**Purpose:** We aim to extend our investigations related to the Relative Intensity of Collaboration (RIC) indicator, by constructing a confidence interval for the obtained values.

**Design/methodology/approach:** We use Mantel-Haenszel statistics as applied recently by Smolinsky, Klingenberg, and Marx.

**Findings:** We obtain confidence intervals for the RIC indicator

**Research limitations:** It is not obvious that data obtained from the Web of Science (or any other database) can be considered a random sample.

**Practical implications:** We explain how to calculate confidence intervals. Bibliometric indicators are more often than not presented as precise values instead of an approximation depending on the database and the time of measurement. Our approach presents a suggestion to solve this problem.

**Originality/value:** Our approach combines the statistics of binary categorical data and bibliometric studies of collaboration.