To interpret statistics we must know how they were produced.
Here I look at the statistics in the World Report. Below, in the Appendix, I reproduce the methodological sections most relevant for data on the number of libraries.
This information constitues only a small part of the report, however. The main purpose of the series is to report on freedom of – and access to – information throughout the world.
Statistics on the number of public and research libraries have been collected since 2005.
The number of public libraries was from the start defined as the number of different service points.
Research libraries were defined as the libraries of educational institutions such as schools and universities. In 2007 separate subcategories for
- university research libraries,
- school libraries and
- government‐funded research libraries
were created. Another addition was the question regarding the source of these numbers. In 2009 the third category was also called statutory research council libraries.
The 2001 report looks very different to the 2003 and 2005 editions [sic].
Each country included in 2001 was accompanied by short factual information and statistics collected from various international sources and edited by the IFLA/FAIFE Office. Aside from this, the reports differed widely on content and length, each one a product of the individual contributor. The content of country reports in the 2001 edition differed greatly, making comparisons country-by-country or region by region difficult.
Evaluating the report in 2002, the IFLA/FAIFE editorial team recommended that future World Reports would gain by being based on factual information that would make comparisons and follow-up on developments in individual countries and regions easier. The use of questionnaires was identified as an appropriate working method combined with monitoring, reports and research of incidents and violations of intellectual freedom in individual countries.
The 2003 methodology was therefore the first time a questionnaire process had been used for the World Report. This method also allowed for a shorter
production process safeguarding the topicality of the information provided by respondents. The questionnaires used were carefully drafted and pretested, and focused specifically on Internet access in the international library community, with a final section focusing on the adoption of codes of ethics and IFLA initiatives. …
The questionnaires were sent as Microsoft Word documents via email to national library associations in each country where possible, followed by national libraries, IFLA institutional members and then individual members where the national library association was not contactable or did not exist. In all cases, senior library professionals were targeted to answer to the questionnaire.
The use of a questionnaire was a success, and respondents generally had few problems completing and returning it via email. There were, however, occasional problems in sending back the completed questionnaire as an attachment. Time and effort was spent writing emails explaining to a number of contacts how to correctly open the attachments and fill out the questionnaire.
Questionnaires were also returned by fax, or scanned and then sent as an attachment. In these cases handwriting was sometimes difficult to interpret. Language was also understandably an issue.
Resources to deal with French speaking African countries, for example, were simply not available, nor was it possible to translate the questionnaire into Russian or Chinese and responses back into English. Thanks to the help of a colleague in New York a Spanish translation was eventually sent out to 22 Spanish-speaking countries and this helped increase the overall response rate.
Another problem was dealing with faulty email addresses and the constant search for an address that worked. The Internet offered the data collection process a number of advantages such as the ability to send the questionnaire instantly and cheaply to hundreds of potential respondents, but many of the addresses did not work and many of the countries included in the mail out suffered from telecommunications problems that impaired their ability to receive and send email. A large amount of time and effort was expended on following up contacts, finding new contacts and persuading people to participate in the exercise.
The first page of the 2005 questionnaire differed quite a lot from the 2003 version.
In 2003 the responding institution remained anonymous and in the country reports only the name of the country was displayed. It was felt at the time that the anonymity this process gave might encourage respondents to reveal information that maybe they would hold back if they were named. This approach did not yield any extra information however and, following an evaluation meeting of the FAIFE editorial work team in 2003, it was decided to ask for the responding institution’s name to add to the country report entry. It should be stressed though that the option of remaining anonymous was still open – but in the end this option was taken by only two respondents this year.
The other difference on the front page was the invitation to estimate the number of public and research libraries in the respondent’s country. In 2003 respondents were asked to estimate the percentage of their public and research libraries offering Internet access but without the background (i.e. the number of libraries being referred to) information this response was deemed to be lacking something.
This year, therefore, respondents could state the number of public libraries – defined as public library service points, including branch libraries – and research libraries – defined as the libraries of educational institutions such as schools and universities in their country.
Not only would these answers give some context to the percentage of libraries in each country offering access to the Internet, but FAIFE could also take advantage of the data collection process to collect the most up to date statistics from IFLA members around the world relating to the size of a country’s library service.
Section 2: Estimated number of libraries
In addition to the 2005 questionnaire, this section asked respondents not only to estimate the number of libraries in the two categories of public and research libraries, but also to provide an even more detailed account of research libraries in their country.
Separate subcategories for university research libraries, school libraries and government‐funded research libraries were created. Another addition was the question regarding the source of these numbers. The reason for including this question was quality control, i.e. to disclose whether the numbers were based on a guesstimate or on a survey, research and/or official figures.
Source: IFLA/FAIFE World Report Series
The previous Reports in this series were branded as the IFLA/FAIFE World Report since the emphasis was primarily on FAIFE-related issues (see http://www.ifla.org/en/faife/world-report).
At the request of IFLA Headquarters (HQ) the questionnaire was expanded considerably, as explained below, to include non-FAIFE-related issues. IFLA HQ therefore decided that the Report should be rebranded as the IFLA World Report. The current Report nevertheless still has a strong focus on FAIFE-related issues.
Data collection for the World Report 2010 started in April 2009 and continued until the end of 2009, with a number of countries asking for extension into 2010. Data for the individual country reports therefore to a very large extent reflect the situation as in 2009. Therefore, in the comparative tables, the 2009 date was kept since this is the more accurate date for the data from most countries.
Additional research for and authoring of the country reports took place in the latter half of 2009 and the first half of 2010. The World Report is launched at the IFLA World Library and Information Congress 2010, 76th IFLA General Conference and Assembly, in Gothenburg, Sweden, in August 2010. The 2007 Report for which the data was collected during 2007, was also only launched the following year, at the IFLA WLIC in August 2008.
2.2 Section 2: Library statistics
The questions in this Section asked data about the national libraries and data for the numbers of public and research libraries per country.
The current questionnaire differs from the previous one in that questions about the existence of a national library in a country and its functions were included, as well as whether the national library has an online presence. Of the 122 respondents, 112 indicated that they do have a national library and 10 that they didn’t. 91 respondents provided web addresses for their national libraries. In cases where no national library exists, the typical functions of a national library are performed by, for example, a public library or a university/college library.
The category of research libraries was subdivided into university research libraries, school libraries and government‐funded research libraries (statutory research council libraries), as in 2007.
These numbers were not analysed in any detail in previous reports and comparisons with the 2007 Report is therefore problematical. In addition, before 2007 only one category of research libraries was defined, which included all the sub-categories defined since 2007; comparisons between the different reports are therefore problematical.
A superficial comparison indicates that library numbers have remained fairly constant during the past two years, with, on average, limited growth or decline. In a few cases numbers differed fairly substantially, but this seems to be linked to different respondents or more updated statistics being available.
These numbers are not further analysed.
Source: IFLA World Report 2010. Analysis and conclusions. PDF file.