You are in :

Study of the use of HE libraries by external researchers

Research specification

Background

1. The UK higher education (HE) funding bodies wish to commission a study of the use made of HE libraries by external researchers. The aim is to estimate the relative burden on HE libraries of such use in order to derive a method of allocating compensatory funding, through the access strand of the Research Support Libraries Programme (RSLP).

2. The survey results will be used to inform the allocation of funds in 1999-2000, 2000-01 and, subject to commitment of funds by the funding bodies, for one further year.

3. A preliminary study was carried out by Coopers and Lybrand in 1996-97. This established that the proportion of visits by external researchers varied greatly between libraries - from less than 0.1 per cent to more than 6 per cent. The study also indicated that external researchers make more requests for assistance than other library users. It concluded that libraries should be compensated in a way that reflects the level and cost of use by external researchers, and proposed that level and cost of use should be estimated by an independent survey of researchers.

Research objectives

Main objectives

4. The central objective of the study is to produce accurate and sufficient data on how much researchers use libraries other than in their home institution. This will be done by undertaking a survey of researchers. The funding bodies will then analyse the data and produce a ‘burden of use’ score for each institution, as a basis for deriving a method to allocate compensatory funding.

Analysis

5. Candidates may choose to explain in their tender how they might analyse the survey data, in order to rank the institutions in terms of the burden of use on library facilities by external researchers. This is entirely optional: the information will not be used to evaluate applications for the main survey. However, if tenderers do include data analysis, they should provide full details (see also paragraph 24.)

Data to be collected

6. The study should collect data from a sample of individual researchers on the frequency, nature and volume of use that they make of HE libraries other than those that are part of their home institution. The data must relate to the libraries of each higher education institution (HEI) in the UK (England, Scotland, Wales and Northern Ireland). In addition, it will be necessary to analyse the data by the researcher’s discipline(s). This will be defined as the Unit of Assessment used in the 1996 Research Assessment Exercise (RAE).

7. The following paragraphs outline the data that it is envisaged will be collected by the survey. Tenderers are asked to comment on the suitability of the proposed data, and to suggest alternatives if appropriate. Alternative suggestions must be accompanied by a discussion of their relative merits.

8. It is proposed that frequency of use should be measured by the number of personal visits to a library, plus contacts by telephone, e-mail and fax over a fixed period. The survey should obtain the total number of instances of use of each library by each researcher over the previous 12 months.

9. The amount of time spent in each library in the last 12 months should be estimated, and should be broken down into:

  • time spent accessing or using collections,
  • time spent consulting staff,
  • other time,

In addition:

  • The number of individual visits to each library,
  • The number of other contacts (for example, by letter, telephone, fax or email but excluding remote catalogue access) with each library should be enumerated.

10. Researchers should also be asked to identify and rank up to five ‘most important’ libraries for their discipline(s) in the UK, including libraries outside the HE sector.

The survey population

11. The survey should be designed to produce results that represent the use made of HE libraries (other than home institution libraries) by people who are either:

  • Research active staff employed by an HEI, or
  • Third-year research students (full-time or part-time) registered with an HEI.

Suggested methodology

12. Methods used for the survey should be of the highest possible quality and must be defensible with respect to the research aims of the study. Both the design and the implementation of the survey must be able to stand up to scrutiny and criticism. With this in mind, the contractor will be responsible for documenting fully the methodological aspects of the survey.

Sampling

13. The Higher Education Funding Council for England (HEFCE) holds a database of all research-active staff involved in the 1996 RAE. This might be used as a sampling frame for research staff in the survey. The client would be able to select the sample to a design agreed with the contractor. The database includes indicators of discipline, age and institution for each researcher; tenderers are asked to consider how these might best be used in sample selection.

14. To comply with data protection requirements associated with use of the RAE data, tenderers should note that sampled researchers must be offered the chance to opt out of the survey before being contacted in person or by phone. It is anticipated that such initial contact would be by letter. If a postal survey is proposed, then this letter could be included in the same mailing as the questionnaire.

15. One way of sampling research students might be to ask each sampled member of staff for the names of students for whom he or she is responsible as tutor or supervisor, and then to sample from those. This aspect of the study is likely to require considerable thought, to overcome the problems associated with gaining a sufficient and minimally biased sample of this group of the survey population. Tenderers are asked to comment on this approach and are encouraged to suggest alternatives.

16. A total sample size of around 5,000 completed interviews or questionnaires (research staff and research students) is envisaged. Tenderers are invited to comment on the balance between the two component populations on which the survey should be based in order to achieve sufficiently robust results. The 1996 RAE identified approximately 48,000 research active staff and approximately 55,000 postgraduate research students associated with those research staff. (Both these figures are full-time equivalents.) The total number of enrolled postgraduate research students as of 31 December 1996 was around 100,000 - the extra 45,000 being accounted for by students whose tutor or supervisor was not included in the RAE. Tenderers are invited to comment on the importance of including this group in the survey, and to propose methods for doing so, if that is felt desirable.

Data collection

17. Telephone interviewing is likely to be the most cost-efficient method of data collection. (The RAE database does not include phone numbers, so the contractor would be responsible for obtaining these.) However, tenderers are invited to suggest alternative methods, particularly for the research students, who may be particularly difficult to contact by telephone.

18. Tenders should be prepared on the assumption that survey interviews will last no more than 12 minutes on average.

19. Tenderers should describe the training and competence of the interviewers who would work on the survey, and the times/days of the week when the interviewing would be carried out. Tenderers should also state the number of fully trained interviewers available to work on the survey. The tendering organisation should also describe its general experience of telephone interviewing.

Response rates

20. The survey must obtain at least 5,000 completed interviews or questionnaires. Tenderers are asked to state the response rates that they anticipate achieving, to describe the methods that they would use to maximise response, and to show evidence of having achieved good response rates on similar surveys.

21. Tenderers should also outline how they might assess the impact on the survey results of any non-response bias.

22. The client will require the contractor to account for every member of the selected sample, in terms of the outcome of the attempt to gain an interview or questionnaire response. The outcome categories will be agreed with the client. Every two weeks during the fieldwork period, the contractor will provide the client with a count of the number of sample members in each outcome category. The contractor will incorporate the final counts into the survey technical report (see paragraph 25). The contractor will also provide a table classifying all sample members who remain uncontacted at the end of the survey period by the number of call attempts made. There should be no sample members for whom this number is less than four, unless it is ascertained that there is no possibility of contact during the fieldwork period (for example, if the person is abroad for three months).

Piloting

23. The survey methods and data collection tools should be adequately tested and piloted in order to minimise the risk of unforeseen problems. Tenderers should state their proposals for testing and piloting and allow for these in their costs and timetable.

Staffing

24. Tenderers should outline the qualifications, experience and expertise of key staff who would work on the study. The workplan should contain details of the breakdown of tasks including timing, duration, and which team member(s) would be involved. Experience of carrying out statistical surveys to high standards, within time and resource constraints, is essential. Knowledge of the HE sector is essential, and knowledge of issues surrounding library use would be an advantage. Tenderers should nominate a contact person who would be responsible for all day-to-day liaison with the client.

Outputs

25. The contractor will supply four main outputs from the survey to the client:

  1. A complete, fully-documented, data file in both Excel and SPSS formats. For each sample case, the file will include the data from all the survey questions, any derived variables, and any relevant information from the sampling frame, such as indicators of institution and discipline.
  2. A matrix of the data on the three variables set out in paragraph 11 above, by libraries used and by institution, with marginal analysis of the variables.
  3. Marginal analysis of the data set out in the matrix.
  4. A technical report, documenting methodological aspects of the survey: sample design, response to the opt-out procedure, response to the survey, fieldwork procedures, fieldwork period, number of interviews per interviewer, assessment of non-response bias, interview content, interview length, and so on.

26. If tenderers choose to include data analysis in their bids (see paragraph 5) this would represent one further output.

Timetable

27. The deadline for receipt of tenders is 1200 on 4 December 1998. Tenderers may be invited to present and discuss their proposals on the morning of 17 December. It is intended that the contract should be awarded as soon as possible thereafter, so work can start at the beginning of January 1999.

28. The client must receive the final clean data file, a matrix of the data and marginal analysis of the variables by 30 April 1999, and the technical report by 14 May 1999.

29. Tenderers are asked to propose a detailed timetable for the study design, piloting, data collection, data preparation, analysis and reporting.

Costs

30. The tender document should present the costs (including VAT if applicable) of carrying out the study, broken down into different types of expenditure and phases of the study as appropriate.

Organisational details

31. The tender document must include the name, status and address of the tendering organisation, plus a contact name, contact address (if different), telephone number, fax number and e-mail address.

Delivery of tender

32. Tenderers should send nine copies of their proposal to Paul Hubbard at the Higher Education Funding Council for England, Northavon House, Coldharbour Lane, Bristol BS16 1QD, to arrive no later than 1200 on 4 December 1998. Fax and e-mail copies will not be accepted. Any queries or requests for clarification of points in this document should be made to Ronald Milne, Director, Research Support Libraries Programme, telephone 0131 651 1493/1494 or e-mail ronald.milne@ed.ac.uk

HEFCE
10 November 1998

Last updated