Incorporating nonparametric statistics into Delphi studies in library and information science

Introduction. The Delphi technique is widely used in library and information science research. However, many researchers in the field fail to employ standard statistical tests when using this technique. This makes the technique vulnerable to criticisms of its reliability and validity. The general goal of this article is to explore how nonparametric statistical techniques could mitigate this drawback and be incorporated into Delphi studies in library and information science. Method. To reach that goal we provide an example study which used a Delphi method for data collection and nonparametric statistics for data analysis. The example study investigated the barriers and challenges encountered by scientists when using various information and communication technologies for their distributed collaboration activities. Twenty-four participants were recruited into two domain groups (social and behavioural sciences vs. science and engineering). Results. After three rounds of data collection and analysis systematically using nonparametric statistical measurements, fourteen items were identified as the most important. Different rankings of the items were observed between the groups involved, including Kendall's Ws. Conclusions. The incorporation of the nonparametric statistics into the Delphi technique may enhance its rigour as a research method and provide evidence for its reliability and validity.