• A
• A
• A
• ABC
• ABC
• ABC
• А
• А
• А
• А
• А
Regular version of the site
Campus inMoscow

# ‘The Future Belongs to Network Analysis’

What do staff efficiency, power of the Medici family, and the Ebola epidemic have in common? It is that they can be studied with network analysis. In 2017, HSE launched a new English-taught master’s programme ‘Applied Statistics with Network Analysis’. Valentina Kuskova, head of the International laboratory for Applied Network Research, told the HSE news service how network research works in social studies.

The programme is run under the supervision of Professor Stanley Wasserman, on the basis of the International laboratory for Applied Network Research. It implements the structure and the subjects characteristic for master’s programmes in statistic consulting and applied network analysis taught at U.S. universities.

Valentina Kuskova

## Applied statistics and network analysis

We are Russia’s first programme in applied statistics. How is applied statistics different from the traditional one? Classical statistics teaches us to rely on the method itself and to prove theorems. Applied analysis, on the contrary, relies on the data available and on the solution of the specific task. We work with real data, from which we have to learn to mine information; we are guided not by the method, but by the theory and by the problem.

Network statistics, which takes up a big part of our curriculum, is the same old statistics, based on the same laws and formulas, but it offers a different level of analysis, including not just isolated observations, but at least pairs of observations and possible interactions between them. We learn not only to analyze the objects for which we’ve collected the data, but rather to find links between them. That’s why network methods are more powerful than simple analysis. They include the opportunity of a relativist approach to data analysis, and allow the creation of a more insightful picture, be it the analysis of bloggers’ influence on their subscribers, public attitude to a new governmental initiative, or a task of explaining low labour efficiency in a company team.

## What can be studied with network analysis

A social network in scientific understanding is a result of interaction between an individual and their environment. For example, if we analyze communication channels in companies, we can influence the effectiveness of joint work and the quality of outcomes in the company. If we study the network structure of the elites in the 15th-century Republic of Florence, we can track how the Medici family, which wasn’t extraordinarily rich, managed to take a key position in the network of Florentine families and came to exclusively control all political and economic processes (the network of relationships in Florence is a classic example of network relations). Network analysis helped to eliminate the last outbreak of Ebola in Africa (2014-2015). Network analysis experts tracked the primary point of the virus appearance (a sick two-year-old boy who had been bitten by a monkey), and then tracked all the channels of further virus distribution.

The internet and its social networks (communities, links, etc) is a huge field for network analysis, since there we can track the processes that we can’t see in everyday life, as we don’t have such access to information. For example, network researchers analyzed Twitter hashtags with Democratic and Republican presidential candidates (more specifically, they analyzed the hashtag-related emotions and network relations between people expressing a certain emotion) and predicted Hillary Clinton and Donald Trump as the ultimate candidates. Network researchers also predicted Trump’s victory, while traditional statistics predicted that Clinton would win. There are too many examples to list, but the future belongs to network analysis.

## What we will teach

Our programme has been created by the example of programmes at Indiana and Illinois, also developed by Prof. Stanley Wasserman, academic supervisor of our master’s programme. In the United States, applied statistics programmes are popular, since the global demand for highly qualified experts in statistical consulting is extremely high and, frankly, there is a lack of them. And the programmes developed by Prof. Wasserman train statistical consultants – professionals who understand theoretical statistics well, but can also work with any data, as well as with clients from various areas of business, as well as public and academic institutions.

Our programme is unique because it brings together various areas of knowledge: data analysis, statistics, mathematics, sociology, management, and many others. We invite professional mathematicians, as well as those proficient in their respective fields (sociology, political science, psychology and all the other ‘-logies’, management, and economics). The term ‘big data’ has gained a lot of popularity recently, and we’ll also study big data, since it doesn’t demand anything special, apart from powerful computers and additional software – the main principles are the same. We’ll provide our students with statistical tools, both network and non-network, which will allow them to work in their key field and turn the raw data into answers to the questions.

The demand for network analysts is so high today, that we can offer our students a choice of projects and internships of all sorts

Our graduates will be able to set the research tasks correctly. I’ve recently met some peers from Deloitte, and they are launching their centre of data analysis. They complain that they manage to employ good mathematicians who can work with software, but are unable to set research questions or solve them. If they enroll graduates in social science, they are great at setting research questions, but don’t have tools for their solution. This is a problem the corporations and leaders of this analytical centre have faced, and they are ready to enroll our students for internships right away in September.

There are 12 applied statistics programmes in Europe, including the University of Ljubljana, one of our international partners. This is the only such programme in Russia. The teaching staff consists of the strongest international experts, such as Stanley Wasserman himself, Vladimir Batagelj (professor of discrete and computational mathematics and creator of Pajek – software for working with big data in networks), Anuška Ferligoj (who launched a similar master’s programme at the University of Ljubljana), Doug Steinley, who uses mathematical methods for analysis in psychology (University of Missouri), as well as some outstanding representatives of the Russian Academy of Sciences, such as Alexander Chkhartishvili and Vladimir Ulyanov, who have been involved in applied statistics and analysis for many years.

## Learning process and individual tracks

The master’s programme has been created on the basis of an international research laboratory, which means that we really have an opportunity to offer individual tracks for our students.

When we create the tracks, we are guided by the student’s goals. We test students at the entrance: if someone is falling behind in mathematics, we improve it by means of workshops taught at other programmes, individual classes, and Coursera courses, until the student achieves a level that allows them to understand their tasks well.

Students don’t have to learn to program, since we can use ready-made statistical packages. Nevertheless, we offer the relevant knowledge for those interested.

## Certification by partners

We’ve signed a partnershop agreement with SAS, a global leader in developing software and applications for Business Intelligence, Data Quality and Business Analytics. Our programme offers a course by SAS as an additional course, and those who pass it will get an official SAS certificate. This certification received directly at the company is very expensive, but we offer it as part of our programme.

## Where to work

We already cooperate with several companies which are in need of our graduates. For example, the Russian Government Analytical Centre, which lacks professionals in this area. They are in an urgent need of people who understand how to set a research problem, how to collect data, to analyze it, and then present it in a useful way. The analytical centre is already attracting our research interns, third-year undergraduate students, to its applied projects. I’ve already mentioned Deloitte, and we’ll continue developing partnerships with other companies to get our students employed even before they graduate. The Russian market has a deficit of professionals able to set questions for problems, to collect the right data, analyze it, and present as reports. By the way, academic careers are also open to our students. One of the options for the programme application is to use it as preparation for PhD and doctoral studies.

The students can work on research projects at our laboratory in various fields. If someone is interested in politics, we have projects that study the mutual evolution of students’ political attitudes and social relations, the impact of political protest on social change, or the building of a certain politician’s image in online communities and media. There is a team that studies networking between rock musicians, and particularly, they study the conditions that influence rock musicians’ success in their fight to be recognized. We also have an interesting large-scale project entitled ‘Sociological research in Russia: structure of the academic community’. It aims to assess the contemporary state of sociological research in the Russian academic community by means of analyzing the publications of social problems in Russian periodicals. We have several other projects, but it is already clear that the topics are very varied.

We accept all certificates of English proficiency. Moreover, internships taken abroad in English are sufficient proof of the necessary level

Following the example of Indiana University, we are opening our own Centre of Statistical Consulting based on our programme, where our students and graduates will be able to work.

Our Centre’s first project is a study entitled ‘Internal staff communications in an organization, and their impact on interaction and team work’ supervised by Ivan Kuznetsov. The research is a new tool for organizational management aimed at monitoring the real internal communication channels, which, together with studying the organizational environment, can increase the management’s awareness on the character of work being done and the quality of staff interaction, all od which helps to improve efficiency. A specific feature of this project is that the laboratory gets paid for its consulting work, and these opportunities will be available for our students.

But, I’ll say it again: the demand for network analysts is so high today, that we can offer our students a choice of projects and internships of all sorts.

## Portfolio and entrance exams

We enroll students via a portfolio competition. Since the students are very different, we offer different options. Someone can pass a set of standardized tests, like GMAT, since it assesses students’ knowledge from various perspectives. Someone can submit certificates of the necessary qualifications. A motivational letter is an essential component, since our programme is for those who have a very clear understanding of why they need to know data analysis and what results they want to get in the process of studies. Portfolios help us understand the level of the applicants, and further determine the relevant individual tracks for them.

Most of the programme teachers are network analysis experts from leading international universities. Unfortunately, Russian social science has been falling behind its Western counterparts in terms of quantitative methods teaching for a long time, so there is a shortage of Russian-language publications on network analysis. We are also planning to attract international students to make studies diversified, and that’s why our programme is English-taught. We accept all certificates of English proficiency. Moreover, internships taken abroad in English are sufficient proof of the necessary level.

Over the last two years, we’ve seen not only a growing demand for data analysts, but also a growing understanding among universities that programmes educating such experts are highly needed. HSE has already launched the Data Culture project, which provides in-depth teaching of statistics at an undergraduate level. The number of programmes in various areas of statistics is also growing. They still offer limited training for professionals in certain areas or demand a very serious knowledge of mathematics. However, statistics should be available to anyone, and this is an advantage of our programme.