The object of this study is to determine how people interpret measurement data, which factors influence their interpretation and which do not. This study forms part of an investigation series undertaken at the paper pulp company Södra Cell in Sweden (N = 1200 employees), the investigated parameter being "paper pulp quality". The findings presented in the study are the outcomes of a questionnaire that was repeated four times (n1 = 54, n2 = 53, n3 = 43, n4 = 38), an interview (ni = 32) and regular group discussions that included 10 people (nd = 10) and that took place once a month over a three-year period. In order to quantify the quality of the product - paper pulp - different measurements must be combined. This study reveals that different employees prefer different parameters for the quantification of paper pulp quality. This study furthermore shows that there is a correlation between the employees' choice of parameters and their belonging to a particular pulp mill within the company. We also found that external factors, e.g. the opinion of authorities, affected which parameters were favoured when product quality was determined. Furthermore, the data analysis indicates that there is a correlation between the job-position of employees and the way they interpret measurement data related to product quality. Controllers and operators make similar judgments, with the exception of chemists whose opinions appear to deviate. However, the greatest variation found in the study is related to the individual person - independently of education or physical or psychological condition. The study finally proposes a number of preventive measures to improve the uniformity and reliability of measurements conducted at a process company, the most important of which is improved education of all employees who are in a position to influence product quality. © 2007 Elsevier Ltd. All rights reserved.