What is Big Data (and what is not) and why is it useful in education?

The Big Data specialist, Walter Sosa Escudero, director of the Economics Department of the University of San Andrés, spoke in front of the TICMAS booth of the Book Fair on the growing role of collecting money. data and its uses.

Sosa Escudero was interviewed by Patricio Zunini of TICMAS, a modular platform that supports students, teachers and educational institutions in digital transformation through a personalized teaching model.

What myth does Big Data have?

It has a lot, like any technology that challenges us. He has fanatics to unbelievers, with very few people in the middle.

The myth is that Big Data will solve all the problems of traditional science and daily tasks. Now that we have a lot of information, any question will be solved by the data. But this idea has an excess of trust and a lot of mythology.

It is the proliferation of data resulting from a spontaneous interaction with interconnected devices. But they are generated data, not for the purpose of creating them, unlike administrative records, surveys, or the results of a laboratory experiment, which have been accurately designed to be generated.

Can this huge amount of data make you adopt a certain behavior?

It's one thing to predict and another to influence. For example, with big data data and algorithms, based on the intensity with which people use umbrellas or rain drivers, I can predict the rain. But I can not rain. Although I can predict well that it will rain and that it will not rain, I am not even able to understand why it is raining. It also works with social problems. I can predict your political intentions with a high percentage, but from there, to manipulate them or turn them into something else, it is very different.

With satellite data, I can measure the poverty of a country. Which means great progress to measure, to draw conclusions to understand and finally to design state policies to overcome this problem. But one thing is to measure and another to understand. What Big Data is far from it, for the moment.

How long would it take?

I would not be optimistic there. It's not a question of lack of data, it's a question of lack of ideas. The history of science does not mean that data and induction go hand in hand with deduction and generalization. For example, many gravitational theories about the motion of the planets come from a work of several hours of observation at the telescope.

The same thing happens with Big Data. Observation probably gives us better data on where the poor are, how they moved, and so on. But from there to designing educational policies, work, etc., there is another step. It's not a problem of little data, but of few ideas.

What data does a state review?

It's a question that changes depending on the year. The systematic measurement of poverty in Argentina only concerns the last 25 years. The data examined by the government are very traditional, such as GDP, household survey data and prices to measure inflation. They start using data from the Internet. This is a growing trend.

Are the resulting data, for example poverty, conditional?

Does this condition people or the state? The measure of poverty is a very thick brush. To say that in Argentina there is 35% of poverty, it is always a very aggregated average of very different situations. And you do not need to go to someone's home so that the poverty rate increases exponentially if you break it down into a specific item or, on the contrary, it also decreases. That's why I do not think it conditions.

The big contribution of Big Data is to be more specific about these grotesque figures. With the great advantage that we can illuminate where there was no light and the great disadvantage of finding more areas more conditioned.

What big data projects can intervene in the field of education?

There is one that is a sum of projects called Mind Spark, which came into being in India. It is a dynamic learning project, which badyzes the features of the student and helps them. The system learns as it is executed. This has many educational benefits.

Has Big Data been applied to an Argentine educational project?

There are very isolated and small projects. These are pilot projects. But it is far from being a state policy as in India. This would serve to democratize education. On the other hand, you can fill a clbadroom with computers, but you must provide teachers with the tools to understand its use.

The myth is that the phone listens and provides data on everything we do …
From a certain point of view, we are strange bugs, unpredictable and, on the other hand, very predictable. Having a lot of data and getting things done is a big step forward.

We have a social technology such as Universal Assignment by Child (AUH), supported by all political parties. But if you ask me how it works, why it works and what the effects of its implementation have been, I tell you we do not know it. Or we know it with very thick brushstrokes. I am optimistic, not because of the proliferation of data, but because there are questions we still can not answer.

Often, Big Data provides you with an ocean of precise answers to very stupid questions.