The issue of quality of information is a recurrent question in information system management, notably when the systems are decentralised. This is all the more true when the information producers are not information specialists, which is the case in systems based on experience exchanges.

What type of quality control can be applied to information systems? This question is not new and has been the object of abundant research and documentation. Concretely, implementing quality control continuously runs up against the issue of means: training information producers in information treatment, time devoted to updating the data continuously, integrating monitoring and animation around content, etc.

The First Question: Quality Criteria

The difficulty is not so much a matter of theoretical formalisation of a set of information treatment criteria but rather their concrete implementation.

• Quality information is up-to-date information: how can one give oneself the means to regularly monitor the information contained in the information system.
  
  
• In a service with multiple information producers, quality is obtained by sharing information treatment modes (data format, homogeneity in ways of classifying the information such as a thesaurus): how to develop this common culture?
  
  
• When it comes to quality of the information, the ability to describe the public and uses of any given piece of information brings considerably added value: how can one lead information producers to adopt this approach?
  
  
• Quality information is complete and specific information. In the field of experience exchanges notably, the information must provide the lessons learnt from the experience, and analyse its impact and the methods used. It can not merely be content with listing the sequence of events and the results: how can one improve practitioners' ability to adopt this approach when their primary job is not communication? How can one mobilise the means and strengthen skills to do so?
  
  
• In the field of experience exchange, quality information is information from the field presented in a neutral and objective manner: how can one ensure this neutrality?
  

This list of criteria and the questions pertaining to each criterion is far from complete.

The Second Question: How to Balance Quality and Quantity?

The criteria for analysing an experience are strict in order to allow comparisons. It is necessary to forge a common information treatment culture.
However, this can make the production of information, and notably experience records, relatively time-consuming. The proper balance must be found between quality and quantity.

• Experiences that are insufficiently documented, or documented differently in function of the organisations that produce the information, too often leads to settling for insufficiently pertinent information in relation to users' expectations or information that is difficult to compare with other experiences.
  
  
• Inversely, it rapidly become clear that excessive requirements drastically reduce the number of references produced. Indeed, few organisations have the (both methodological and financial) capacity to analyse fully the experiences in which they are involved. Standards that are set too high can discourage contributions to the information service by its partners, which would go against the service's initial purpose of allowing stakeholders--who have little training analysing their observations in detail--to share their field experiences easily.
  

The Third Question: Can Quality Exist without a Formal Validation Process?

Speaking of quality in information systems means being able to display and guarantee quality.

Is it really possible to succeed in this without developing content monitoring and animation?

At what price can content monitoring and animation be taken into account without threatening the sustainability of an information service?

The InterDev service is based on an ensemble of thematic networks. Each is moderated by a coordinator who is notably in charge of validating the information gathered by the various partners. A formal validation procedure has been set up for the experience records.

Our experience after a few months of examination leads us to emphasise the following points:

• validation is vital,
  
  
• one pertinent method is sharing this responsibility between network members: a cross-validation process was therefore implemented, each partner being in charge of validating content in a specific field in function of the partner's specific expertise.
  
  
• overall moderation/coordination is also needed: it makes it possible to homogenise the various partners' different methods and guarantee the overall coherency of the service.
  
  
• the priority challenge in this field therefore lies in the ability to mobilise sustainably the resources needed for such monitoring and coordination.