Reaching a Consensus on the Definition of Genetic Literacy that Is Required from a Twenty-First-Century Citizen

To determine what knowledge of genetics is needed for decision-making on genetic-related issues, a consensus-reaching approach was used. An international group of 57 experts, involved in teaching, studying, or developing genetic education and communication or working with genetic applications in medicine, agriculture, or forensics, answered the questions: “What knowledge of genetics is relevant to those individuals not professionally involved in science?” and “Why is this knowledge relevant?” The answers were classified in different knowledge components following the PISA 2015 science framework. During a workshop with the participants, the results were discussed and applied to seven cases in which genetic knowledge is relevant for decision-making. The analysis of these discussions resulted in a revised framework consisting of nine conceptual knowledge components, three sociocultural components, and four epistemic components. The framework can be used in curricular decisions; its open character allows for including new technologies and applications and facilitates comparisons of different cases.


Introduction
Genetics has evolved from a unique subdiscipline of biology into an integral part of most biological research, covering multiple levels of biological organization. Results from studies in genetics influence societal practices, such as disease diagnosis and treatment, drug development, experts via e-mail: BWhat knowledge of genetics is relevant to those individuals not professionally involved in science?^and BWhy is this knowledge relevant?^The experts (n = 57) included science education researchers (n = 26), developers of educational materials (n = 18), teachers and teacher educators (n = 8), science communicators (n = 6), scientists-including medical geneticists, community geneticists, and genetic counseling experts-forensic science experts and agricultural experts (n = 8), and educational policy-makers (n = 3). The science education researchers and developers were all involved in research and development of genetic education. The total number of experts exceeds 57 because some of the experts had more than a single expertise. To obtain a representative group of participants, we started with a group of researchers from eight different countries who had published on genetic education and asked them to recommend other researchers, developers, teacher educators, and genetic specialists. Participants came mainly from Europe and the USA, along with three participants from Australia and one from Taiwan (Appendix 1).
All participating experts' answers to the first question were initially classified into three types of knowledge, following the PISA 2015 Science Framework (OECD 2016): conceptual knowledge, epistemic knowledge, and procedural knowledge.
The components previously suggested for a learning progression in modern genetics by Duncan et al. (2009) proved useful as a framework to analyze the answers classified under conceptual knowledge. The Duncan et al. (2009) framework is based on the suggestion of Stewart et al. (2005) that knowledge of three integrated conceptual models is necessary to truly understand genetic phenomena: (i) the genetic model, which deals with the patterns of inheritance observed when organisms reproduce sexually and the probabilities with which different patterns are likely to occur; (ii) the meiotic model, which relates to the cellular processes underlying gene recombination, sorting, and transfer from one generation to the next; and (iii) the molecular model, which deals with the mechanisms that link genes to their biological outcomes. The framework of Duncan et al. (2009) expanded these three models to include the environment as the context in which genetic processes take place and emphasized the role of proteins in the development of traits. We found the eight components listed in Duncan et al.'s (2009) suggested framework useful for classifying the answers of the conceptual knowledge type. Responses that could not be classified into any of the components of the framework (see Table 1) were listed separately. The answers that were classified under procedural knowledge and epistemic knowledge were each analyzed bottom-up, allowing the categories to emerge from the data. This process resulted in three tables of knowledge components which were used in the second phase of the study (see Tables 1, 2, and 3).

Phase II (March 2013)
The second phase of the study differed from a traditional Delphi approach in the sense that the participants were no longer anonymous, but met in a workshop to discuss the results of phase I. One week before this workshop, the results of the first phase of the study were e-mailed to all participants. The workshop was attended by 46 out of the 57 experts who responded to the initial survey, plus an additional 8 experts from the local hosting university, who did not participate in the first phase of the study. During the workshop, the experts worked in groups, and each group discussed a different case study in which genetic knowledge is relevant for decision-making. Participants could choose which case they wanted to discuss. Each group was asked to indicate the specific knowledge type-among the three types of knowledge mentioned above-needed to make an informed decision in their specific case. The cases were selected to cover a broad spectrum of those that citizens might encounter in their everyday lives. Cases were selected from the main social practices in which genetic applications play a role in today's society: medical diagnosis, food production, sports, and forensics. Another criterion for selecting cases was to cover different gene-trait relationships: monogenic versus polygenic traits, traits with little versus big environmental influence, germline mutations versus somatic mutations, and functional genes versus short tandem repeats. A third criterion for selecting cases was to cover different sources of genetic information, such as the media, Table 1 The conceptual knowledge components of genetic literacy after the first and second phases of the study. Main changes to Duncan et al.'s (2009)

Phase III (June 2014)
The comments of the participants, both on the proposed categorization of knowledge components and on the specific knowledge per case, formed the input for the third phase. In this phase, the first two authors revised the categorization into four types of knowledge: Expert answers from the first phase were reclassified according to the new categories. Based on the group reports in the second phase, within each knowledge type, a further analysis was performed to identify which specific knowledge was needed in each case. This process resulted in three tables with matrices that describe the content of a specific knowledge component for every case. Each row in Tables 4, 5, and 6 addresses a different part of the specific knowledge component, and each column discusses a different case, providing the specific genetic knowledge needed for decision-making for that knowledge component in each row.

Phase IV (November 2015)
The matrices were sent back to the participants for the final round of comments, producing the final elaboration of genetic literacy required by a twenty-first-century citizen. In this phase,   participants could indicate whether they agreed with the formulation of genetic literacy. Apart from some comments on the categorization, all agreed with the formulation.

Categories of Conceptual Knowledge
After the first phase of the research, three tables were formed, based on the distinction among conceptual, procedural, and epistemic knowledge (OECD 2016). Expert answers on conceptual knowledge could, for the most part, be classified according to the framework of Duncan et al. (2009). Three additional categories (I, J, and K in Table 1) were formed to include answers that did not fit the framework. Based on the discussions in the second phase and comments by e-mail, the following changes were made in the conceptual knowledge components (Table 1,

right column):
& Several knowledge items were redefined to comply with current genetics, and basic concepts such as DNA and gene were included in the description. & A new category on variation was added (Table 1, category 8), as many answers fell into that category, which was not present as such in the original framework. & The new category of polygenic inheritance (Table 1, category K) was combined with the existing category of environmental influence (Table 1, category 9): BMultiple genes and multiple environmental factors interact in the development of most traits.^This category was considered by several participants as the core message of genetic literacy. & Evolution and natural selection (Table 1, category J) were omitted from the conceptual knowledge in order to focus the definition of genetic literacy. It proved difficult to decide what to include in genetic literacy, as genetics is linked to other biological phenomena such as evolution and development.

Categories of Epistemic, Sociocultural, and Procedural Knowledge
The analysis of the answers originally classified as epistemic knowledge resulted into a division in two separate types of knowledge. One type deals with how genetic knowledge is used in societal applications and which issues are generated by these applications. This type of knowledge was given the title of sociocultural knowledge ( Table 2).
The second type consists of knowledge about how genetic information should be interpreted, both in professional and public communication. This type retained the name epistemic knowledge.
From the answers that were classified as procedural knowledge, two main components emerged during the first phase of the study: (i) science-communication skills and (ii) argumentation and decision-making skills. During the later phases, two kinds of comments gave reason to change this initial classification. One comment was that argumentation should be a part of epistemic knowledge and therefore be transferred from procedural to epistemic knowledge (Table 3). The second comment was that the other elements of procedural knowledge appeared to be general skills, not specific for genetic-related situations. Therefore, specific procedural knowledge was not included as part of genetic literacy.

Outcomes of the Discussion on the Case Studies
Tables 4, 5, and 6 describe the different types of genetic knowledge needed for decisionmaking in the seven case studies (see descriptions of each case study in Appendix 2). Where no specific genetic knowledge seems to be needed for a case, the cells remain empty. The three knowledge categories can be described as follows: & Conceptual knowledge: knowledge of genetic concepts. & Sociocultural knowledge: knowledge of how applications of genetic technologies are used in societal activities and in what ways they influence human lives. & Epistemic knowledge: knowledge of the meaning of genetic information. This concerns the knowledge needed to interpret genetic information from different sources and how to use these in argument and decision-making. This knowledge includes Nature of Science aspects such as the certainty and uncertainty of genetic information and how genetic concepts have evolved.
In Tables 4, 5, and 6, the different types of knowledge are described in relation to situations in which this knowledge has to be applied in personal or societal decision-making on geneticrelated issues.

Conclusions
Based on the results, according to the participants of this study, the following answer can be given to the research question Bwhich genetic knowledge is needed for decision-making on genetic-related issues?.^Genetic literacy can be described by the three aforementioned types of knowledge, with the knowledge components presented in Tables 1, 2, and 3. Starting with sociocultural knowledge, the results indicated that genetic literacy includes awareness of the fact that new genetic technologies are frequently used in practices that can affect everyone, such as medical diagnosis and therapy, food production, and forensics, and that most of these applications have benefits as well as risks or disadvantages, for individuals and for groups, which makes careful decision-making both complex and necessary. The cases illustrated that decision-making on these issues requires knowledge of certain (more or less case-specific) genetic concepts, together with an understanding of the meaning of genetic information and the misleading way in which this information sometimes reaches the public. For example, a focus on the Mendelian inheritance model, whereby one allele leads to a specific trait without environmental influences, can contribute to a misleading image of most gene-trait relationships. This does not mean that Mendelian inheritance has become obsolete (it is needed to understand certain types of genetic-related issues), but that Mendelian inheritance should be regarded as the exception, not the rule. Many participants indicated that conceptual knowledge component 9 (multiple genes and multiple environmental factors interact in the development of most traits) summarizes the most important message of this study. A similar formulation can be found in the K-12 Framework of the NRC in the part on the core idea LS3: Heredity: Inheritance and Variation of Traits: BComplex relationships between genes and interactions of genes with the environment determine how an organism will develop and function^ (NRC 2012, p. 159). The cases demonstrated many examples of obstacles to genetic literacy brought about by transferring the Mendelian inheritance model to all gene-trait relationships and by misleading images created by the media. Table 7 summarizes some of these images, which biology education should avoid or correct using the genetic literacy elements described in this article.

Discussion
The distinction in this framework between conceptual, sociocultural, and epistemic knowledge fits well with the experts' answers, as well as with their discussion of the seven cases. This distinction is comparable with the components of genetic literacy for patients and providers described by McInerney (2002): knowledge of genetics, variation, and disease (comparable with conceptual knowledge); ethical, legal, and social issues (comparable with sociocultural knowledge); and knowledge on the Nature of Science (comparable with epistemic knowledge). There are also many parallels between the framework developed in this study and what is written about genetics in the Next Generation Science Standards (NGSS). The standards are based on the framework K-12 in which disciplinary core ideas, practices, and crosscutting concepts are formulated (NRC 2012). In the NGSS, these elements are integrated in required performances such as the following (aimed for middle school students who learn genetics): BGather and synthesize information about technologies that have changed the way humans influence the inheritance of desired traits in organisms^ (NGSS Lead States 2013). This performance includes knowledge about traits, about genetic applications, and about obtaining and evaluating information. As in the genetic literacy framework presented here, genetic education according to the NGSS is not limited to learning core concepts, but is linked to learning about engineering, technology, and various applications of science (as in the sociocultural knowledge components in our framework) and to the nature of science (as in the epistemic knowledge components in our framework). The importance of the genetic literacy framework presented here lies in the case Most predictions based on gene variants will remain uncertain due to the complex gene-trait relationships studies which indicate which specific genetic knowledge is needed and used in a special situation and indicate which choices can be made, with the risk and benefits of each decision. This article describes a consensus reaching process. Consensus is a relative notion, because approving this final text does not mean that there is a consensus on all matters. The main points of this discussion concern the limits and purpose of genetic literacy.

The Limits of Genetic Literacy
The participants' opinions differed on whether procedural knowledge should be included. The procedural knowledge appeared, for the most part, to be of a general kind, and after some debate, we decided not to include this type of knowledge in our definition of genetic literacy. Without communication skills, however, a twenty-first-century citizen would probably be unable to put genetic knowledge to use in a societal context.
Another topic about which participants disagreed was whether natural selection should be a part of genetic literacy. We decided not to include evolution and natural selection, as many other core biological ideas are also linked to genetic concepts, and including them would result in expansion of the description of genetic literacy to the whole of biology. This distinction is in line with the NRC framework for K-12 science education (NRC 2012), which lists four core ideas that represent basic fields of investigation in the life sciences, among which heredity and evolution appear as two separate core ideas. Nevertheless, separating genetic literacy from evolution literacy seemed odd to some participants of this study, since for decision-making on the use of antibiotics, knowledge on mutation and variation (as included in this article) cannot be separated from knowledge about selection.

The Purpose of Genetic Literacy
Another point of discussion among the participants concerned the purpose of genetic literacy. The methodology that was chosen framed the results in several ways. The question of defining the genetic knowledge that a citizen needs already implied a focus on the knowledge needed in societal contexts and for decision-making. Therefore, using the case studies to discuss genetic literacy implied a Bfilter^through which genetic knowledge was selected. Some participants wanted to go even further to create an antidote to racist policies, meaning that genetic literacy should also inform certain societal choices. Others thought that the goal of informed decisionmaking is too narrow and that education also has the broader task of introducing students to the cultural heritage of which genetic science is a part. Thus, the procedure of selecting genetic knowledge through the filter of societal cases could exclude genetic knowledge that is needed for goals other than informed decision-making. There were also doubts as to whether the genetic literacy goal of keeping students up to date is feasible, given the speed of the developments in science; it was suggested that it would be more realistic to concentrate on generative and powerful knowledge of core ideas and a critical stance toward the presentation of genetic information.

Limitations of the Study
In a traditional Delphi approach, the participants remain anonymous and interact in writing (Osborne et al. 2003;Bolte 2008). In our study, participants also met each other in person and participated in group discussions during the workshop that took place in phase II of the study, so as to promote a deeper understanding of what genetic literacy should entail. The group members were not asked to reach a consensus, but to generate informed and reviewed opinions. Comments before and after the workshop were analyzed and presented anonymously. Almost all of the participants were from countries with comparable social and scientific characteristics. There were no representatives of African or South American countries, where the problems related to genetics can be different. Input from these countries might have influenced the outcomes.

Implications of the Study
Having produced these tables, the next step was to determine what they say about the genetic literacy required by a twenty-first-century citizen. The framework describes both familiar and new elements of genetic education. Familiar elements include knowledge that genetic information is inherited from both parents with basic patterns of Mendelian inheritance (needed, for example, to understand concepts such as carrier and risk in genetic counseling). Elements that receive much more attention in this framework than in traditional genetic education are the complex relationships between genes, environment, and traits, including the concept of heritability (needed, for example, to detect incorrect wording in media reports such as Bgene for^ (Dupré 2012). The framework does not elaborate specifically on mechanisms such as transcription and meiosis. Namely, knowing that Bdifferent genes are switched on and off in different cells^and that Beach parent contributes a set of genes, leading to a double set in the offspring^(components 4 and 5, Table 1), is sufficient for a twenty-first-century citizen confronted with genetic-related issues, while knowledge about the specific mechanisms governing these facts is not. Thus, it is proposed that the meiotic model suggested by Stewart et al. (2005), which emphasized the cellular processes underlying gene recombination, sorting, and transfer from one generation to the next, should be acquired during advanced courses in genetics, but is not required for every citizen.
The genetic literacy described in this study can have several functions in education, both at the curriculum level and in the preparation of biology teachers. As a basis for curriculum innovation, genetic literacy is described here as a conceptual framework, not a definition. The framework can consider several cases, making comparisons between cases easy and giving a more complete image of the role of genetics in society. The structure of the framework makes it easy to add new elements, as well as to explore new cases. For example, a new application that occurs in the media can be explored by creating a new case column, specifying the relevant knowledge using the knowledge components of the tables, and comparing the results with those of already described cases. The consequences of a new technology such as CRISPR-Cas (Burgess 2013) can be studied by checking how this would influence benefits and risks in Table 5. The produced framework for genetic literacy is therefore an open one that allows for curricular discussion, revision, and extension.
With regards to the preparation of biology teachers, the framework presented here can also help teachers and teacher trainers to analyze genetic issues that are relevant for their students by using the extensive overview in this framework of genetic technologies, their applications, and the issues that they generate. This may enable the development of educational activities within context-based education and the guidance of classroom discussions on these issues. Communication and argumentation skills are essential for citizenship education, and biology teacher education should prepare teachers for discussions on moral issues generated by genetic technologies (Van der Zande et al. 2009). Finally, the framework may also help to improve biology teacher education by pointing out common misconceptions.
Although this is all very challenging for science teachers, many examples of context-based genetic education already exist. In using contexts such as the seven described cases in biology education, the different types of knowledge can be distinguished but should not be treated separately. Discussing a new application of genetic technology can generate a need for new conceptual knowledge as well as stimulate discussion regarding its advantages and disadvantages.
Special attention should be paid to developing skills for taking uncertainty into account. The combination of more genetic information reaching us with less certain predictions makes the task of preparing students for citizenship difficult. McInerney (2002) mentioned the Bmanagement of uncertainty^as a fundamental part of genetic literacy. Teachers might have to elaborate on different sources of uncertainty, for example, uncertainty stemming from the complexity of the gene-environment-trait interplay, uncertainty caused by the meiotic process leading to different genotypes, and uncertainty caused by a technology that has false-positive and false-negative outcomes. The risk of overstating this uncertainty is that it might lead to rejection of all scientific information.
Another topic that needs careful treatment involves the concepts of race and ethnicity (Donovan 2016). Obvious visible differences such as skin color and results of sport events such as marathon running might easily lead students to the idea that important genetic differences exist between groups of people. The issue is complicated by the difference between sociological and biological meanings of race and the political issues related to them. This explains why this topic is often avoided in education. However, careful training in argumentation can prepare students for discussions on this topic (Puig and Jiménez-Aleixandre 2011).
We feel that the reported procedure for defining genetic literacy, i.e., defining the conceptual core and analyzing cases in which specific knowledge is needed, could also be used for other science topics which are linked to important societal implications.
During a social meeting, you hear someone claiming Bof course races differ in IQ. IQ has proven to be largely genetically determined and anyone can see that races differ. Just look at the Olympics, and which countries always win the marathon.^How do you react?
http://www.runnersworld.com/peak-performance/why-are-kenyan-distance-runners-so-fast 7. Genetic counseling for spinal muscular atrophy A healthy couple who want a child undergo a reproductive carrier screen. The test shows that both are carriers for spinal muscular atrophy (SMA), a disease in which muscles progressively deteriorate, often in early infan84cy. The couple is deliberating about whether they should have a child, knowing their genetic condition. http://ghr.nlm.nih.gov/condition/spinal-muscular-atrophy