Scientometrics

, Volume 86, Issue 1, pp 211–226

Scientific publications of engineers in South Africa, 1975–2005

Authors

    • Sociology Programme, Howard CampusUniversity of KwaZulu-Natal
Article

DOI: 10.1007/s11192-010-0288-3

Cite this article as:
Sooryamoorthy, R. Scientometrics (2011) 86: 211. doi:10.1007/s11192-010-0288-3

Abstract

In the production of scientific knowledge, as revealed by publication output, South Africa is at the forefront of many other countries in the developing world and in the African continent. This study examines for the first time the publication trends of South African engineering researchers for a period of 30 years since 1975. Drawing data from the ISI Web of Knowledge, this paper specifically looks at the publication patterns of engineering researchers in South Africa.

Keywords

EngineeringScientific productionCollaborationSouth Africa

Introduction

Engineering and its variants are key fields of the modern economy and global trade (Bengisu 2003). The roles engineers assume in the local and global economy is as vital as ever (Lister and Donaldson 2003). In the domain of engineering South Africa has an impressive history. This record is believed to be a result of South Africa’s response to supplement the needs of the major industries of diamond, gold and coal mining located in the country and their growth has been reassured by basic studies and applied knowledge of several branches of engineering (SAJS 2006). South Africa’s engineering profile compares well with that of some European countries, particularly on the innovation front (Rooks et al. 2005). When 44% of the firms in South Africa introduced technological innovations in their respective field it was 46% for the firms in the EU, and South Africa’s share was higher than other countries such as France, Belgium, and Spain (Rooks et al. 2005).

There are a few studies that relate to the area of engineering science. Analyzing the PASCAL database for a 5-year period between 1991 and 1996, Arvanitis et al. (2000) reported a surprising increase in publications in engineering science mainly because of the expansion of engineering sciences in north Africa. Tsay (2009) conducted a scientometric study of journals including engineering using statistical procedures to analyze citations and the impact factor of publications. This analysis showed the differences between scientists in engineering, and physics and chemistry in their citation behaviour. Mapping the collaboration linkages between engineers, organizations and countries in oceanographic engineering research, Dastidar and Ramachandran (2005) revealed the linkages between productivity and socioeconomic imperatives but without offering any concrete relationship. Rao and Suma (1999) looked at the Indian engineering research for a short period and elaborated on the variation in the growth of output within the sub-branches of engineering. In certain areas of technological innovation South Africa ranks top at the international level, as Pouris (2009) indicated. Pouris (2009) further noted that South Africa has a remarkable record of producing patents in the last 15 years. This, in other words, marks development in engineering. Most remarkably, as this study pointed out, South Africa is in 12th position among other countries in the world in class 075, specialized metallurgical processes, compositions for use therein, consolidated metal powder compositions, and loose metal particulate mixtures.

In South–south collaboration Boshoff (2009a) underlined the nature of collaboration of Southern African Development Community (SADC) countries including South Africa. His study also provides evidence that South Africans tend to collaborate more with the global core (North) than the global periphery (South). Abramo et al. (2009) in their examination of the relation between productivity and collaboration noted the strength of correlation between these two. Drawing on the variation across different areas Abramo et al. (2009) located a strong relationship between productivity and collaboration in industrial and information engineering. Revisiting the data from the 1996 South African National Research and Technology Audit on the South African research system, Mouton (2000) documented the distinctive features of the inter-disciplinary, inter-institutional and inter-sectoral collaboration in the country. To view the South African academic system as an isolationist system, as Mouton (2000) concludes, is unsubstantiated given South Africa’s record in domestic and international collaboration. Other related studies include Pouris (2003), Bunt (1977), and Blankley and Moses (2009).

Against the backdrop of these studies the question that emerges is whether South African engineering in any manner reflects the production of scientific publications? One is at a loss to find plausible answers to this question as studies that chronicle the specific contribution of the discipline of engineering to the production of scientific knowledge is yet to appear in the literature. This study therefore comes out of this need by making a precursory effort to analyze the production patterns engineering research is taking in South Africa, spanning the last 30 years since 1975. Without a study of this kind it is difficult to comprehend the current standing of engineering research in South Africa and what measures should be taken to further its progress, growth, and the directions it requires in the present uncertain times. This paper seeks to make a contribution towards this understanding.

In the following part, the source of data and methods of analysis are described, which is followed by results. In the results section, the analyses are organized in subsections of: (i) publications in engineering that outline their distinctive features in comparison to the total publications of South Africans in all subjects; (ii) citations received by the engineering publications during the period of analysis and against those of all South African publications; (iii) affiliation details of the authors with differentiation in sectors such as university, research institute, industry, and government; (iv) publication outlets of journals that South African authors preferred to publish their pieces of research; and (v) collaborative dimensions of the authors that include the type of collaboration and their sectoral affiliation. These findings are discussed in detail in the discussion part.

Source, data, and methods

Data for this analysis were drawn in several stages from the ISI Web of Knowledge, the Science Citation Index Expanded (1945-present). Firstly, an appropriate period of data was chosen. During 1945–1965, there were no papers of the South African scholars in the database, for reasons of either not publishing or not appearing in the SCI journals yet (Sooryamoorthy 2009a). Until 1971 there were only a few publication records in this database and since then the number of publications has began to increase. Therefore the year 1975 presents itself as the right starting point, followed by another sample year for every five subsequent years. Thus, the data for a three-decade period with representative years of 1975, 1980, 1985, 1990, 1995, 2000 and 2005 were extracted. However, it should be noted that during this 30-year period the ISI indexing of the journals has changed. Some journals were dropped from the database while new ones entered the list (see the appendix as online supplement).

In the second stage, the type of publications was decided upon. Publications grouped as ‘articles’ and ‘reviews’ for the selected years were thus retrieved.

The data in the ISI is stored under several categories including articles, notes and reviews. In order to qualify for the current analysis these articles and reviews should have a minimum of one South African author in the address box of the publication record. There were 18,466 such publications by South African authors and their partners in all the seven selected years, of which 1,320 belonged to the branch of engineering.

In the next phase of data processing the subject of the publication was to be determined. This being a study of engineering publications by South Africans, all the engineering publications listed in the subject category of the ISI records were gathered. The subject category of engineering included publications in engineering, computer sciences, crystallography, instrumentation, materials science, mineralogy, remote sensing, transportation, and medical technology. The ISI which currently covers 12,000 international and regional journals in all areas of the natural sciences, social sciences, and arts and humanities. The principle behind the selection of journals by the ISI is Bradford’s Law that that the core literature for any given scientific discipline is composed of fewer than 1,000 journals and of these there are relatively a few with a very strong relevance to the subject. The articles in the ISI database are classified according to the total number of citations and the impact factor, which are then ranked according to the criteria of immediacy index and cited half-life. The list of the ISI journals that come under these subject categories is in the online appendix.

Next, all the basic but relevant details of the publication records from the database were collected and entered manually into a statistical software programme for further analysis. This was necessary as the features of the ISI database do not allow for advanced statistical analyses. In the last stage, many more new variables were created to form a new dataset from the basic data pertaining to authors, journals, year of publication, affiliation, subject, and citations. The citation calculations of the ISI take place on two levels. Firstly, it looks at the citations to the journal, as expressed by Impact Factor and/or total citations received. Secondly, it examines the citation record of the contributing authors, which is important in the evaluation of new journals about which a citation history does not exist. However ISI takes into account the variations between journals as it believes that the quantitative citation data to measure impact is useful only in the context of journals in the same general discipline. In smaller fields of science like crystallography not many articles or citations are generated as in other larger fields such as biotechnology or genetics. Also, in some areas it takes a longer time to attract a considerable number of citations, which is not the case with papers in life sciences.

Results

Publications in engineering

South African publications in engineering constituted 7% of the total publications (1,320 of 18,466) during the seven sampled years beginning from 1975 to 2005. Within engineering a steady movement in the number of publications is evident. From 52 in 1975 to 340 publications in 2005, the proliferation in number was almost seven-fold for the entire period of analysis. The first big leap within the years happened in 1980 by nearly a three-fold increase (from 52 to 141), and then in 1995 with about five-fold increase of the 1975 figure (52 and 251).

The basic features of engineering publications were derived from the variables, namely the number of authors per publication, the fractional count of papers, the number of countries involved in the production of publications, the count of citations received, sectoral affiliation of authors, and the type of journals in which these publications appeared.

The average number of authors taking part in the final production of research papers in engineering increased by 49% from 1.87 in 1975 to 2.78 in 2005 (row 1, Table 1). The mean value for the whole period was 2.35. Year on year, the increase was most conspicuous in at least the 2 years of 1990 (15%) and 2000 (21%) over their immediate previous years. Only in 1980 the mean number declined by eight percentage points of the 1975 level. The mean number of authors per publication began to exceed the value of 2 after 1990 and in 2005 it came closer to 2.8.
Table 1

Publications in engineering in South Africa, 1975–2005

 

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

 

Year

ANOVA

1975 (N = 52)

1980 (N = 141)

1985 (N = 152)

1990 (N = 164)

1995 (N = 251)

2000 (N = 220)

2005 (N = 340)

All (N = 1320)

F

Sig

Mean

SD

Mean

SD

Mean

SD

Mean

SD

Mean

SD

Mean

SD

Mean

SD

Mean

SD

Publications

                  

1. No. of authors/publication

1.87

0.86

1.72

0.99

1.89

1.07

2.16

1.16

2.30

1.11

2.69

2.54

2.78

1.72

2.35

1.63

12.914

0.000

2. Fractional count

0.67

0.29

0.74

0.30

0.67

0.29

0.59

0.29

0.55

0.27

0.51

0.26

0.48

0.26

0.57

0.29

21.674

0.000

3. No. of countries/publication

1.10

0.30

1.10

0.34

1.11

0.31

1.09

0.29

1.16

0.44

1.36

0.88

1.36

0.85

1.22

0.63

8.351

0.000

4. No. of times cited/publication

11.50

17.25

4.56

11.38

5.46

8.94

5.23

8.39

5.16

8.79

3.36

4.69

0.86

1.95

3.98

8.26

19.364

0.000

Sector of affiliation of authors

5. University

0.52

0.94

0.55

0.72

0.68

0.84

0.87

0.71

1.07

0.94

1.70

1.03

1.92

1.09

1.25

1.08

69.764

0.000

6. Government

0.23

0.70

0.16

0.45

0.08

0.27

0.10

0.32

0.06

0.24

0.03

0.20

0.01

0.09

0.70

0.29

8.418

0.000

7. Industry

0.54

0.98

0.32

0.57

0.24

0.52

0.16

0.37

0.24

0.51

0.25

0.70

0.19

0.54

0.24

0.57

3.720

0.000

8. Research institute

0.48

0.80

0.26

0.51

0.34

0.54

0.16

0.45

0.15

0.48

0.22

0.67

0.18

0.59

0.22

0.57

4.128

0.000

Publication outlets

N

%

N

%

N

%

N

%

N

%

N

%

N

%

N

%

130.76

0.000

9. South African journal

21

40.40

74

52.50

32

21.10

32

19.50

26

10.40

36

16.40

46

13.50

267

20.20

130.76

0.000

10. Non-South African journal

31

59.60

67

47.50

120

78.90

132

80.50

225

89.60

184

83.60

294

86.50

1053

79.80

 

South African engineers attract people from other countries to become their research partners. During 1975–2005 they worked with their counterparts from 32 other countries. There has been a maximum collaboration of 12 countries in the production of a single engineering paper. The count of countries per publication reveals that there has been an involvement of 1.22 countries per publication for engineering (row 3, col. 15, Table 1). The rise in the mean number of countries per publication from 1.1 in 1975 to 1.36, augmenting about one-fifth of the original figure of 1975 (row 3, Table 1) presents the expanding scenario of participation in international collaboration in South African engineering. In 2000 the mean number of countries moved to 1.36 (24% over the 1975 figure), the highest level ever since 1975. International participation will be analyzed in detail later.

The declining fractional count (calculated by dividing the number of papers by the number of authors) of publications—0.67 in 1975 to 0.48 in 2005—further confirms the disposition of engineering scholars for collaboration that leads to the production of scientific papers (row 2, Table 1). The fractional count in 2005 is compressed by about two-thirds of the mean value for 1975 (0.68 and 0.48) again supporting this tendency of engineers to collaborate more in research.

Citations received

The citation count of the publications of South African engineers is of a regressive order. From 11.5 citations per publication in 1975 to 3.36 in 2000 (2005 is left out as it is too early to consider the citations figures) the percentage of regression was close to 70% (row 4, Table 1). All the 1,320 publications in engineering jointly collected 5,254 citations during the entire period of analysis. However, amongst the engineering publications, 529 did not attract any citation, i.e. zero citation. Put it in percentile terms, these 529 publications with zero citations constituted 40% of the total publications in engineering. At the other end of this scale, one publication was cited 87 times by other researchers.

Sectoral affiliation of all authors

The scientific system of South Africa includes state-owned corporations, science councils, traditional universities, universities of technology, comprehensive universities, and domain-specific research organizations. There are 23 universities in the country (Inglesi and Pouris 2008). The major national research institutes in the country are Bureau of Economic Research (1944), Council for Scientific and Industrial Research (CSIR 1945), Human Sciences Research Council (HSRC 1968), Medical Research Council (MRC 1969), National Mineral Research Organisation (Mintek 1981), Agricultural Research Council (ARC 1992), Council for Geoscience (CGS 1993), National Research Foundation (NRF 1999), National Health Laboratory Services (NHLS 1999), and the National Energy Research Institute (NERI 2004).

Authors (both South Africans and their international partners) are affiliated to four sectors: university, research institute, industry, and government. The university sector covers universities and technikons/universities of technology; research institutes are usually regional and national centres where research is a major activity; the government sector includes provincial and national government departments; and industry means companies, firms, and business enterprises. Computing the sectors of the first five authors resulted in the actual number of authors coming from each of these sectors, which are presented in Table 1 (rows 5–8). In the aggregate period of analysis, the university sector comes top with the maximum number of authors, followed by government sector, industry, and lastly the research institute sector. The average number of authors belonging to the university sector (calculated by adding the number of authors divided by number of publications belonging to the sector) for the full period of analysis is 1.25 in contrast to 0.7 for the government sector, and 0.24 for industry. The research institute sector is on the bottom rung with a value of 0.22.

The contribution of the university sector in engineering research in the country has been growing in a linear fashion. From 0.52 authors in 1975 it has appreciated substantially to 1.92 in 2005—an increase of 269%. The biggest change ever was first noted in 2000 with a 227% increase over the 1975 values.

While the role of the university sector has been expanding steadily in the production of scientific publications in engineering, it was contracting for the research institute sector. By 2005, research institutes lost nearly one-third of their original share of 1975 (0.48–0.18). As far as engineering research is concerned, the position of the industrial sector was significant (row 7, Table 1). In the government sector, the production of scientific research in engineering has clearly diminished, at least during the period of analysis (from 0.23 in 1975 to 0.01 in 2005) (row 6, Table 1).

Looking from another plane on the affiliating centres of the authors, it is clear that in 38% of the engineering publications there was only one author who represented the university sector, two authors in 21% of publications, three authors in 10%, four authors in 2%, and 0.5% had five authors from this sector. In the sectors of research institute and industry, one author formed 14 and 15% of the engineering publications, 1.4 and 2.7% had two authors, and 0.8 and 1% reported three authors respectively. About 5% of the publications returned with one author, and less than 1% has two or more authors designating the government in engineering research.

Publication outlets

South African engineering researchers publish their research findings in both South African originated and foreign-originated journals accredited and indexed in the ISI but in differing proportions. Over the years—between 1975 and 2005—the preference of South African engineering scholars has swung in favour of foreign-originated journals (60% in 1975 to 80% in 2005, row 10, Table 1). This association is obvious in the Chi-square test (publication outlets, col. 17, Table 1). Inversely, publications in South African-originated journals plummeted (40% in 1975 to 20% in 2005, row 9, Table 1) despite a rise in the absolute number of papers.

Authorships and collaborative patterns

Nearly three quarters of the engineering publications are co-authored (row 2, col. 16, Table 2). Put differently, three out of every four papers are a collective product of collaboration. Co-authored publications of South African engineers have been mounting up in number during the period of analysis. The proportion of such co-authored publications to the total publications in engineering is scaled from 58% in 1975 to 82% in 2005. If the 1975 figures are taken as the base (1975 = 100), the increase in 2005 was nine-fold. Between the years, the change was 213% in 1980 and 293 in 1985, 380 in 1990, 640 in 1995, and 600 in 2000, with a notable increase of 22% in 1995 over the immediate reference period of 1990.
Table 2

Authorships and collaborative patterns in engineering in South Africa, 1975–2005

 

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

 

Year

1975

1980

1985

1990

1995

2000

2005

All

N

%

N

%

N

%

N

%

N

%

N

%

N

%

N

%

Publication records

Authorship

 1. Sole authored papers

22

42.3

77

54.6

64

42.1

50

30.5

59

23.5

40

18.2

60

17.6

372

28.2

 2. Coauthored papers

30

57.7

64

45.4

88

57.9

114

69.5

192

76.5

180

81.8

280

82.4

948

71.8

 3. All South African authors***,#

25

83.3

53

82.8

72

81.8

98

86.7

159

82.8

121

67.2

185

66.1

713

75.3

 4. SA authors within the same organization***,a,#,&

22

73.8

11

34.4

11

26.2

11

30.6

21

24.4

153

85.0

223

79.6

452

65.9

Collaboration types

 5. Any collaboration

30

57.7

64

45.4

88

57.9

114

69.5

192

76.5

180

81.8

280

82.4

948

71.8

 6. Domestic collaboration#

25

83.3

53

82.8

72

81.8

98

86.0

160

83.3

125

69.4

195

69.6

728

76.8

 7. Internal institutional collaboration#,***

22

73.3

44

68.8

57

64.8

88

77.2

126

65.6

98

54.4

146

52.1

581

61.3

 8. External institutional collaboration#,***

3

10.0

9

14.1

15

17.0

11

9.6

36

18.8

27

15.0

56

20.0

157

16.6

 9. International collaboration#

5

16.7

11

17.2

16

18.2

15

13.2

33

17.2

59

32.8

95

33.9

234

24.7

 10. Multi-country international collaboration#,$

0

0.0

2

18.2

0

0.0

0

0.0

4

12.1

8

13.6

13

13.7

27

11.5

 

Mean

SD

Mean

SD

Mean

SD

Mean

SD

Mean

SD

Mean

SD

Mean

SD

Mean

SD

 11. Degree of collaboration (mean)#

0.58

0.50

0.47

0.53

0.58

0.50

0.70

0.47

0.79

0.48

0.87

0.48

0.91

0.53

0.76

0.52

Notes: ***, **, * significant at the .01, .05, .1 levels respectively. a Chi-square test; # If if refers to single authored paper they are not applicable under this classification

#Percentage out of ‘any collaboration’; $ percentage out of ‘international collaboration’; & excludes international collaboration

Among the co-authored publications in engineering 75% (row 3, col. 16, Table 2) of the authors are South Africans (for the whole period of analysis), working in different institutions in South Africa. In 1975 this percentage was 83 (of the co-authored publications). The following years witnessed an inconsistent pattern of ups and downs of decreasing, increasing and decreasing again before reaching the bottom level of 66% in 2005. By 2005, the percentage has contracted by 79% of the 1975 figure. In 66% of the co-publications in engineering, South African authors belonged to the same department or institution. This percentage is on an upward trail, from 74% in 1975 to 85% in 2000, and then declining to 80% in 2005 (row 4, Table 2). This is termed as internal-institutional collaboration which is part of domestic collaboration.

The nature and type of collaboration can be deduced from the variable of co-authored publications. Firstly, whether any type of collaboration that resulted in the production of papers is identified. Secondly, what kind of collaboration exists among engineering researchers is considered. The type of collaboration involved in the production of papers can occur in several forms. If it is within the same department or institution within South Africa, or with other institutions in South Africa it can be internal-institutional or external-institutional. Both of them are domestic collaboration, as they take place internally within the country. If collaboration involves participation of at least one foreign country it is international collaboration. International collaboration may be single-foreign country or multiple-foreign country depending on the number of foreign countries involved in the production of papers. An index of the degree of collaboration was also calculated by adding up the measures of internal-institutional, external-institutional, international, and multi-country collaborations.

Seventy-seven percent of the collaborated papers in engineering were generated in domestic collaboration, i.e. authors belonging to the institutions or departments within South Africa (row 6, col. 16, Table 2). Domestic collaboration in recent years, however, showed the symptoms of degeneration, more significantly from 1995 and after: 83% in 1975 to 70% in 2005. South Africa lost about 13% of its domestic collaboration that existed in engineering in 1975. If domestic collaboration is split into internal-institutional and external-institutional the divide for engineering is quite sharp: 56 and 16% respectively. While internal domestic collaboration is prone to decline, external collaboration turned the opposite way (rows 7 and 8, Table 2). For the internal-institutional type the magnitude of depreciation was more than a quarter (629%) of the 1975 figures. Chi-square tests also confirmed this association between institutional collaboration and year of publication. External-institutional collaboration at the same time appreciated by 100%. This clearly shows the changing nature of domestic collaboration as evident in the growth of external-institutional collaboration vis-à-vis internal-institutional collaboration.

Collaborated publications in engineering that brought researchers from other countries together—international collaboration—do not seem to be as great as domestically collaborated papers. These internationally collaborated papers constituted only 17% of the total co-authored papers in 1975 but by 2005 it rose to 34% (row 9, Table 2). Since 1975 the growth in international collaboration has been phenomenal in engineering, with more than 100% over the percentage that reported in 1975. The increase was most conspicuous in 2000 (33%) and 2005 (34%).

In some cases international association might engage more than one foreign country in the collective research activity. Such multi-country collaboration has ensued remarkably both in absolute number and percentage of those internationally collaborated papers, particularly in 1995 and thereafter (row 10, Table 2). About 12% of international publications in engineering had solicited the collaboration of more than one country. Over the years, from 1975 to 2005, the percentage moved from 2 to 14. In particular, the growth in multi-country collaboration was significant in 1995 (12%), 2000 (14%), and in 2005 (14%).

An index of the degree of collaboration was also developed for this analysis by combining the variables of internal-institutional, external-institutional, international, and multi-country collaborations. On a continuum of 1–4, the maximum value of this index could be 4, if a paper is the outcome of all four types of collaborations, namely, internal-institutional, external-institutional, international, and multi-country. Of the co-publications 94% came up with the value of 1.00 and 5% with 2.00 The mean value of 1.00 (row 11, col. 15, Table 2) means collaboration is uniformly present in all publications in more than one count of internal-institutional, external-institutional, international, or multi-country collaborations. The mean values have varied positively from 1975 as well. This substantiates the previous finding that collaboration has been growing among South African engineers. In particular, the change is seen towards external collaboration outside the institutions and the country, as pointed out in the case of external-institutional and international collaborations.

Sectoral affiliation of South African authors and their international partners

Scientific research that leads to the production of publications in engineering takes place at universities, research institutes, industry, and in several government departments. These sectors display differing levels of production of scientific publications in the country. Counting the sectoral affiliation of the South African authors and their international partners provided details of the centres where research is concentrated and publications are generated. Table 3 presents the sectoral attributes of the first five South African authors and their first five respective partners. The sectors of the first five authors have already been demonstrated in Table 1.
Table 3

Sectoral affiliation of South African authors and their international partners, 1975–2005

 

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

 

Year

ANOVA

1975 (N = 52)

1980 (N = 141)

1985 (N = 152)

1990 (N = 164)

1995 (N = 251)

2000 (N = 220)

2005 (N = 340)

All (N = 1320)

F

Sig

Mean

SD

Mean

SD

Mean

SD

Mean

SD

Mean

SD

Mean

SD

Mean

SD

Mean

SD

Publications

I. Sector of affiliation of SA authors

 1. University

0.44

0.80

0.48

0.62

0.61

0.76

0.79

0.65

0.93

0.82

1.44

1.01

1.59

1.08

1.06

0.98

50.885

0.000

 2. Industry

0.54

0.98

0.30

0.54

0.24

0.50

0.14

0.35

0.24

0.51

0.20

0.57

0.16

0.49

0.22

0.53

5.087

0.000

 3. Research institute

0.48

0.80

0.24

0.49

0.32

0.54

0.13

0.35

0.12

0.46

0.14

0.46

0.13

0.48

0.18

0.49

7.646

0.000

 4. Government

0.23

0.70

0.15

0.45

0.06

0.24

0.09

0.29

0.06

0.24

0.03

0.16

0.01

0.08

0.06

0.28

9.069

0.000

II. Sector of affiliation of intl partners

 5. University

0.06

0.24

0.07

0.31

0.07

0.26

0.07

0.27

0.13

0.40

0.25

0.59

0.34

0.67

0.18

0.50

11.954

0.000

 6. Research institute

0.02

0.14

0.01

0.12

0.01

0.08

0.04

0.29

0.03

0.18

0.09

0.48

0.06

0.35

0.04

0.30

1.603

0.143

 7. Industry

0.02

0.14

0.02

0.15

0.01

0.08

0.02

0.16

0.00

0.06

0.05

0.31

0.03

0.19

0.02

0.18

1.224

0.291

 8. Government

0.00

0.00

0.01

0.08

0.02

0.14

0.01

0.08

0.00

0.00

0.00

0.07

0.00

0.05

0.01

0.07

1.352

0.231

III. Regional location of intl partners

 12. North America

0.06

0.24

0.04

0.28

0.06

0.24

0.05

0.25

0.06

0.29

0.15

0.50

0.14

0.53

0.09

0.39

3.224

0.004

 13. Europe

0.02

0.14

0.04

0.24

0.02

0.17

0.04

0.34

0.06

0.29

0.12

0.38

0.15

0.48

0.08

0.35

4.655

0.000

 14. Australasia

0.00

0.00

0.00

0.00

0.00

0.00

0.01

0.11

0.01

0.09

0.06

0.36

0.02

0.18

0.02

0.18

2.624

0.016

 15. Asia

0.02

0.14

0.01

0.08

0.00

0.00

0.00

0.00

0.02

0.12

0.02

0.18

0.04

0.22

0.02

0.15

1.975

0.066

 16. Latin America

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.05

0.01

0.03

0.479

0.824

 17. Middle East

0.00

0.00

0.01

0.11

0.00

0.00

0.01

0.07

0.00

0.06

0.02

0.13

0.01

0.08

0.01

0.09

1.045

0.394

 18. East Europe

0.00

0.00

0.00

0.00

0.01

0.08

0.00

0.00

0.02

0.15

0.00

0.07

0.00

0.00

0.00

0.07

1.314

0.247

 19. Africa

0.00

0.00

0.01

0.11

0.00

0.00

0.01

0.08

0.00

0.00

0.00

0.00

0.06

0.11

0.00

0.07

0.859

0.525

As seen in the mean values of the sectors of South African authors (panel 1, Table 3), universities produce the highest number of researchers in engineering, followed by research institutes, industry, and government in descending order. The university sector, over the years, has expanded from 0.44 to 1.59 recording an increase of 260% above the original mean number of authors for 1975. Although second in the order after the university, the research sector does not seem to be in any way closer to universities in its contribution of authors to engineering research in the country. As against the university sector with a mean value of 1.06 authors for the aggregate period, it is only 0.18 authors for research institutes. It has (for the research institutes) declined from 0.48 in 1975 to 0.13 in 2005, losing out about 73% of its original portion of 1975. Industry participated a little better than the research institutes in engineering research, but it too slowed down from 0.54 in 1975 to 0.16 authors in 2005. The same applies to the government sector as well.

A large majority of the international partners of South African engineers originate from the university sector than from any other listed sectors. As seen in the panel II of Table 3, the university sector scored a net average value of 0.18 authors, while the next major sector of the research institutes ended up with 0.04 authors. Industry has 0.02 authors while the government sector with its 0.01 authors becomes the poorest amongst all sectors in this regard.

Most of the international research partners of South African researchers arrive from North America, as the regional location of the collaborators in panel III of Table 3 illustrates. After North America, in descending order, are Europe, Australasia, Asia, Latin America, the Middle East, Eastern Europe and Africa. Partners from Africa and Eastern Europe are few and far between. Close on the heels of the North American engineers are European engineers who sought collaboration with South Africans. Next come their Australasian and Asian partners, but not as conspicuous as Europeans or North Americans in their numbers. The mean values of the partnering authors from North America, Europe, Australasia and Asia, for the aggregate period, were 0.09, 0.08, 0.02 and 0.02 respectively (col. 15, Table 3); and for the remaining continents the figures remain low.

The country-wise list of the first five international partners of South African engineers (Table 4) shows 32 countries that had collaborative links with South African engineers. One-third of the total collaboration occurred with the US. England produced 12%, Germany 9, Australia 7, Canada 6, the Netherlands 5, and the rest of the countries in descending number. European participation is ensured by researchers from England, Germany, the Netherlands, Switzerland, Italy, Scotland, France, Belgium, Denmark, Ireland, Sweden, Spain, and The Wales. The link between South Africa and other African countries is very weak having one of the lowest collaborations in engineering. The African countries with which South Africa has collaboration in engineering include Zimbabwe, Malawi, Namibia and Botswana. Each of these countries contributed just 1% or less to the total collaboration with South African engineers.
Table 4

Partnering countries of South African engineers, 1975–2005

No.

Countries

Partner 1

Partner 2

Partner 3

Partner 4

Partner 5

Total

No.

%

No.

%

No.

%

No.

%

No.

%

No.

%

1

USA

69

31.65

20

36.36

8

40.00

4

50.00

3

50.00

104

33.88

2

England

30

13.76

5

9.09

1

5.00

 

0.00

 

0.00

36

11.73

3

Germany

19

8.72

7

12.73

2

10.00

1

12.50

 

0.00

29

9.45

4

Australia

11

5.05

5

9.09

2

10.00

1

12.50

1

16.67

20

6.51

5

Canada

17

7.80

1

1.82

 

0.00

 

0.00

1

16.67

19

6.19

6

The Netherlands

10

4.59

4

7.27

 

0.00

 

0.00

 

0.00

14

4.56

7

Israel

10

4.59

 

0.00

 

0.00

 

0.00

 

0.00

10

3.26

8

China

6

2.75

1

1.82

 

0.00

1

12.50

 

0.00

8

2.61

9

Italy

5

2.29

1

1.82

1

5.00

 

0.00

 

0.00

7

2.28

10

Switzerland

4

1.83

 

0.00

2

10.00

 

0.00

 

0.00

6

1.95

11

Japan

1

0.46

3

5.45

1

5.00

1

12.50

 

0.00

6

1.95

12

Scotland

3

1.38

2

3.64

 

0.00

 

0.00

 

0.00

5

1.63

13

India

3

1.38

 

0.00

2

10.00

 

0.00

 

0.00

5

1.63

14

France

4

1.83

 

0.00

 

0.00

 

0.00

 

0.00

4

1.30

15

Belgium

4

1.83

 

0.00

 

0.00

 

0.00

 

0.00

4

1.30

16

Russia

3

1.38

1

1.82

 

0.00

 

0.00

 

0.00

4

1.30

17

Denmark

3

1.38

 

0.00

1

5.00

 

0.00

 

0.00

4

1.30

18

South Korea

3

1.38

1

1.82

 

0.00

 

0.00

 

0.00

4

1.30

19

Ireland

1

0.46

2

3.64

 

0.00

 

0.00

 

0.00

3

0.98

20

New Zealand

1

0.46

 

0.00

 

0.00

 

0.00

1

16.67

2

0.65

21

Zimbabwe

1

0.46

1

1.82

 

0.00

 

0.00

 

0.00

2

0.65

22

Sweden

1

0.46

 

0.00

 

0.00

 

0.00

 

0.00

1

0.33

23

Spain

1

0.46

 

0.00

 

0.00

 

0.00

 

0.00

1

0.33

24

The Wales

1

0.46

 

0.00

 

0.00

 

0.00

 

0.00

1

0.33

25

Brazil

1

0.46

 

0.00

 

0.00

 

0.00

 

0.00

1

0.33

26

India

1

0.46

 

0.00

 

0.00

 

0.00

 

0.00

1

0.33

27

Malawi

1

0.46

 

0.00

 

0.00

 

0.00

 

0.00

1

0.33

28

Namibia

1

0.46

 

0.00

 

0.00

 

0.00

 

0.00

1

0.33

29

Yugoslavia

1

0.46

 

0.00

 

0.00

 

0.00

 

0.00

1

0.33

30

Czechoslovakia

1

0.46

 

0.00

 

0.00

 

0.00

 

0.00

1

0.33

31

Turkey

1

0.46

 

0.00

 

0.00

 

0.00

 

0.00

1

0.33

32

Botswana

  

1

1.8

 

0

 

0

 

0

1

0.32

 

Total

218

100

55

100.0

20

100

8

100

6

100

307

100

Discussion

This study examined the publication patterns of the engineering researchers in South Africa, as seen in the scientific publications stored in the ISI database for a 30-year period, spanning from 1975 to 2005. Engineering publications during the period of analysis grew considerably in number. Both the average count of authors per publication and the fractional count of South African engineers present the scenario that partnership in engineering research is still expanding. Compared to the level of 1975, the number of authors per publication rose prominently in 2005. On the other hand, the production of research articles in engineering in Africa for the period 1980–2004 was 7 as against 13% for the world (Tijssen 2007). India’s output as against the world output in several branches of engineering (civil, industrial and mechanical) between 1990 and 1994 has increased (Rao and Suma 1999). South African engineers usher in the scientific fraternity from overseas as seen in the number of countries engaged in the production of scientific publications. This propensity to associate in research has been building up particularly from 1990 and later. The production of collaborated papers also began to increase, even at a higher level than the production of all engineering papers. Engineering researchers in South Africa during this period of analysis have partnered with other researchers in 32 countries. This can be compared with a similar study of South African scholars (Sooryamoorthy 2010) which reported 70 countries in all subjects and 54 countries in medicine. India’s collaboration with China during 1994–1999 produced 67% collaborated publications in engineering and technology (Gupta and Dhawan 2003).

There is of course a cause of concern in regard to the citations received by the engineering publications of South Africans. The citation count of the publications for the chosen years was drawn directly from the ISI database irrespective of the journals in which the publications appeared. The year 2005 was neglected, as it is too early to gain the optimum number of citations against the other chosen years. The declining citation count of engineering publications of South Africans and their partners reflects the diminishing visibility and importance of the publications and the research that gave rise to these publications. This is a matter of interest when 40% of the publications remain in the knowledge domain without attracting even a single citation. Citation studies (Schmoch and Schubert 2008; Sooryamoorthy 2009b, for instance) provide proof that the count of citation tends to improve when the publication is the product of collaboration, particularly international. Also, citations depend on the impact factor of the journals in which the publications appear. In engineering, collaborated papers formed about three-fourths of its publications. Of this international collaboration is reported in only 17% of the collaborated papers. If we go by the evidence of previous studies that indicated the relationship between increasing citations and international collaboration, citation of engineering publication could rise in correspondence to an expansion in international collaboration.

A few of the South African journals that are listed in the ISI database are in engineering and allied branches. The impact factors of these journals in 2005, however, were not very high. For example, the Journal of the South African Institute of Mining and Metallurgy’s impact factor was 0.007, while the South African Journal of Geology had 1.086. A related issue here is in which journals South African engineers publish their research outcomes. It is clear that they prefer to see their papers appearing in journals that are originated elsewhere than in South Africa. How do we explain this? As noted above the local South African originated journals are also rated high in academic publishing, and are indexed by the ISI that follows international standards for listing of journals. Despite this, if South African engineers opt to publish in non-local journals, there can be other reasons. Does the changing preference of South African engineers to publish in foreign-originated journals than South African ones have something to do with collaboration? To find this out a Chi-square test was run between the variables of collaboration and journal’s origin. The test results (χ2 = 13.791, df = 2, p < 0.001) show that coauthored papers and non-South African journals are associated. That is, 75% of the non-South African journal articles are coauthored, while only 65% South African journal articles are collaborated ones. A previous study (Mouton 2000) that investigated the disciplinary variation in the publication preferences of South African scholars disclosed similar findings. According to this study, engineering, medical sciences, and natural sciences record the higher counts of publications in overseas journals. This would suggest that collaborated papers tend to appear in non-South African journals rather than in South African journals. This can be explored further for the association between international collaboration and the origin of journals. Among the total 234 internationally collaborated publications, 206 (88%) are published in non-South African journals and the rest of 28 (12%) in South African journals. Clearly, collaboration influences the choice of the journals, and this choice is definitely in favour of non-local journal if there was any international collaboration in its production.

Collaboration has come to stay as an integral part of engineering research in South Africa. The lowering fractional count indicates the growing level of collaborative efforts in engineering research. Seen in the measures of collaboration index and other related variables is that collaboration is the preferred mode of research rather than individualized research in engineering. While collaboration as a whole recorded an ascending trend in engineering, domestic collaboration has witnessed a decline. Within domestic collaboration a clear border between the internal-institutional and external-institutional partnership is visible, both in the percentage of the publications, and in the trends within the years of analysis. Although the internal one formed the largest share of the domestic collaboration, its share is shrinking gradually. On the other hand, external-institutional is small in size, but showing signs of improvement. Marked in this pattern is the orientation of engineering researchers looking outwards more than ever and beyond their institutional confines for prospective collaborative ventures, notably since 1995.

International collaboration, as opposed to domestic collaboration, is not very significant but growing strongly over the years. Multi-country participation is taking roots in engineering. The preferred partners of South African engineers are North Americans, and this partnership is now fortified. This cannot simply be explained away as the perpetuation of the past ties South Africa has had with these countries. Not surprising though, as Boshoff (2009b) noted in his study of Central Africa, is that 80% of the collaborated papers in Central Africa are in partnership with those outside the region of Africa. Dastidar and Ramachandran (2005) reported that the USA and the UK are the top collaborators in international engineering research. What is evident in these international alliances of South African engineers is the minimal partnerships with their neighbours. The predominant African collaborators of South Africa are Zimbabwe, Malawi, Namibia and Botswana. This finding endorses again that of Boshoff (2009a) whose analysis referred to the collaboration of South Africa with other SADC countries. South African collaboration with Asian countries, the new giants in science, has also been very poor so far.

The analysis of the affiliation background of the authors including that of their partners revealed the centres where engineering publications are emanating from. It is unmistakably the university sector that produced the maximum number of publications, as depicted in the high mean value of authors belonging to universities. The university sector in the production of scientific publications has seen spectacular growth, particularly after 1985. This impressive record placed the university sector parallel to that of the other major sectors, but the research institutes in the country did not fare well. Rather surprisingly, research institutes lost their contribution. While the government sector performed better than the research institutes industry lagged behind.

The sectoral scenario becomes clearer when the affiliations of South African authors are taken into account separately. As for the sectors of South African authors, universities continue to be the major reservoir of engineers. The university sector has appreciated its contribution to scientific research. The other sectors such as research institutes, industry, and the government trail behind universities. Research institutes and industry with their 21 and 17% share respectively of the total share of the universities raise questions about their shrinking presence in the production of engineering knowledge in South Africa. Industry, as one would normally expect, has not adequately provided the leadership that engineering research required in the country. Despite the benefits that firms accrue from conducting and publishing research, industry in South Africa has not taken a serious, not to mention a central role in scientific research. This is when industries the world over are becoming important players in knowledge production (Okubo and Sjoberg 2000).

In South Africa three sectors invest in R&D. The R&D expenditure in the country in 2004–2005 showed that most of the research in engineering science is supported by the business sector, followed by science councils, and lastly the higher education sector. Compared to other fields such as basic sciences that include chemistry, physics, mathematics, and marine science, ICT, applied, medical, life and social sciences, the business sector funded 30% of its total R&D expenditure on engineering research (HSRC 2006 cited in Kahn 2007). The contribution of the science councils and the higher education sector to engineering research was 18 and 12% of their total funding respectively. If this is the case wherein the business sector—industry in other words—should have been in the forefront of scientific research, not the higher education sector namely the universities. Again, the expenditure of firms in South Africa on innovations, based on a survey of industrial firms, indicates that it leaves much to be desired. In 2000, South African business spent an average of 2.6% of their total sales, while it was 3.7% for the European Union (Rooks et al. 2005). As my analysis suggests, the government sector that includes its various departments is pulling out of its responsibility of conducting and publishing research in engineering science.

One would find a different pattern in other countries. Okubo and Sjoberg (2000) report that in Sweden the contribution of the private and public firms to the scientific knowledge—in terms of scientific publications—was fairly stable as their participation in collaborative scientific activities.

Conclusions

This study on the scientific publications of South African engineers as seen through their research publications indexed in the ISI database brings out some unique features of engineering research in the country. The output of engineering research is largely influenced by the opportunities for collaboration. This collaboration can be both national and international. Another point of concern that evolves from this analysis is the declining contribution of research institutes, industry, and the government, and the heavy reliance on the university sector for the growth of engineering in the country. A serious shortage of skilled and trained people in the country in the scientific domain (Philander 2009) also has its impact on engineering research. Though limited to industrial engineering and not to engineering in general some, like Lister and Donaldson (2003) note, education is the key for engineers to provide leadership, improvement, innovation, growth and sustainability in engineering. Perhaps this might improve the status of the engineering branch of science in the country, through research and innovation.

Supplementary material

11192_2010_288_MOESM1_ESM.doc (566 kb)
Supplementary material 1 (DOC 565 kb)

Copyright information

© Akadémiai Kiadó, Budapest, Hungary 2010