I would think MARKETS generate all sorts of #s

In ML we play with #s for classifying used for prediction and projection , this helps to accommodate the "control" factor . Or other variables that were assumed and then change so that there is a leveraging .

I would think MARKETS generate all sorts of #s

In ML we play with #s for classifying used for prediction and projection , this helps to accommodate the "control" factor . Or other variables that were assumed and then change so that there is a leveraging .

Games.

Following to know

All possible data outlets (controlled and uncontrolled environments) from real world to virtual world.

I support what Brother Dickson said; data can be gathered from both human and nonhuman activities.

this link is useful

https://en.wikipedia.org/wiki/Machine-generated_data

regards

The term data generating process is used in statistical and scientific literature to convey a number of different ideas: the data collection process, being routes and procedures by which data reach a database (particularly where these may change over time);

Unlike natural science, I think that it is difficult to take concrete data in the economic field.

Previously, I worked at the central laboratory of the company.

So I asked a similar story from a plant engineer.

It is completely unknown why the defect of the product is generated. Multiple regression analysis was performed with a possible factor as a parameter, but the cause was unknown. However, it seems that the cause unexpected became clear by the hint of the memo of the engineer of the plant.

Anyway, it seems to be important to observe first.

In physics reality has become something like a "wave collapse" - depends on when you're looking and where your viewpoint stands, the range of expectation, etc data are in need of a language, (math for instance) - history is generating facts - every science looks for "significant things" - and facts get significance by history.

I don't think I am helping you out here, dear prof.Ahmad Battal but I can very well understand the difficutly of this question.

I think also of Thaler's distinction between "Econs and Humans".

Data are very much that what people (economists) are willing to take in consideration. That what "counts".

best wishes

Willy

I think data is everywhere, you need to ensure that you firstly form the research question, followed by the techniques that will ensure that whatever data you get is deemed acceptable. It is important that how you get your data is ethical, as this needs to be known. Your test subjects also need to know that they are part of your study, it's 101 of ethics in data gathering.

Generally, you gather data with instruments such as surveys, questionnaires and interviews, depending if your research is qualitative/quantitative. I like to use automation tools such as computer mediated surveys.

For economics, you want to explore the domain of social sciences, so, predominantly, your data will be derived from instruments or government/institutional sources and then you will have to test it. Note that you need not necessarily have to address causes, since available institutional data sources already derive such conclusions. You might need to emphasize on relationships to actually work with this kind of research.

Either way, I would probably suggest working on relationships between data sets, as in economics, there seems to be an overload of data that is available emphasizing on causes (ie World Bank/IMF etc).

My humble opinion is also that experiments in economics tend to be difficult due to the scale of such endeavours, ideally, you can just study trends. Micro economics tends to encompass other areas in the social sciences whilst studies in Macro economics are way to big for specific experiments to even be undertaken (unless you have massive funds to expend).

Hope this helps.