-80, and on the other, it has decreased in the noncollaborated authors subset, with no clear trend, for the same block years. Now we have studied further the distribution of productivity of funded and nonfunded authors within the collaborated authors subset and data on this aspect has been presented in Table 15, and the calculated value of average productivity per author in Table 16. It is observed that the number of authors involved in collaborated-funded research have increased faster than authors involved in collaborated non-funded research as we move from block year 1956-60 to 1976-80. As a result we find that the average productivity per author is higher in collaborated-funded authors subset and smaller in collaborated non-funded authors subset, than the average productivity per

364

Scientometrics 42 (1998)

B. M. GUPTA, C. R. KARISIDDAPPA:COLLABORATION1N THEORETICALPOPULATIONGENETICS

author for total collaborated authors subset. There is an increasing trend in the average productivity per author in the collaborated funded authors subset, and a decreasing and increasing trend in the collaborated non-funded authors subset with time. Table 15 Distribution of productivity of collaborated funded and collaborated non-funded authors Collaborated and funded authors

NP l 2 3 4 5 6 7 8 9 l0 11 12 13 14 15 16 17 18 19 22 23 24 26 30

A 23 6 5

B

1

39 18 6 4 1

2 1

1

1

Group C 94 40 14 11 8 9 5 3 4 2 1 2 l

D 143 38 18 11 6 5 3 3 3 4 2 2

1 1

Collaborated and non-funded authors

E

A

B

221 64 32 12 12 10 4 4 4 7 3 2 3 2 2 2 2

33 5 2 3 1 1

30 6 1 2 1

Group C 66 11 3 1

D 80 19 11 6 2 2

E 103 20 7 1

l

1 2 1 1 1 l 1

T

38

70

195

243

390

45

41

81

120

133

AP

2.02

1.90

2.73

2.68

2.79

1.60

1.63

1.24

1.64

1.35

NP=Number of papers; T=Total; AP=Averge productivity per author. Group A=1956-60; B=1961-65; Group C=1966-70; Group D=1971-75; and Group E=1976 80.

Scientometrics 42 (1998)

Group

365

B. M. GUPTA,C. R. KARISIDDAPPA:COLLABORATIONIN THEORETICALPOPULATIONGENETICS

Table 16 Average productivity per author of total collaborated, collaboratedfunded and collaborated non-funded authors Period

Average productivity per author Total Collaborated Collaborated collaborated funded non-funded authors authors authors

1956-60 1961-65

1966-70 1971-75 1976-80

1.79 (n=83) 1.80 (n=ll l) 2.26 (n=276) 2.34 (n=363) 2.42 (n=523)

2.02 (n-38)

t.60 (n=45)

1.90

1.63

(n=70) 2.73 (n=195) 2.60 (n=243) 2.79 (u=390)

(n=41) 1.24 (n=81) 1.64 (n= 120) 1.35 (n=133)

Nature and type of collaboration lnternal collaboration

Out of total 543 internal collaborated publications, 347 publications (63.90%) are contributed by United States, 95 publications (17.49%) by United Kingdom, 29 publications (5.34%) by Japan, 20 publications (3.68%) by Australia, and 11 publications (2.02%) by Canada, and the rest of the countries have contributed less than 2%. Internal collaborative publications occupy a major percentage share in total collaborative publications of some of the major producing countries. Its percentage contribution in total collaborative publications lies between: 71%-80% in Germany and Japan, 51 ~ in United States, United Kingdom, and India, 41~ in Australia, 31-40% in Canada, Italy, and France, and 11%-20% in Israel. The reason why this collaboration is dominant because it requires the least resources to collaborate. Domestic collaboration

Compared to other types of collaboration, priority to domestic collaborative publications are given only in few countries. Out of total 257 domestic collaborated publications, 181 publications (70.43%) are contributed by United States, 31 publications (12.06%) by United Kingdom, 11 publications (4.28%) by Australia, 8

366

Scientometrics 42 (1998)

B. M. GUPTA, C. R. KARISIDDAPPA: COLLABORATION 1N THEORETICAL POPULATION GENETICS

publications (3.11%) by Israel, 7 publications (2.72%) by Canada, and rest of the countries below this percentage. In individual countries, the priority to this type of collaborative publications is not very high. Its percent contribution in total collaborative publications lies between: 31%-40% in Israel, 21%-30% in United States, Italy, Canada, France, and Australia, 11%,20% in United Kingdom, and Japan, 1%-10% in New Zealand and 0% in Germany and India. In recent years, science and technology policies of many countries are focusing on the need to do active collaborative work of this type to economise on resources, facilities, and human resources, and also encourage increasing domestic collaboration, with same or different objectives. The United Kingdom, for example, encourages universities and research institutions to work more closely with industry, to generate applications of research efforts carried out in the country. The European Union, on the other hand, suggests and encourages collaboration between institutions and organisations Iocated in more- and less-favoured regions in the research efforts it supports. 17 International collaboration

International collaboration is increasingly becoming more frequent over a period of time, and playing an important role in production of scientific knowledge across the world. There are many factors both, internal and external, to the scientific enterprise, that stimulate international collaboration. The internal factors are cognitive and social, while external factors are economic and political. International collaboration is motivated by many intra-scientific factors, such as the desire to increase knowledge, exchange skills and data, and enhance professional development. It is further facilitated by decreasing cost of communication (travel, fax, and e-mail). Besides government support for international contacts and facilities, bilateral and multilateral agreements between countries, and initiatives of the international agencies have further encouraged collaboration. 18 The number of internationally collaborated publications world-wide has grown from 7 publications during 1956-60 (14.58% of total) to 83 publications (21.22% of the total), with a mean annual average growth rate of 117.04%. United States The United States contributed 41.66% to the world's total internationally coauthored publications. It is among the countries with a significant percentage of coauthorship (47.02%), but it ranks in the low end in terms of overall percentage of

Scientometrics 42 H998)

367

B. M. GUPTA, C. R. KARISIDDAPPA: COLLABORATION IN THEORETICAL POPULATION GENETICS

articles involving international collaboration (11.53% during 1956-60 to 15.81% during 1976-80). This occurred solely as a result of its large publication base. The United Kingdom contributed 16.23% to the world's international co-authored publications. It is among the countries with 32.15% of its total publications appearing as co-authored publications, but it ranks in lower end in terms of its overall percentage of publications involving international collaboration (6.25% during 1956-60 to 24.13% during 1976-80). Australia contributed 7.45% to the world's total international coauthored publications. Its percentage of co-authored publications are 23.41%, but its contribution in terms of publications involving international collaboration have decreased (from 50% during 1956-60 to 33.33% during 1976-80. Japan contributed 2.19% to the world's total international co-authored publications and its percentage of co-authored publications are 39.58%. Canada contributed 13.59% to the world's total co-authored publications, and its percentage of co-authored publications are 43.05%. Scientific collaboration as measured by international co-authorship, centres to a remarkable degree on United States. 19 In 41.66% (95 publications) of all international co-authored publications, United States authors collaborated with authors from 15 countries. The largest number of international co-authored publications of United States involve collaboration with: Japan (22 publications out of 95), United Kingdom (15 publications), Israel (14 publications), Australia (8 publications), New Zealand (7 publications), Canada (6 publications), Italy (5 publications), and Denmark (4 publications), so on. Authors participating in nearly 32.89% of all international co-authored publications from other countries collaborate with United States authors. From the total international co-authored publications combined from 1956-60 to 1976-80, United States authors participated and collaborated in: 100% of the international co-authored publications of Japan, in 90.90% in Israel, in 85.71% of New Zealand, in 66.66% of Italy, in 64.70% of Australia, in 54.05% of United Kingdom, in 50% of India and Denmark, in 38.46% of Canada, and in 20% of France, as shown in Tables 17 and 18. Other countries Out of total collaborated publications, the number of international co-authored publications are: 37 publications (16.22%) of United Kingdom, 17 publications (5.70%) of Australia, 13 publications (5.70%) of Canada, and 11 publications (4.82%) of Israel, as shown in Table 17. This means that authors from other countries collaborated with these countries authors in these publications.

368

ScienWmetrics 42 (1998)

B. M. GUPTA,C. R. KARISIDDAPPA:COLLABORATION1NTHEORETICALPOPULATIONGENETICS

Similarly, in 35 publications (15.35%), 16 publications (7.01%), 13 publications (5.70%), 16 publications (7.01%), 24 publications (10.52%) and 11 publications (4.82%) from other countries collaborate with authors from United Kingdom, Australia, Canada, Israel, Japan and New Zealand, as shown in Table 18. From the total sample of collaborated publications from 1956-60 to 1976-80 taken together, United Kingdom authors collaborated in: 15.78% (15 out of 95) of all international co-authored publications of United States, in 17.64% (3 out of 17) of Australia, in 30.76% (4 out of 13) of Canada, in 9.09% (1 out of 11) of Israel, in 16.66% (1 out of 6) of Italy, in 20% (1 out of 5) of France, and 100% (2 out of 2) of Federal Republic of Germany, as shown in Table 17. Similarly, Australian authors collaborated in 8.51% (8 out of 95) of all international co-authored publications of United States, in 13.51% (5 out of 37) of United Kingdom, and in 14.28% (1 out of 7) of New Zealand. Canadian authors, similarly, collaborated in 9.57% (9 out of 95) and 5.40% (2 out of 37) of all international co-authored publications of United States and United Kingdom. Japanese and Israeli authors, similarly, collaborated in 23.40% (22 out of 95) and 15.95% (15 out of 95) of all internationally co-authored publications of United States. Table 17 Collaboration linkages made with other countries Source country

NICP

USA

95

JPN(22), ISR(15), UK(I 5), CAN(9), AUS(8), NZL(8), ITA(4), DNK(4), BRA(2), MEX(2), FRG(2), etc.

UK

37

USA(20),AUS(5),CAN(2),NOR(2),JPN(1),NED(I)

AUST CAN ISR NZL IND JPN ITA FR FRG

17 13 11

7 6 6 6 5 2

Collaborating countries

POE(l), IND(1), NGR(1), SDN(1), MEX(1), SAFR(I) USA(11), UK(3), ISR(1), NZL(1), NGR(I) USA(5), UK(4), NZL(2), FRG(a), FRA(I) USA(10), UK(I) USA(6), AUS(1) USA(3), GDR(2), SWITZ(1) USA(6) USA(4), UK(1), NLD(1) USA(l), UK(I), BEE(3) UK(2)

NICP = Number of international co-authored publications or linkages.

Scientometrics 42 (1998)

369

B. M. GUPTA,C. R. KAR1SIDDAPPA:COLLABORATIONIN THEORETICALPOPULATIONGENETICS

Table 18 Collaboration linkages received from other countries

Collaborating country

Source countries

NCL

USA

75

UK

35

JPN AUS ISR CAN NZL DNK ITL

24 16 16 13 11 4 4

UK(20), AUS(11), ISR(10), JPN(6), ITA(4) CAN(5), NZL(4), IND(3), etc. USA(15), AUS(3), CAN(4), FRG(2), SWED(2), NOR(l), ITL(1), FRA(1), ISR(1), SPN(1), MEX(1) USA(22), UK(1), SRILANK(1) USA(g), UK(5), NZL(I), SWE(1) USA(15), AUS(1) USA(9), UK(2), SGP(1), GRC(1) USA(8), CAN(2), AUS(1) USA(4) USA(4)

NCL = Number of collaboration linkages received.

Priorities accorded to different type of collaborative publications From the data presented in Table 19, it is clear that although there is an increase in the number of internal collaborated publications fi'om 30 publications (62.59% of the total) during 1956-60 to 202 publications (51.66% of the total) during 1976-80, but in terms of percentage contribution there is a decrease from 62.59% to 51.66% from 1956-60 to 1976-80. In contrast, the number of domestic and international collaborated publications have increased from 11 publications (22.91% of the total) and 7 publications (14.58 % of the total) during 195 6-60 to 106 public ations (27.10%) and 83 publications (21.22% of the total) during 1976-80, showing an increasing trend in the percentage share in both the cases. Table 19 Change in different type of collaborative publications with time Period

NCP

1956-60 1961-65 1966-70 1971-75 1976-80

48 72 204 313 391

Type of collaboration Internal Domestic International in percentage in percentage in percentage 62.50 56.94 56.37 49.52 51.66

22.9l 27.77 19.60 25.55 27.10

t4.58 15.27 24.01 24.92 21.22

NCP = Number of collaborated publications.

370

Scientometrics 42 (1998)

B. M. GUPTA,C. R. KARISIDDAPPA:COLLABORATIONINTHEORETICALPOPULATIONGENETICS

Table 20 Funded support available to differenttype of collaborativepublications with time Period

NCP

1956-60 1961-65 1966-70 1971-75 1976-80

48 72 204 313 391

Type of collaboration Internal Domestic International percentage of publications funded 36.66 56.09 56.52 52.25 62.37

18.18 50.00 72.50 75.00 77.35

57.14 72.72 71.42 75.64 86.74

NCP = Numberof collaborated publications. Studying the funded support available to different type of collaborative publications, we observe that it is increasing with time in all type of collaborative publications as shown in Table 20. In internal collaboration, the funded publications has increased from 36.66% during 1956-60 to 62.37% during 1976-80. In domestic and international collaboration, the funded publications have increased from 18.18% and 57.14% during 1956-60 to 77.35% and 86.74% during 1976-80. The comparative larger funding support is available to international collaborative publications from the beginning. Now we will focus on the changing priorities accorded to different type of collaborative publications in Group A and Group B. Based on the data presented in Table 21 and 22, we can make the following observations. (a) Internal collaborative publications occupy a significant percentage share of total collaborative publications in both groups. On one hand in Group A, its contribution has consistently decreased from 69.23% during 1956-60 to 53.35% during 1976-80, and on the other hand, its contribution fluctuated between 37.50% to 54.54% during the same period in Group B. (b) Domestic collaborative publications, on one hand, have consistently increased their percentage contribution from 19.23% during 1956-60 to 30.83% during 1976-80 in Group A, and on the other hand, its percentage contribution have decreased from 27.27% during 1956-60 to 20.28% during 1976-80, in Group B, with the exception of one block year 1961-65. (c) The percentage of international collaborative publications have shown an increasing trend in both the groups~ In Group A, its percentage contribution have increased from 11.53% during 1956-60 to 15.81% during 1976-80, while in Group B the increase for the same period was from 18.18% to 31.15%, a marginal increase in Group A.

Scientometrics 42 (1998)

371

B. M. GUPTA, C. R. KARISIDDAPPA:COLLABORATIONIN THEORETICALPOPULATION GENETICS

Table 21 Change in different type of collaborative publications in Group A with time Period

NCP

1956 60 1961-65 1966-70 1971-75 1976-80

26 48 116 180 253

Type of collaboration Internal Domestic International in percentage in percentage in percentage 69.23 66.66 57.76 52.77 53.35

19.23 18.75 23.27 34.44 30.83

11.53 14.58 18.96 12.77 15.81

NCP = Number of collaborated publications.

Table 22 Change in different type of collaborative publications in Group B with time Period

NCP

1956-60 1961-65 1966-70 1971-75 197680

22 24 88 133 138

Type of collaboration Internal Domestic International in percentage in percentage in percentage 54.54 37.50 67.04 45.11 48.50

27.27 45.83 14.77 13.53 20.28

18.18 16.16 30.68 41.35 31.15

NCP = Number of collaborated publications.

Now we will study the funding support available to different type of collaborative publications in Groups A and B. From the data presented in Tables 23 and 24, we observe that the number of funded publications have shown an increasing trend in both the groups from 1956--60 to 1976-80. We can, however, make the following observations: (a) In internal collaborative publications, the percentage of funded publications have increased from 50% to 72.59% in Group A and from 16.16% to 41.79% in Group B, from 1956-60 to 1976-80. (b) The percentage of funded publications in domestic collaborated publications have increased from 40% to 89.74% in Group A, and from 0% to 42.85% in Group B, from 1956-60 to 1976-80. The systematic increase in funded support to publications is obselwed only in Group A, but in Group B there are fluctuations in the growth.

372

Scientometrics 42 (1998)

B. M. GUPTA, C. R. KAR1SIDDAPPA:COLLABORATIONIN THEORETICALPOPULATION GENETICS

(c) A comparative higher percentage of funded publications are noticed in international collaborative publications. The percentage of funded publications has increased from 33.33% to 100% in Group A from 1956-60 to 1976-80, while in Group B, the percentage contribution has first declined and then came back to the same position for the same period. Table 23 Funded support available to different type of collaborative publications in Group A with time Period

NCP

1956-60 1961 65 1966 70 1971 75 1976-80

26 48 116 180 253

Type of collaboration Internal Domestic International percentage of publications funded 50.00 65.62 68.65 65.26 72.59

40.00 66.66 74.07 80.64 89.74

33.33 85.71 81.81 95.65 100.00

NCP = Number of collaborated publications.

Table 24 Funded support available to different type of collaborative publications in Group B with time Period

1956-60 1961-65 1966-70 1971-75 1976-80

NCP

22 24 88 133 138

Type of collaboration Internal Domestic International percentage o f publications funded t6.I6 22.22 41.66 31.66 41.79

0.00 36.36 69.23 55.55 42.85

75.00 50.00 62.96 67.27 74.21

NCP = Number of collaborated publications.

Conclusions

The analysis presented in this publication offers evidence to support the hypothesis that with the increase in the total number of publications and authors, there is not a proportional increase in the number of funded and collaborated research publications and authors and there is a strong correlation between the two. O f the total publications,

Scientometrics 42 (1998.)

373

B. M. GUPTA, C. R. KARISIDDAPPA: COLLABORATION IN THEORETICAL POPULATION GENETICS

41.88% and 39.76% publications are found to be funded and collaborated. The percentage of collaborated publications is not very high because the field is theoretical in nature. Among the single-authored and multi-authored publications, 33.64% and 64.88% publications are observed to be funded. Among the countries involved in theoretical population genetics, United States is the largest contributor with maximum number of publications, followed by United Kingdom, Australia, Japan, Canada, etc. The largest funding support was noticed in case of publications from United States, Japan, and Israel, where more than 50% of their publications were funded. Similarly, the countries where more than 40% of their total publications are collaborated are: United States, Canada, Israel, and India. When the extent of authorship was considered, it was observed that on one hand, there is a decrease in the percentage of single-author publications, and on the other hand, simultaneously there is an increase in the percentage of multi-authored publications with time. The percentage share of two-authored publications increased from 23.28% during 1961-65 to 28.75% during 1976-80; three-authored publications increased from 3.81% to 11.08%; four-authored publications from 0.40% to 4.29%, etc. for the same period. The impact of collaboration and funding on the productivity of authors was clearly visible in the data. Generally the authors who were found to be most productive are observed to be largely collaborative and funded. The average productivity per author is found to be generally higher in the subset of authors associated with funded and collaborated research and smaller in the subset of authors associated with nonCollaborative and non-funded research, then the average productivity per author for the total authors subset in all the five block years studied from 1956-60 to 1976-80. Also there is a consistent increase in the average productivity per author in the funded and collaborated authors subset as one moves from time block 1956-60 to 1976-80. The average productivity per author in the funded and collaborated authors subset has increased from 1.95 and 1.79 during 1956-60 to 2.67 and 2.42 during 1976--80. United States has been the largest contributor in all type of collaborative publications. Domestic collaboration, in particular, has been observed and encouraged maximum in United States, and is seen as an important policy instrument for regional development and for innovative research in the field. Scientific collaboration, as measured by international co-authorship, centres to a remarkable degree on United States. On one hand, we observed that 41.66% (95 out of 228) of all international collaborated publications are from United States, and these publications involves collaboration with authors from 15 other countries. United States important and significant collaborated partners, as reflected in co-authorship, are Japan,

374

Scien tometrics 42 (1998)

B. M. GUPTA, C. R. KAILISIDDAPPA:COLLABORATIONIN THEORETICAL POPULATION GENETICS

UK, Israel, Australia, etc. On the other hand, United States authors attract collaboration in 56.39% (75 out of 133) of all international collaborated publications of other countries. United States authors collaborated in 70% of all internationally collaborated publications of Japan, Israel, and New Zealand. The analysis also clearly bring forth the changes taking place in the nature of collaborative research over the years. A distinctive shift is observed to be taking place from internal collaborative publications to domestic and international collaborative publications, supported by provision of increasing funds from government agencies. Science policies of various countries are promoting domestic collaboration in particular to meet their policy objectives.

References 1. D. DE. B. BEAVER, R. ROSEN, Studies in scientific collaboration 1. The professional origions of scientific authorship, Scientometrics, 1 (1978) 65-84. 2. D. DE. B. BEAVER, R. ROSEN, Studies in scientific collaboration 2. Scientific co-authorship, research productivity, and visibility in the French scientific elites, Scientometrics, I (1978) 133-149. 3. D. DE. B. BEAVER, R. ROSEN, Studies in scientific collaboration 3. Professionalisation and the natural history of scientific co-authorship, Scientometrics, 1 (1978) 231-245. 4. M. L. PAO, Global and local collaboration: A study o f scientific collaboration, Information Processing and Management, 28 (1992) 99-109. 5. D. J. DE SOLLA PRICE, D. DE B. BEAVER,Collaboration in an invisible college, American P.wchologist, 21 (1966) 1000-1018. 6. H. A. ZUCKERMAN,Noble laureates in science: Patterns of productivity, collaboration and authorship, American Sociological Review, 32 (1967) 391-403. 7. M. L. PAO, Collaboration in computational musicology, Journal of the American Society for lnfiJrmation Science, 33 (1982) 38-43. 8. N. PRAVDIC, V. OLUIC-VUKOVIC, Dual approach to multiple authorship in the study of collaborator/scientific output relationship. Scientometrics, 10 (1986) 259-280. 9. D. J. DE SOLLA PRICE, Citation measures of hard science, soft science, technology and non-science. In C. E. NELSON, D. K. POLLAK (Eds), Communication among Scientists and Engineers, Lexington, M. A.: D. C. Health, 1970, p. 3-22. 10. A. G. HAFNER, Funded research, multiple authorship, and sub-authorship collaboration in four disciplines, Scientometrics, 3 (1981) 5-12. l 1. K. S. SUBRAMANYAM, E. M. STEPHENS, Research collaboration and funding in biochemistry and chemical engineering, International Forum on Information and Documentation, 7 (1982) 26-29. 12. B. M. GUPTA, S. KUMAR, C. R. KARISIDDAPPA,Collaboration profile of theoretical population genetics speciality, Scientometrics, 39 (1997) 293-314. 13. K. SUBRAMANYAM,Biblionaetric studies of research collaboration: A review, Journal of Information Science, 6 (1983) 33-38. 14. J. QIN, An investigation of research collaboration in the sciences through the Philosophical Transactions, Scientometrics, 29 (l 994) 219-238. 15. J. SYLVANKATZ, BEN R. MARTIN, What is research collaboration? Research Policy, 26 (1997) 1 18.

Scientometrics 42 (1998)

375

B. M. GUPTA, C. R. KARISIDDAPPA: COLLABORATION 1N THEORETICAL POPULATION GENETICS

16. J. FELSENSTEIN, Bibliography o f theoretical population genetics. Pennsylvania; Dowen, Hutchinson & Ross, Inc. 1981. 17. DIANA HICKS, J. SYLVAN KAT'Z, Science policy for a highly collaborated science system. Science and Public Policy, 23 (February 1996) 39-44. 18. P. S. NAGPAL, LAL1TASHARMA, Research output and transnational cooperation in physics subfields: A multidimensional analysis. Scientometrics, 31 (1994) 97-122. 19. NATIONAL SCIENCE BOARD, Science and engineering indicators - 1996. Washington, DC.; U.S. Government Printing Office, 1997.

376

Scientometrics 42 (1998)