We place a special emphasis on developing wherever possible a cordial and relaxed relationship with our clients and work together to design research and evaluations which meet their real needs.

proMENTE has experience with a wide range of research and evaluation approaches, combing qualitative and quantitative methods.

Our researchers spend a lot of time in the field and they know how to engage in communication with different groups of people; they talk easily with street children as well as with state ministers. Often our task is to build bridges between different world views, for example, between the worlds of donors and their potential beneficiaries.

Ethical concerns always come first. In particular, we take data protection and child protection very seriously. The evaluation methodology is planned in accordance with the ethical guidelines of the UK Evaluation Society (http://www.evaluation.org.uk). More precisely, this means: full respect for the protection of data of interest groups, which are fully informed about the evaluation and have access to the "Ombudsman of Evaluation" if necessary.

We are proud to employ innovative approaches in our work. As far as we know, we are the first organisation in the Balkans to implement Outcome Mapping and Most Significant Changes methodology. For many of our projects we develop websites to improve interaction with stakeholders and have long experience in employing web-based questionnaires. Most of our statistical analyses our conducted in R (www.r-project.org), the software of choice for most modern statisticians. Above all, we believe that the key to a successful project is to understand what the client really wants and needs to find out and to carefully negotiate the details of the Terms of Reference in order to make sure these results are delivered. We have been using web-based research methods since 2002 (for example at www.mojakarijera.com and www.qimpl.com) which we usually implement on our own. We use four different systems for web opinion research from which we chose the one most appropriate for our clients. Most of our research and evaluation projects include web modules. Recently we have been introducing Geographic Information Systems (GIS) approaches into our work. Our clients find that presenting data on maps can give an extra dimension of insights. Social Network Analysis is much in vogue at the moment, and with good reason. Many of our clients’ projects are delivered through networks or aim to strengthen networks. We have some simple and useful tools to help understand network data.

Nearly all our statistical reports follow the principles of reproducible research. This means that the final pdf file is automatically generated by computer from the original raw data following instructions written by us in a computer script. So there is neither any cutting-and-pasting or editing of data in the data files and nor is there, for example, any manual editing of data or graphics. This makes mistakes much less likely; the original data is untouched; our clients or other researchers can check exactly how we arrived at these results. Given the same database, software and script they could also produce the same report with one click. And the software is free and open-source. Reproducible research is transparent, verifiable and less prone to error.

We include stakeholders in our research, evaluation and training wherever possible and try to engage interviewed stakeholders in active debate about how to improve the quality of people’s lives.

Reliability and validity of information and conclusions drawn from it are of the highest importance. The validity of information is mirrored in its relevance and appropriateness for defined research questions, as well as in the directness and strength of association with the concepts which are subjects of the evaluation process. proMENTE uses all available information within the evaluation process, whose validity may sometimes be weak. In those cases, we use a wider range of measures in order to to reduce dependance on any one of them and solve the problem of data validity.