 We introduced a novel metric for hypergraphs that takes into account both the distance between pairs of nodes and the distance between groups of nodes. This metric can be used to calculate the efficiency, closeness and between the centralities of nodes in a hypergraph. When applied to real-world data, this metric revealed differences between the values of these centralities obtained from the original hypergraph and those obtained from its clique projection. These differences were more pronounced when the hyperedges had larger sizes and fewer connections between them. This article was authored by Ekaterina Vasilieva, Miguel Romance, Ivan Samalanco and others.