Data visualization has been gaining popularity also in official statistics. Official statistics are generally considered of high quality but may not be easy to understand, particularly economic statistics that relate to complex concepts. Communication of these data requires effective data visualization to convey the right message, especially when directed to general public. The mechanisms of effective data visualization are still not entirely clear. While acknowledging findings on visual perception of graphical elements, more research seems necessary on cognitive processes involved in interpretation of visualizations. Our proposition is that the effectiveness of data visualisation depends on the graphical characteristic of a chart display, content domain requirements, and data set characteristics, while knowledge, both domain and statistical, as well as experiences and skills in the field, mediate the process. To investigate this proposition, we present results of two studies in two groups: first-year undergraduate students and Mechanical Turk Workers. Our tests included several examples of official statistics data visualizations by Eurostat, mainly from economic statistics.