Learning to See: Convolutional Neural Networks for the Analysis of Social Science Data

2021 ◽  
pp. 1-19
Author(s):  
Michelle Torres ◽  
Francisco Cantú
Keyword(s):  
Social Sciences ◽  
Neural Networks ◽  
Social Science ◽  
Convolutional Neural Networks ◽  
Visual Information ◽  
Science Data ◽  
Social Science Data

Abstract We provide an introduction of the functioning, implementation, and challenges of convolutional neural networks (CNNs) to classify visual information in social sciences. This tool can help scholars to make more efficient the tedious task of classifying images and extracting information from them. We illustrate the implementation and impact of this methodology by coding handwritten information from vote tallies. Our paper not only demonstrates the contributions of CNNs to both scholars and policy practitioners, but also presents the practical challenges and limitations of the method, providing advice on how to deal with these issues.

scholarly journals Studying the History of Social Science Data Archives as Knowledge Infrastructure

2016 ◽  
Vol 29 (2) ◽  
pp. 62-73
Author(s):  
Kalpana Shankar ◽  
Kristin R. Eschenfelder ◽  
Greg Downey
Keyword(s):  
Social Sciences ◽  
Social Science ◽  
International Development ◽  
Science Data ◽  
Data Intensive ◽  
Knowledge Infrastructure ◽  
Data Archives ◽  
Entry Points ◽  
Institutional Sustainability ◽  
Social Science Data

We map out a new arena of analysis for knowledge and cyberinfrastructure scholars: Social Science Data Archives (SSDA). SSDA have influenced the international development of the social sciences, research methods, and data standards in the latter half of the twentieth century. They provide entry points to understand how fields organise themselves to be ‘data intensive’. Longitudinal studies of SSDA can increase our understanding of the sustainability of knowledge infrastructure more generally. We argue for special attention to the following themes: the co-shaping of data use and users, the materiality of shifting revenue sources, field level relationships as an important component of infrastructure, and the implications of centralisation and federation of institutions and resources. We briefly describe our ongoing study of primarily quantitative social science data archives. We conclude by discussing how cross-institutional and longitudinal analyses can contribute to the scholarship of knowledge infrastructure.Keywords: social sciences; data archives; institutional sustainability

The Haitun dichotomy and the relevance of Bradford's law

1984 ◽  
Vol 8 (1) ◽  
pp. 19-24 ◽  
Author(s):  
B.C. Brookes
Keyword(s):  
Social Sciences ◽  
Statistical Theory ◽  
Social Science ◽  
Critical Review ◽  
Science Data ◽  
Bradford’S Law ◽  
The Social ◽  
Social Science Data

In a critical review of all the empirical laws of bibliometrics and scientometrics, the Russian statistician S.D. Haitun has shown that the application of modern statistical theory to social science data is 'inadmissible', i.e. it 'does not work'. Haitun thus points to the need to develop a wholly new statistical theory for the social sciences in general and for informetrics in particular. This paper discusses the implications of Haitun's work and explains why the older Bradford law still has an important role to play in the development of a new theory.

scholarly journals Guidelines on the Provision and Handling of Research Data in Sociology

2020 ◽  
Author(s):  
Academy of Sociology
Keyword(s):  
Social Sciences ◽  
Social Science ◽  
Research Data ◽  
Professional Association ◽  
Science Data ◽  
The Social ◽  
Social Science Data

With these guidelines the Academy of Sociology (a German professional association) gives recommendations on how social science data could be made open. The aim is to make the Social Sciences more open.

Social Science Data and the Courts

1976 ◽  
Vol 5 (5) ◽  
pp. 11-13
Author(s):  
PATRICIA E. STIVERS
Keyword(s):  
Social Science ◽  
Science Data ◽  
Social Science Data

scholarly journals Uncomfortable Decisions

2022 ◽  
Author(s):  
Paul Bloom ◽  
Laurie Paul
Keyword(s):  
Decision Making ◽  
Social Science ◽  
Career Change ◽  
Social Science Research ◽  
Science Research ◽  
Science Data ◽  
Decision Making Processes ◽  
Social Science Data ◽  
Personal Decision ◽  
Algorithmic Process

Some decision-making processes are uncomfortable. Many of us do not like to make significant decisions, such as whether to have a child, solely based on social science research. We do not like to choose randomly, even in cases where flipping a coin is plainly the wisest choice. We are often reluctant to defer to another person, even if we believe that the other person is wiser, and have similar reservations about appealing to powerful algorithms. And, while we are comfortable with considering and weighing different options, there is something strange about deciding solely on a purely algorithmic process, even one that takes place in our own heads.What is the source of our discomfort? We do not present a decisive theory here—and, indeed, the authors have clashing views over some of these issues—but we lay out the arguments for two (consistent) explanations. The first is that such impersonal decision-making processes are felt to be a threat to our autonomy. In all of the examples above, it is not you who is making the decision, it is someone or something else. This is to be contrasted with personal decision-making, where, to put it colloquially, you “own” your decision, though of course you may be informed by social science data, recommendations of others, and so on. A second possibility is that such impersonal decision-making processes are not seen as authentic, where authentic decision making is one in which you intentionally and knowledgably choose an option in a way that is “true to yourself.” Such decision making can be particularly important in contexts where one is making a life-changing decision of great import, such as the choice to emigrate, start a family, or embark on a major career change.

Imperatives for Theories in Social Science: The North-South Divide Notwithstanding

2020 ◽  
Vol 19 (3) ◽  
pp. 195-217
Author(s):  
Aaron Ola Ogundiwin, ◽  
Joel N. Nwachukwu
Keyword(s):  
Social Science ◽  
Conceptual Analysis ◽  
Social Research ◽  
Science Data ◽  
Social Scientists ◽  
The North ◽  
Common Mistake ◽  
History Of ◽  
Research Questions ◽  
Social Science Data

Abstract The paper underscores the place of theories in organizing social science data and experience. It holds that theories are indispensable to social research (The North-South divide notwithstanding), in view of the fact that the framework of knowledge and experience within which theories are established make a meaningful explanation of the world phenomenon reasonably possible. It delineates political philosophy and history of ideas from theory and thus, takes care of common mistake social scientists make differentiating between them. Furthermore, the paper on one hand, takes on the scientific requisites of theory such as assumption, concepts (and their functions), hypothesis (and its characteristics typology), law, models, paradigm and provides lucid conceptual analysis of each with a view to showing their relatedness to theory but not as synonyms to it. On the other hand, we singled out dependency theory in its emanation from knowledge and experience of underdevelopment of Third World countries, as the first and perhaps most relevant theoretic explanation of Africa’s underdevelopment. The paper posits that a good theory that will serve as a rudder for formulation of research questions, problem statement, as well as sustain the data analysis, and findings must parade some, if not all of the following qualities: precision and testability, empirical validity, parsimony, stimulation, and practicability.

Sign in / Sign up

Export Citation Format

Share Document