Gap assessments, impact assessments and scientific research
ValueMonitor uses written sources as an input (e.g. newspaper articles, policy documents, scientific articles). Distributions of words (so-called ‘topics’) are then applied to examine values addressed in these documents. ValueMonitor can be used to produce policy gap assessments, impact assessments and for scientific research, for example to study how values have changed over time.
ValueMonitor can be used to systematically identify gaps between how values are addressed in different text corpora, for example newspaper articles, scientific and technical articles, and policy documents. An example can be found below.
The figure to the right shows an analysis of 12 values relevant to blockchain technology, in newspaper articles (NEWS), scientific literature on ethics of technology (ETHICS), and in techno-scientific literature (TECH). It suggests some striking differences in what values are addressed in these three realms in relation to blockchain technology. For example, in the news and the ethics literature, democracy is often discussed in relation to blockchain technology, while in the techno-scientific literature there is hardly attention for this value. Conversely, a value like justice and fairness is surprisingly absent in the ethics literature. (This may be explained by the small number of ethics literature on blockchain).
Such outcomes might help to assess the effectiveness of current policies and to identify values or topics that might require more attention. This can be used for policy advice or new regulatory approaches to technology.
The analysis shown is part of the report Ethical and societal challenges of the approaching technological storm written for the Panel for the Future of Science and Technology of the European Parliament.
ValueMonitor can systematically trace in text corpora whether certain values that are central in certain policies or strategies are more often mentioned after the introduction of these polices or strategies. We have not yet done impact assessments of policies with ValueMonitor, but the following example shows how ValueMonitor can be used to trace changes in how a value is conceptualized over time; in this case showing the impact of the information age on how privacy is conceptualized.
We created eight supervised topics that relate to the typology of privacy proposed by Koops et al. ( 2017). Bodily privacy refers to the freedom of movement and touch of human body by others. Spatial privacy is the restriction of control and access to individual space. Similarly, communicational privacy is about transmitted information. Proprietary privacy refers to the use of property to hide aspects of individual lives. Intellectual privacy relates to thoughts, beliefs, and opinions. Decisional privacy relates to the intimacy of decisions in human relationships. Associational privacy refers to freedom of interaction in social networks. Behavioral privacy relates to publicly visible activities.
The figure shows that there is a growing emphasis on communicational privacy, possibly due to the rise of the information age. The increasing popularity of social media could explain the rise in associational privacy in the late 2000s. Overall, there does not seem to be a succession of conceptualizations of privacy that have emerged or disappeared over time. Rather, the conceptualizations seem to be coexistent.
This analysis can be found in: de Wildt, T. E., van de Poel, I. R., & Chappin, E. J. L. (2021). Tracing Long-term Value Change in (Energy) Technologies: Opportunities of Probabilistic Topic Models Using Large Data Sets. Science, Technology, & Human Values. https://doi.org/10.1177/01622439211054439
An example of how ValueMonitor can be used for the scientific study of value change is a study we did on changing values due to the Corona pandemic .
The figure shows the results for a large corpus with new articles in the period 2016–20, including non-COVID news. It very clearly shows the effect of the COVID-19 pandemic on the frequency in which specific values are addressed in news articles. Health & safety increase from below 10 % to above 50% in three months. Hedonism, mental health and economic welfare also show an increase in frequency in early 2020, although the frequency of these values does not deviate from their bandwidth in the period before 2020. The other values show a drop in frequency. For democracy, privacy and socio-economic equality, this is a drop well below the bandwidth of the values in the period 2016–20.
Our further results suggests that this punctuated value change started to cancel out after the first months of the pandemic. We also found some distinct differences between countries, which we could in part explain by cultural differences between these countries. We also made a comparison with value change trends from surveys. Interestingly, we did not find evidence for a dilemmatic choice between “safety and health” versus “economic welfare”, as was sometimes suggested in the popular media.
Readers interested in the other outcomes can read our forthcoming chapter: Van de Poel, I., De Wildt, T. E., & Van Kooten Pássaro, D. (Forthcoming). COVID-19 and changing values. In M. Dennis, J. v. d. Hoven, G. Ishmaev, & S. Umbrello (Eds.), Values for a Post-Pandemic Future: Ethics, Technology, and the ‘New Normal’. Springer.