Towards a Logic of Meta-Analysis
- Uncertainty, vagueness, many-valued and fuzzy logics-General
- Knowledge representation languages-General
We currently have access to a plethora of statistical analyses based on sampling limited parts of a population. Meta-analysis is the task of combining several statistical results to obtain a more precise and reliable picture of the population. By the nature of sampling, all these results are uncertain, and difficult to combine with other knowledge. In this position paper, we propose a first approach for automated reasoning in meta-analyses.