Also, no amount of post-research data maneuvering can save your conclusions if your data collecting method is seriously flawed and the data corrupted. If you work with corrupted data, you get unreliable results, plain and simple, as I believe you researchers definitely understand with all your academic training. Pretty basic stuff, right? That is until you start getting attached to your results: they just fit your cultural presumptions so well. Northern European countries are populated by altruistic angels and east Asia, by petty shoplifters who steals napkins from fancy restaurants. Also, what can you do? All those years and research funds for nothing? Better make the results sound plausible, well, at least to a western audience. What a bummer it’ll be if no significant difference is found?