Tell mi if it is a foolishness.
Nowadays, Internet websites (expecially forums and social networks) record an huge amount of quantitative data about humans' social behaviour. As I know, no studious has already massively analyzed these data to make up sociological theories.
Sociology is the social science by definition, but with these resources it could become more mathematical, more or less as the economy; and this could improve our understanding of social interactions in our species.
What do you think?

Sure, off you go.

What do you think will be a meaningful analysis?

The internet does not suffice as a meaningful basis for discerning human behaviour, precisely because 'being on the internet' is indicative of only one kind of human behaviour (being on the internet). Okay, one may be able to access and assess the opinions/likes/dislikes of those who are willing to give them, but to what extent?

The OP assumes several things about the internet:

1) It is frequented by a huge population of humans.
2) All/most of them either do, or are willing, to give a honest and extensive evaluation of their everyday behaviour.
3) That something definitive about humanity as a whole can be hypothesised via analysis of such data.

However:

http://www.internetworldstats.com/stats.htm

... This recent data suggests that, even now, less than a third of the world's population has access to the internet. What it doesn't tell us, is how many of them have regular access to it, nor how their opinions are constrained by the particular restrictions/impositions/fears/etc. which are specific to their locale.

Indeed, one should consider that even we in the 'free West' understand that it is not okay to give an honest opinion about everything, given the taboos and political correctness which are currently moulding Western thought. It would not be wise, for example, to proclaim to the internet-community that one was a facist/nazi/racist/paedophile/etc.. The point being that even 'the lands of the free' impose their own paricular kinds of restrictions upon what we are allowed to be honest about.

The bottom-line is that any data we have of human behaviour, discerned via a meta-analysis of global [internet] input, is not going to unveil anything significant about humans themselves other than the fact that [in general] their input does not transcend social/geographical/political/historical/legal/religious concerns. However, we already know that [now]. The internet just serves to emphasise this knowledge.

Sociology will never be [purely] like maths/science, even if it partakes of it, because there is no maths or science to Marxism/religion/Nazism/justice/equality/love/hate/fear/etc..

Sociology is a mongrel, in that it feeds from numerous/diverse practices (I've just studied it as a 3rd-level module for my degree, so am fully-aware of its extensiveness and limitations).

DanDare wrote:What do you think will be a meaningful analysis?

For example, is there a relation between the number of users and the average number of staff members in a community?
Or, in what way does the average number of "friends" in a social network decrease with the distance?

jamest wrote:The internet does not suffice as a meaningful basis for discerning human behaviour, precisely because 'being on the internet' is indicative of only one kind of human behaviour (being on the internet).

Even so, it could be interesting. Maybe there is a correlation between what one does on the Internet and what he does in other situations.

jamest wrote:This recent data suggests that, even now, less than a third of the world's population has access to the internet.

Even one undredth of the human population is an huge sample.

jamest wrote:Indeed, one should consider that even we in the 'free West' understand that it is not okay to give an honest opinion about everything, given the taboos and political correctness which are currently moulding Western thought. It would not be wise, for example, to proclaim to the internet-community that one was a facist/nazi/racist/paedophile/etc.. The point being that even 'the lands of the free' impose their own paricular kinds of restrictions upon what we are allowed to be honest about.

The bottom-line is that any data we have of human behaviour, discerned via a meta-analysis of global [internet] input, is not going to unveil anything significant about humans themselves other than the fact that [in general] their input does not transcend social/geographical/political/historical/legal/religious concerns. However, we already know that [now]. The internet just serves to emphasise this knowledge

Don't it applies also off from the Internet?

jamest wrote:Sociology will never be [purely] like maths/science

I agree, but might it approach to economy?

jamest wrote:Sociology is a mongrel, in that it feeds from numerous/diverse practices (I've just studied it as a 3rd-level module for my degree, so am fully-aware of its extensiveness and limitations).

Thanks for correcting my crap.

Microfarad wrote:Tell mi if it is a foolishness.
Nowadays, Internet websites (expecially forums and social networks) record an huge amount of quantitative data about humans' social behaviour. As I know, no studious has already massively analyzed these data to make up sociological theories.
Sociology is the social science by definition, but with these resources it could become more mathematical, more or less as the economy; and this could improve our understanding of social interactions in our species.
What do you think?

My bold: Why should we want sociology to be more mathematical?

Sociologists research soft concepts/relations mostly. Such as corruption & trust, culture & happiness, organisational structure & efficiency, social mobility & poverty, social capital & wealth, migration & integration patterns et cetera.

These things can be measured of course and they are, but sociologists shouldn't run the risk of becoming like economists who pretense they can reduce complex social structures to workable simple mathematical models. I think economics should become less mathematical instead

I think the most important thing for sociology is to establish a stable set of fixed definitions that is worked with globally, in different contexts and for an extended amount of time, so that a decent base of data will be built from which comparisons can be made. It may even be possible to draw meaningful conclusions one day or...the sociologist's wet dream..to predict the future accurately!

Sorry for late, I gone on holiday.

mindhack wrote:My bold: Why should we want sociology to be more mathematical?

Because it could make precise predictions. Is not this the aim of science?

mindhack wrote:These things can be measured of course and they are, but sociologists shouldn't run the risk of becoming like economists who pretense they can reduce complex social structures to workable simple mathematical models.

I don't know if this is possible for sociology. But if it is, a simple mathematical model would be better than no mathematical models, in my opinion.

Microfarad wrote:Sorry for late, I gone on holiday.

mindhack wrote:My bold: Why should we want sociology to be more mathematical?

Because it could make precise predictions. Is not this the aim of science?

Yeah...not in sociology really. From personal experience I say that the research I've encountered was often aimed at just explaining observations as precise and meaningfully as possible.

The predictions part is difficult for many reasons, but most notably the abundance of variables in involved social reality, contexts are often unique, variables often difficult or ambiguous to quantify so room is left for subjective interpretation. There are more difficulties..

mindhack wrote:

Microfarad wrote:I don't know if this is possible for sociology. But if it is, a simple mathematical model would be better than no mathematical models, in my opinion.

These things can be measured of course and they are, but sociologists shouldn't run the risk of becoming like economists who pretense they can reduce complex social structures to workable simple mathematical models.

There are many models, mostly regression models of all sorts. Methodology claims a huge part of the study. But it's just really difficult or unrealistic even to think there're social laws in neat little functions like E=MC2.

Sociology is more like Environmental science, an applied science to support professionals in the field with advice and rigorous explanations.

This subject reminds me of Science gleans 60TB of behavior data from Everquest 2 logs | Ars Technica

Noshir Contractor described how the data was allowing him to explore social network dynamics within the game. He described a variety of factors that are thought to influence the growth and extent of social networks, such as collective action, social exchange, the search for similar people, physical proximity, friend-of-a-friend (FoaF) interactions, and so on. Because these are well-developed concepts, statistical tools exist that can extract their signature from the raw data by looking at interactions like instant messaging, partnerships, and trade.

This seems like the sort of thing that one can research using various Internet resources.