Is this an example of “how to do a poor research by simply using numbers that fit your theory”? Using just one ranking is a terrible idea since the ranking itself can be biased as well as being a poor representation of reality. In this case, to create its ranking, Triposo asked just 700 people from 62 countries that uses its service. This means just 12 people per country but more likely that a number of countries had just 1 or 2 people answering. These numbers are nowhere near enough to have a good representation from various countries. And as far as I can see without installing its app, Triposo seems to be English language only service. This means that those that answered for the ranking are likely to be mainly English speakers who would be likely to be traveling with fellow English speakers. So the fact that US and UK finished in worst 2 with Australia in 5th can be not due to how bad they really are, but how frequently Americans, British, and Australians saw fellow Americans and British and Australians behave poorly.
What you need to do is find or do a survey that asks tourists themselves on how bad they think they are compared to others and then compare with how bad hotel staffs, restraunt staffs and tourist spot staffs rate each tourists from each country. For example, if tourists from a country think that they are bad and staffs agree, this shows that the self image and reality matches. This is bad but can be improved because tourists know that they need to improve. If another group of tourists thinks that they are OK but staffs think that they are terrible, this shows a bigger problem because this group of tourists is unaware or may even believes that they are discriminated.