Abstract
The use of the content analysis method in library and information science research is discussed, and the method illustrated by a content analysis of Australian library research articles published over the ten-year period 1985-1994 in The Australian Library Journal and Australian Academic & Research Libraries. The analysis is made using categories devised by two Finnish researchers, Jiirvelin and Vakkari, for analysing articles in 37 core library and information science journals published in 1985.
Introduction
This paper presents a study of Australian library and information science research during the ten-year period 1985-1994, as represented in research articles published in the major journals The Australian Library Journal (AU) and Australian Academic & Research Libraries (AARL). The problem to be investigated is what interests Australian researchers; the topics investigated by them and the research strategies used. The methodology used is that employed by two Finnish researchers, Kalervo Jarvelin and Pertti Vakkari, for a content analysis of international research articles in 1985 in library and information science. Footnote1 Content analysis research method will first be examined, and then this particular study reported.
Content Analysis
Content analysis consists of extracting and evaluating in a systematic and generally quantitative manner the occurrences of the manifest and latent content of a body of textual or audiovisual material, in order to uncover its key symbols and themes and to compare them to one another. It can use different types of analysis-among others, classification, lexical and propositional. Footnote2 It is an excellent method of gathering data unobtrusively; the writer/producer does not know that their output will be analysed on publication.
Content analysis was used in research conducted in schools of journalism in the USA, and reflected their preoccupation with quantitative rather than qualitative research. Thus column inches of reporting on various topics was measured. Now qualitative analysis is employed also in analysing the data, and the method is used widely in the mass media. The research method is empirical.Footnote3 The first report of the use of content analysis quantitatively for printed materials was in Sweden in the 18th century, namely an analysis of a collection of 90 hymns. They were analysed for religious symbols and compared to symbols in established song collections. No differences were found, and thus the hymn collection was cleared of charges of carrying unorthodox ideas.Footnote4
Content analysis has been used in library and information science research to analyse the content of monographs, newspapers and journals, so as to investigate such areas as themes and bias. Fundamental to content analysis is the set of categories by which the content is analysed. Categories are set up according to rules, are designed for the particular problem to be investigated, and must be ‘exhaustive and mutually exclusive.’ Footnote5
To carry out a content analysis study the research questions and/or hypotheses must be defined; categories of analysis decided; materials selected to be examined; then analysis of the materials according to the categories decided; quantified results; compared with the hypotheses; and finally interpreted.Footnote6
An example of Australian library content analysis research is that employed by Joyce Kirk in ‘Portrayal of Aged Characters in Australian Award-Winning Children’s Novels 1946-1985.’ Footnote7 The materials examined were children’s novels which won Children’s Book Council of Australia awards from 1946 to 1985. The demographic, personal and behavioural characteristics of aged characters were identified using a 34 item checklist, giving a quantified portrayal score for the depiction of aged characters. Variance tests were then applied to the portrayal scores, showing the scores were significantly related to the position of importance of the aged characters in the novels. Kirk found that the aged characters did not adequately reflect the demographic characteristics of aged people in the Australian population; for personal characteristics the characters were in a restricted range of social settings, and their range of behaviours was also restricted. Kirk concluded that since children’s novels are a socialisation agent, the limited depiction of the aged in the novels is a cause for concern. There had been no improvement over time.
Assumptions of This Study
The area defined by library and information science has fluid boundaries, and it is therefore easy to use material published in library and information science journals to define the field. This has been done by previous researchers and is necessary to provide some comparative findings and also to provide cumulative research results. A major assumption is that journal articles are the main method of reporting research to the general library community. In Australia, reporting of library and information science research is done in formal research reports, in conference papers and in theses, and consultants’ reports may also include research results. One would hope that most of this research would also be published as a journal article to make it widely available. Another major assumption is that most Australian library and information science research will be published in the core Australian library journals. This assumption may be incorrect, with Australian researchers publishing in overseas journals, or in the specialist library journals such as Education for Library and Information Services: Australia. It is also to be noted that the school library area is neglected by taking only mainstream journals. Thus a more complete study could make a content analysis of all research journal articles published by Australian library professionals, in the major Australian and overseas journals. It could also include other formats such as monographs, research reports and theses.
Research Study
As it was decided to use a content analysis methodology already applied internationally, only research articles published in the two core Australian library journals, AU and AARL, were analysed. The aim was to make the research studies cumulative, and to attempt some comparison of results and analyse any differences found. There is a requirement that content analysis be replicable, that another researcher applying the same technique to the same data would come up with the same results. Thus coding of data must be reliable. However, there may be differences in how different people perceive messages in communications.Footnote8
To enable cross checking of categories allocated to the journal articles, the researcher and the research assistant, Mona Brown, independently decided whether an article was a research one or not, and to which category to allocate the subject of the research, as shown in Table . Both were familiar with the nature of the material. The definition of research adopted was that used in the original research study of 1985 by Jarvelin and Vakkari: ‘research is an inquiry, where the goal is to elicit, through a systematic method, some new facts, concepts or ideas. ‘ Footnote9 The topics used to classify the research articles were also those used by Jarvelin and Vakkari. Footnote10 There were a few differences of perception between the researcher and the research assistant in the allocation of articles to either of these categories. A ten-year period was used to see whether there had been trends over time in the number of research articles to non-research articles published in the two journals.
The data were summarised separately for AU and AARL for research and non-research articles, in Tables and , and cumulated in Table . The subject of research topics investigated is shown in Table . This enabled patterns and trends for the two journals to be discovered, and also a comparison with the data from other international studies.
Table 1 The Australian Library Journal Articles 1985-1994
Table 2 Australian Academic & Research Libraries Articles 1985-1994
Table 3 Total Research Articles AU and AARL 1985-1994
Table 4 The Distribution of Library and Information Science Topics in Research Articles in AU and AARL 1985-1994
Findings
As shown in Tables and the percentage of research articles in the two journals 1985-1994 was 22% for AU and 27% for AARL, giving a combined tot~l of 24%, as shown in Table . When a comparison was made of two five-year penods, there was only an increase of 5% in the percentage of research articles in the second five-year period, so we would need a future comparison to see if there is an increase over time.
An analysis of the Canadian Library Journal for the period 1981-91 examined many variables, including research-based approaches in articles. It was found that 24% of all articles used a research-based approach, the same as in Australia. Footnote11 A study of 1404 articles in core library journals published in 1980 also found that 24% were research based. Footnote12 However, the analysis by Jlirvelin and Vakkari of 833 articles published in 37 core international library and information science research journals in 1985 found that 54% of the articles were based on research. Footnote13
When the research articles were analysed by topic, using the classification of Jlirvelin and Vakkari, we find differing results for AU and AARL. Thus research on library and information service activities was the concern of 19% of articles in AU, but 58% of articles in AARL. The difference was not so marked for research on information seeking, the topic of 16% of articles in AU and 23% in AARL. A surprise was the high percentage of articles in AU concerned with library history, 23%, compared with 7.25% in AARL.
Comparison with other studies that used the same classification of library and information science topics reveals differences in results. The Jarvelin and Vakkari study of articles for 1985 showed that the largest percentage, 29%, concerned information storage and retrieval, with 27% concerned with library and information service activities. The Australian research articles were concentrated on library and information service activities (40%), followed by research information seeking (20%), and library history (14.2%). The high proportion on service activities perhaps reflects a concentration on applied research to bring about improvement. In 1985 library history research articles comprised only 3.8% of topics. However, high figures for library history research, 14%, were found in a content analysis of Danish research publications over a 25-year period, 1965-1989. Footnote14 The low percentage of articles on education in library and information science, 1.5%, compared to the international one of 4.7%, may be accounted for by publication in the specialist library education journal. The higher percentage of Australian articles on the topic of information seeking may reflect the later period of data collection. The small proportion of articles in these Australian journals concerned with information storage and retrieval is a matter of concern; this should surely be a key area of research in the information age.
Research Strategies Used
The research strategies used in the 126 research articles were analysed, also adopting the classification of Jlirvelin and Vakkari. Irena Ali was the research assistant who helped with this analysis. Empirical research strategies were most popular, used by 83% of Australian investigators. Jlirvelin and Vakkari found 56% of the international research articles in 1985 used empirical strategies. Of the empirical strategies the most popular in Australia was the survey, employed by 44% of the researches, compared to 23% for the international studies. In Australia the survey seems to be regarded as a standard strategy for all library and information science problems. The second most popular Australian research strategy was historical method, at 14%, higher than the 11% found by Jlirvelin and Vakkari. There was very little use in the Australian articles of other empirical methods such as qualitative, case study, or bibliometrics. System and software analysis and design were seldom used, not surprising because of the low number of Australian studies concerned with information storage and retrieval.
Conclusions
A fundamental characteristic of a profession is support for research and its dissemination, so one would expect that AU, the journal of the Australian Library and Information Association, and AARL, the journal of the University, College and Research Libraries Section of the association, would support publication of research articles. The findings of this analysis of the two journals show this support for research and its dissemination. The Australian figures for 1985-1994 of 24% of articles being research-based tallies with the figure of 24% for the Canadian Library Journal 1981-91, and with 24% for core library journals in 1980. However, articles from core international research journals for 1985 give a finding of 54% being research based, an emphasis we would expect in research-oriented journals.
When the aspect of the subject topics of research reported in the articles in AU and AARL are analysed some interesting results are revealed. The most popular Australian topic is research on library and information service activities (40%), followed by research on information seeking (20%) and library history (14.2%), followed by small percentages in other areas. The small percentage in the two Australian journals, 3.2%, concerned with information storage and retrieval warrants further investigation. As for method, one could hope that Australian researchers become more adventurous, and explore the use of research strategies beside the survey and historical method. Qualitative research, for example, has great potential, and is increasingly used in the social sciences area. It will be interesting to see if there are changes in the next ten years.
The research topic classification of Jlirvelin and Vakkari was built upon for the purpose of analysing the Spanish library and information science research literature as published in three Spanish library and information science journals in a paper by Virginia Cano and C Rey, ‘Ten Years of Spanish Library and Information Science Research’, delivered at the open meeting of the Section on Library Theory and Research at the IFLA Conference in Barcelona in August 1993. Cano and Rey found weaknesses with the topic classification used by Jlirvelin and Vakkari. They also analysed why the Spanish research topics and methodology were different from the international library and information science journal findings. This national focus of analysis is to be continued by the Section on Library Theory and Research of IFLA, with papers for the UK and Turkey planned to be presented at the IFLA Conference in Istanbul in August 1995, and one on China promised for the IFLA conference in Beijing in August 1996.
The versatility of the content analysis methodology for use in library and information science research has been demonstrated by this study. The discipline enforced by use of categories allows for cumulative research and attempts at analysing international comparisons and trends, keeping in mind the differences that may arise from allocation to categories, and the differing time periods used.
Additional information
Notes on contributors
Maxine K Rochester
Associate Professor in Library and Information Science, Charles Sturt University - Riverina
Notes
1. K Jarvelin and P Vakkari 'Content Analysis of Research Articles in Library and Information Science' Library and Information Science Research vol12 1990 pp395-421
2. P Bernhard 'AIa Recherche des Methodes de Recherche Utilisees en Sciences de )'Information' Canadian Journal of Information and Librmy Science vol 18 no 3 1993 pll
3. K Krippendorff Content Analysis: An Introduction to its Methodology Beverly Hills Sage Publication 1980 pp7-12
4. ibid p13
5. F N Kerlinger Foundations of Behavioural Research: Educational and Psychological Inquiry New York Holt Rinehart and Winston 1967 p696, cited inCH Busha and S P Harter Research Methods in Librarianship: Techniques and Interpretation New York Academic Press 1980 p72
6. Busha and Harter Research Methods in Librarianship p173
7. J Kirk 'Portrayal of Aged Characters in Australian Award-Winning Children's Novels 1946-1985' MA (Librarianship) thesis Canberra CAE 1988
8. Krippendorff Content Analysis Chapter 1
9. Jarvelin and Vakkari 'Content Analysis' p401
10. ibid p402
11. M S Stephenson 'The Canadian Library Journal 1981-91: An Analysis' Canadian Journal of Information and Library Science vol 18 no 2 1993 pp 12-14
12. M Nour 'A Quantitative Analysis of the Research Articles Published in Core Library Journals of 1980' Library and Information Science Research vol 7 1985 p262
13. Jarvelin and Vakkari 'Content Analysis' p401
14. L Kajberg 'Library and Information Science Research in Denmark 1965-1989: A Content Analysis of R & D Publications in Technology and Competence' Proceedings of the 8th Nordic Conference on Information and Documentation, Helsingborg, 19-21 May 1991 Stockholm Teknska Litteratursallskapet 1992 p236