How to compare apples with bananas
Bandy will likely achieve high media audience in the Olympics.
What we know is that bandy achieves high audience numbers when sent on national broadcasters. Moreover, those ball-sports achieve the highest television audience of all sports categories. Finally, that bandy has high participation and achieves higher tickets sales on stadiums. All this suggests that bandy will be successful in the Games. What’s sure is that it is not meaningful to compare audiences between Olympic sports and non-Olympic sports. It would be rejected as a circular hypothesis. It’s easy to confuse cause and outcome. The context is entirely different. It is the taxpayers that fund the Games and enable the IOC to attract 400 broadcasters. Researchers would typically try to adjust for different context.
Correlation* test of relationships
High attendance at stadiums is positive for broadcasting and considered a successful outcome of the Games. Bandy achieves high-ticket sales. The research concludes that larger stadium attendances have positive impacts on the size of television audiences (Buraimo 2007).
Statistical methods such as correlation can give an idea about the probability of high attendance in the next Olympic Winter Games (OWG). Therefore it is of interest to identify what Universality criteria actually co-vary with high attendance at the arenas in OWG. It turned out that the correlation between participation (number of athletes) and attendance had a positive correlation of 0,97. There are reasons to believe that there is a positive correlation between participation and media audience and attendance versus media audience.
*The correlation test of relationships is a principal statistical probability method to measure the strength and direction of the interdependency between two data sets. Correlation is a necessary, but not sufficient precondition to entirely conclude on causality, that is which variable causes the other to vary. In some tests, one may conclude causality if evident that one variable cannot influence another. Regardless of causality,
as long that there is a strong correlation that is deemed significant, still this correlation