Action

In Ferligoj’s (2015) paper titled “How to improve statistical literacy”, the actors that can contribute to better statistical literacy include educational institutions, statistical offices, statistical associations, and the media. Furthermore, recommendations were made on action to be taken which included: advising different segments of the population separately about the proper use and interpretation of statistical data as well as presenting typical abuses or misunderstandings of statistical concepts and data.

This section of the review encompasses action that has been taken, in the UK and internationally to improve statistical literacy and statistical communication.

5.1 United Kingdom

The Royal Statistical Society (RSS) launched the getstats campaign in 2010 which had a 10-year goal of improving statistical literacy in the UK. It aimed to do this by providing training and running workshops to statistics communicators, such as journalists, MPs, and their research staff. There would be wider benefits from RSS publishing the outcomes and learnings from this campaign.

In 2021, a user engagement strategy for “ensuring official statistics meet society’s needs” was announced by the Government Statistical Service for the producers of official statistics within the UK. The three main goals of this strategy are: building collaboration across public sector bodies, developing the capabilities of statistics producers to understand the audience, and strengthening user engagement culture. Furthermore, the phases of the strategy include showcasing and building on existing good work as well as establishing baseline activity, establishing the USER hub to lead the way on user engagement support packages, and lastly, depending on statistics producers to use the USER hub to improve engagement. This strategy is still in its early stages and is expected to be implemented over four years (2021-2025).

The Winton Centre for Risk and Evidence Communication, the source of a number of the studies mentioned in this report, have established multiple online tools available on their website (https://wintoncentre.maths.cam.ac.uk/) that incorporate some of the learnings of this research. These include:

• Predict websites: Tools to allow clinicians to compare different treatment options for breast cancer (https://breast.predict.nhs.uk/) and prostate cancer (https://prostate.predict.nhs.uk/) both individually, and with patients during their appointments. They have incorporated uncertainty in their estimates using numerical ranges, which is in line with their previous research highlighting that this approach is not harmful to perceived trust in the data (Van Der Bles et al., 2020). Furthermore, work is ongoing to investigate the effect of presenting the statistics in different formats on people’s understanding of the data.
• RealRisk (wintoncentre.uk): An online tool to assist press officers and journalists in communicating risk statistics in health and social areas. Specifically, it can be used to convert risk originally conveyed as relative risk, odds ratios, or hazard ratios into absolute risks. As mentioned in the “format” section of this report, absolute risks are often better understood by the general public (Oxman et al., 2020).
• A tool communicating risk from different COVID-19 transmission routes. Statistical information is presented in different ways for different users by integrating information boxes into their infographics that communicate COVID-19 transmission risk so that more interested/advanced users can find out more about the certainty of evidence. Colour bars were also used rather than numerical estimates of risk to limit the overinterpretation of precision in these statistics (Rutter et al., 2021).

The Plain Numbers Project (https://plainnumbers.org.uk/) was established in 2021 to support consumers who struggle with numeracy. Plain Numbers has worked with the Bank of England as well as firms such as ClearScore and Direct Line to make changes to their communication materials to improve consumer comprehension. The approach is based on three key principles, namely “numbers themselves” which refers to aligning the use of numbers with how humans think which is in stories. The second principle is “numbers in context” which recommends assuming little contextual knowledge from the consumer as well as removing jargon. The third principle is “how we think” and this is related to considering how the information to be presented should be processed by the consumer. In initial studies the Plain Numbers approach was successful and across firms, the approach doubled the number of people who understood customer information (The Plain Numbers Project, 2021).

5.2 International

As part of the work programme of the Conference of European Statisticians, a Steering Group on Statistical Dissemination and Communication, supported by the United Nations Economic Commission for Europe (UNECE), produced four guides to help those communicating statistics to non-statisticians. The first guide “A guide to writing stories about numbers” (2009) was prepared by a group of international experts. The audience of the guide are managers, statisticians, and media relations officers. The guide begins by emphasising the importance of providing a statistical story. The guide provides recommendations for writing with a focus on increasing engagement by being “newsy”. Overall, the guidelines are quite consistent with others in this area. Most relevant is the fourth guide, titled “How to improve statistical literacy: A guide for statistical organizations” which covers statistical literacy not only in the general public but also in decision-makers and within statistical organisations (UNECE, 2014). This guide includes a summary of several international groups working in the area of statistical literacy.

This summary included the International Statistical Literacy Project (ISLP) (https://iase-web.org/islp/) which was initiated by the International Association for Statistical Education (IASE), the education section of the International Statistical Institute (ISI) to promote statistical literacy worldwide. This project aims to increase the widespread statistical literacy of citizens. It has previously focussed on young people in educational settings which has involved activities such as a statistical literacy competition; however, ISLP has shifted more recently to also consider statistical literacy in adults.

The summary also mentions Statlit.org a website for statistical literacy-related articles, books, and activities. The website is led by Milo Schield, who is a former President of the National Numeracy Network and Director of the W.M. Keck Statistical Literacy Project, based at Augsburg College in the USA. The first goal of both the website and the project is to present statistical literacy as an interdisciplinary activity, overlapping with quantitative reasoning, quantitative literacy, numeracy, and statistical reasoning. The second goal is to present statistical literacy as the study of statisticians in everyday arguments.

Further action identified through the current review includes Eurostat, the statistical office of the European Union (EU). Eurostat’s efforts to increase understanding of statistics include training programmes and their guide to European statistics “Statistics Explained” which aims to present statistical topics in a way that is easily understood.

In a less conventional approach, the Italian National Institute of Statistics (Istat) developed a puzzle magazine named “Statistica enigmistica” which uses crosswords and similar puzzles to introduce statistical concepts to non-specialists (Da Valle & Osti, 2016). There seems to be no available evidence on the efficacy of this approach or indication that it was widely distributed.

ProCivicStat is a strategic partnership across six universities and five countries as part of the Erasmus+ Strategic Partnerships. The goal of this project is to promote the inclusion of civil statistics within statistical education. Outcomes of this product include new methods for statistics instruction to help young people understand evidence and statistics about social phenomena. Other outcomes include conceptualising the knowledge requirements for engaging with civic statistics and this research was discussed in the “Definitions” section of this report (ProCivicStat Partners, 2018).

5.3 Summary

Overall, this section of the review identified multiple initiatives, of varied scale, in the statistical literacy space. Some of the initiatives had little information available online about their work in statistical literacy, whether they are currently active and their efficacy. From the information available, there appears to be a lot of variety in these initiatives regarding location, target audience and overarching aims. Many of the programmes may no longer be active and it may be beneficial to gain further information about these in particular. For example, insight into why these programmes ended, and aspects of them that were successful or unsuccessful could increase the efficacy of future action. Furthermore, no work was detected during the review process that aimed to connect these different initiatives. Therefore, this may be a fruitful future endeavour to share learnings across groups.