Certain tendencies' is very general

Yes, you are right. You may consider that particular use of that particular term from me in that case to mean the characteristics that we had previously been discussing, ie: fundamentalism, extremism, and violence. That is what I intended the term to mean.

Another point that you have never answered or even addressed as far as I see is my argument that with your thinking no valid research could be possible for instance on the influence of gender, social class and similar variables upon for instance proneness to violence, for these variables are as well confounded with others. You seem to agree (that's how I understand you) that 'impoverished background' may be a factor explaining that a disproportionate amount of blacks are in US prisons. To state that with some confidence one has to separate the influence of race and social/money etc background by use of a mathematical model. This is completely symmetric, for the same way of looking at the data allows you to factor out race alone.

The problem arises because, in the case of this particular question, there are too many variables (not just two or three), and any decision to leave any of them out would be arbitrary, not objective, and not conducive to producing reliable results. There is no scientific way to define the parameters without making non-scientific judgements. Value judgements.

It would not be possible to determine how to define the terms without making arbitrary distinctions (for instance, how do we define "violence"? How do we define, "fundamentalism"? How do we define "extremism"?). And how do we determine whether or not we define an act as being religiously motivated, or as being motivated by political realities on the ground but being acted by someone from a particular religion?. All of these things require the people conducting the study to make value judgements. And value judgements are, in the final analysis, nothing more than opinions.

Also, you can't really select for a time frame and get a reliable answer because each time frame will produce different results, but the results over time are what will give you accurate information. However, it is not possible to collect the needed data for many of the time periods in the past. In fact, it's probably not possible to collect sufficient data from this time period either, since we are talking about all of the people in all of the countries of the world. Leaving out some populations but using others will also produce different results, in the same way that leaving out certain time periods would.

Just BTW, the modeling for which factors influence the earth's climate is much more difficult and has much more problems and it still is done (and, of course, criticised, as is the good rule in science). But a position that it should not be done because it could be difficult has no supporters.

Modeling for which factors influence the earth's climate, done badly, would not cause discrimination against whole groups of people. The kind of study you are trying to defend in this thread, if done badly, would cause discrimination against at least one whole group of people (or put an "official", "scientific" stamp on already existing prejudice and discrimination).