https://academic.oup.com/rev/article/27/4/298/3978693

I know some of the people who wrote this and would trust that they read everything.

It's really difficult to establish indicator for measuring impact of the social sicences. it's against our perception of social science. It's more a debate.

https://www.hceres.fr/sites/default/files/media/downloads/Seminaire%20SHS%20Synthese_mai2019_1.pdf

Social sciences is a very general concept. The impact can only be looked from specific areas in social sciences

From the analysis of conference materials (https://www.ssh-impact.eu/ ; https://www.ihs.ac.at/events/detail/event/termin/sozial-und-geisteswissenschaften-in-horizon-europe/ ; https://aesisnet.com/events/ssh19/), we can conclude that a social request is being formed for the social and humanitarian sciences to serve the political tasks of social design.

USSR in 1930-1980 already experienced this when the whole complex of social and human sciences was subordinated to the tasks of social construction.

This led to the formation of enduring myths in the field of official science and the departure of outstanding minds underground. Real science existed secretly away from official doctrine.

The field of political programs is only one of the spheres of influence of the social and human sciences. It is characterized by the impact on short- and medium-term social processes: 5-15 years. The role of not all sciences is equivalent. The leading place here is occupied by political science, the theory of international relations, economic theory, the theory of communication, applied sociological, cultural and art history (criticism) studies. The poststructural methods of discourse analysis are applicable here (Foucault, Derrida, Bart, Laclo, Muff, Alexander, etc.).

In the medium term (15–25 years), the theory of law, ideological paradigms of historical science, social and cultural anthropology, prognostic sociological and cultural studies (futurology), etc. prevail. Here we should pay attention to a quantitative analysis of the fixed dynamics of social reflection (for example, cinema, medium-term dynamics of political programs, the dynamics of improving the sphere of legal relations, etc.). Such tasks require extensive research, development of a methodology for a specific direction of observations and organization of scientific discussions to discuss the results.

Fundamental humanitarian knowledge (philosophy, philology, theoretical cultural science, history and philosophy of science, etc.), as a rule, lays down the factors of long-term development. Why does no one set the task of tracking the impact on modern society of Plato, Aristotle, Descartes, Bacon, Kant, Nietzsche and others? Modern technologies for collecting and processing big data allow this.

Dear colleagues, thank you for the helpful answers so far!

Dear Anthony Ichuloi Oure : I couldn't agree more; impact assessment should be field-specific and should anticipate epistemic specifics (btw. on both, the research as well as the society side).

Dear Anne-Paule Mettoux-Petchimoutou : Exactly! I think there are some (few) valuable quantifiable indicators (depending on the unit of analysis) AND of course a lot of valuable practices that cannot be measured / counted. All the more I think it is important that we identify these blind spots and offer solutions for making these count. The REF, for instance, tries to do this through case studies. I'm not saying this is ideal but at least a move away from purely quantitative assessment that typically is not beneficiary to scholars in the social sciences and humanities.

Gennady Bakumenko : Thank you! Do you know of any experts / sources listing the limits of measuring impact or the possibilities of new methods?

Thank you, Dr. Benedikt Fecher, for your attention.

I was more interested in the causes of the impact. In other matters ...

K. Popper in his best-seller pointed out the limiting values ​​of historical science for the prospects of social engineering. One can push off from his position and trace which concepts of the historical process are translated by various sciences. This is my idea. But she is on the surface. Without historical representations, there is no social or human sciences. It is possible, sector by sector, to carry out an epistemological analysis of the research mainstream of the social and human sciences. So we will find out whether an open society is being reconstructed or some form of totalitarianism.

The mainstream is quantified in the thematic sample, and the direction of social engineering (open society / totalitarianism) is determined by the characteristics of the vision of the historical process. Comparison of results implies a certain forecast. Here we can talk about medium and long term prospects.

Philippe Keraudren ; Rob Timans, Johan Heilbron ; Chiara Faggiolani, Giovanni Solimine ; Marcus Morgan ; Geoffrey Williams, Ioana Galleron ; Andrea Bonaccorsi ; Kateřina Cidlinská ; Brigida Blasi, Sandra Romagnosi, Andrea Bonaccorsi ; Fernanda Beigel, Gustavo Sorá ; Eugenio Petrovich ; Sándor Soós, Zsófia Vida, András Schubert ; Alice B. M. Vadrot, Aleksandar Rankovic, Renaud Lapeyre, Pierre-Marie Aubert & Yann Laurans; Patricia van Hemert , Peter Nijkamp & Jolanda Verbraak; Ulrike Felt; Ronald J. Pohoryles & Andreas Schadauer and etc.

Chapter Publication Strategies in SSH: Empirical Evidence and Policy...

Chapter More, Less or Better: The Problem of Evaluating Books in SSH Research

https://link.springer.com/chapter/10.1007/978-3-319-73299-2_6

Article Publication patterns in the social sciences and humanities: ...

and etc.

https://academic.oup.com/rev/article/27/4/298/3978693

I know some of the people who wrote this and would trust that they read everything.

I found "The Four Literatures of Social Science" by Diana Hicks a nice introduction to the particularities of SSH: Article The Four Literatures Of Social Science

Also check the latest special issue in Reseach Evaluation https://academic.oup.com/rev/article/29/1/1/5714805

I like the LSE's policy group position (at least it was their position in 2011 - see http://eprints.lse.ac.uk/35758/1/Handbook_PDF_for_the_LSE_impact_blog_April_2011.pdf ) that it's incredibly costly, and potentially divisive, to measure social science research impacts and we should look to 'recordable moments of influence'. This is the model we use in the Centre for Social Impact network in Australia. The presumed causality of research impact ignores the complexity of systems in which researchers are embedded and to which they respond. One researcher's output might be another's input before it leads to an impact, so who do we reward in that scenario?

Dear Benedikt Fecher (and colleagues),

My position as a scholar is that this is a very important topic that speaks to the social responsibility of publicly-funded research.

As such, I would direct you to the contemporary discussion on public science (and the idea of public approaches to disciplines, such as public diplomacy, or public sociology, which has been developed by Michael Burawoy, since his 2004 presidential address to the American Sociological Association).

As I work in the area of Education in Brazil, in our research group, we have identified some "domains of action" in which educational researchers are prone to develop knowledge. These domains, however, are not necessarily recognized in the state's research assessment exercises (which determine funding). Each knowledge area could identify its domains of action and devise indicators from there, considering which are the possible evidences for societal engagement.

I would like to introduce you to the Brazilian experience. As most research here is conducted in postgraduate programs, science evaluation is closely linked to postgraduate education evaluation. This evaluation was recently reformed (as of 2018-2019), and now includes three titles, one of them being "Social Impact". Social impact criteria are defined by knowledge areas. In the area of Education, there are very comprehensive indicators, but arguably the most proficuous point would be to measure societal engagement outcomes linked to research projects (such as partnerships with educational institutions and professionals fostering innovation).

Then, the limits of measurability lie in the positivist paradigm associated with most measurement. The SSH research may have a broader and deeper social impact when it is socialized with social movements than when it is cited by a scholar. But how to produce a positive proof of that, so that it can be "read" by assessment exercises? You cannot count the impact of a class on an organization, but you can count how many times a researcher visited educational settings to discuss results and describe the procedures they adopted. Some consensus should be met among the policy and scientific communities so as to produce a code for registering that. I understand that the political aspects of such a process would prove a bigger challenge than the technical ones.

Of course, ethical principles should be observed. In this matter, I would recommend you to take a look at was is produced under the "Responsible Research and Innovation", supported by Horizon 2020: https://www.rri-tools.eu/about-rri .