Genetic determinism.

Deterministic understandings of genetics typically focus on a one-to-one relationship between genes, proteins, functions, and traits, as if particular traits or diseases were generally related to a single gene. Genetic determinism is often implied in common expressions such as “the gene for intelligence”, and ignores the influence of environmental and epigenetic factors [1, 26]. Nevertheless, the expression “gene for” [2] has subtleties in use and interpretation that are often lost from sight. The expression “gene for” can be conceived as a shorthand expression for “a locus in which sequence variation causes a difference in phenotype, all other things being equal” [27, p. 298, citing 28]. However, there are many cases in which we cannot interpret phenotypic differences as traceable to a single locus, since a whole series of loci affect the normal functioning of the biochemical processes involved in the development of the trait—for instance, the biochemical pathways that lead to the synthesis of eye pigments. Moreover, epigenetic and environmental factors are as a rule involved in the development of phenotypes and phenotypic differences. If we consider, say, the textbook example of a “gene for blue eyes”, we will see that it corresponds, in fact, to a disjunction of alleles that can be responsible for a decrease of the pigmentation in the iris, and a disjunction is a logical expression, not a material entity to which a concept can refer [29]. This does not deny, however, the usefulness of the “gene for” concept: It helps us understand the results of a crossing between a brown-eyed father and a blue-eyed mother, and we can readily use pedigree analysis, accompanied by the apt simplification of assuming genes that determine the presence of brown or blue eyes.

What is at stake is that the deterministic gene is an instrumental concept that has predictive power and plays an important role in some explanations in genetics and molecular biology, but cannot be conflated with the molecular gene [30], which is a necessary but not sufficient condition for the development of many phenotypic traits [31]. There has been a tendency, however, for both scientists and lay people alike to overemphasize the importance of genes, applying deterministic thinking to conceive the role of genes in the development of complex multifactorial traits and disorders [15, 32], thus extending the deterministic interpretation beyond the domain in which it may be appropriate. Deterministic thinking has been used, for instance, in gene-centric explanations of complex phenomena like criminal behavior, political participation or sexual orientation (as discussed in [33]). This genetic essentialist belief may be problematic for our society since it engenders intolerant attitudes such as racism and prejudice against sexual orientation [16, 33–36].

In the last two decades, however, research in genetics has evolved into genomics and our understanding of genes and genomes and how they relate to development, phenotypic traits, cell physiology, among other features, has radically changed [29, 37–40]. It seems to be under way a general shift within the scientific community from a more deterministic to a more probabilistic understanding of the relationship between genes and traits (Fig 1). Advances in genetics and genomics have made it clear that gene action and function should be conceived as embedded into multiple hierarchical levels, in which complex networks of interactions between components are the rule [41]. Consequently, the probabilistic understanding of the structure, dynamics and functions of genes demands that they are located in complex informational networks and pathways [41].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. From deterministic to probabilistic understanding about genetics and genomics.

https://doi.org/10.1371/journal.pone.0169808.g001

Genetic science now tells us that, despite the usefulness of the deterministic gene as an instrumental concept in some explanatory tasks, in realistic terms it is not possible for any trait—even so-called “single-gene disorders”–to be determined by genes only, due to the influence of epigenetic and environmental factors [1, 2, 42]. There can be many genes associated with one trait, or many different traits associated with the same genes, which in turn are affected by a myriad of environmental factors. As shown in Fig 1, as we move from a “one-to-one” (deterministic) to a “multifactorial” (probabilistic) model of the relation between genes and traits, environmental and epigenetic factors are seen as playing a more important role in the development of traits and diseases. In the probabilistic model, genes are embedded in the context of an internal and external environment, with due attention to the fact that many genetic and epigenetic factors interact with one another, and, also, that genetic and environmental factors often interact non-additively, so that genes show different expressivity and penetrance depending on the influence of environmental factors [1, 4, 5].

Public knowledge about modern genetics and genomics.

Despite the progress of genetics as a research field, the developments in scientific understanding seem not to have reached the public: genetic deterministic beliefs still prevail to a great extent, according to the literature [15, 16, 26, 32, 33, 43–45]. Previous studies examining public knowledge of genetics generally indicate that such knowledge is low [20, 21, 46, 47]. In particular, people have difficulty understanding the mechanisms behind genetic risk [48], as well as the notion that genetic and environmental factors interact [9, 49, 50]. Most previous studies of the public understanding of genetics have been qualitative focus groups, case or interview studies [20, 21, 46, 49], as well as a few large-scale survey or questionnaire-based studies [47, 48].

Studies of mass media portrayals of genetics also generally indicate a predominant discourse of determinism [15, 32, 45]. Nelkin and Lindee [15] have advocated that public discourse on genetics is plagued by genetic determinism in such a way that human beings equate “all their social, historical, and moral complexity, with their genes” (p. 2). They have argued that determinism, or ‘genetic essentialism’ as they call it, is not simply a result of misunderstanding or simplification of science, but could be anchored in deep beliefs and social phenomena. As an example of such deep beliefs, Parrott et al. found that some people believe God plays an important role in how genes are expressed and impact health [51]. They believe, for example, that a higher power protects human genes from the effects of people’s unhealthy behaviours. This shows how predetermined views of biological outcomes may be deeply rooted in religious beliefs.

Studies on how genetics is taught in schools generally indicate that the predominant mode of genetics instruction, as well as the content of genetics textbooks, prime students to think deterministically [17, 43, 44, 52, 53]. It has been also shown that when students learn or talk about molecular genetics they tend to draw on more deterministic Mendelian explanations [54, 55]. In particular, according to Dougherty [44], “many students view phenotypes through the lens of Mendelian inheritance and fail to appreciate that most human traits are the product of polygenic expression modulated by the environment” (p.7). In addition, very few students have heard of newer concepts within genetics such as those relating to epigenetics [22]. Moreover, it has been shown that a significant part of primary and secondary school teachers (in Biology and in language) from different countries have strong genetic deterministic beliefs about some human personality traits [16].

Students in high school find genetics particularly challenging to learn because it requires reasoning across multiple organizational levels: between genes, proteins, cells, tissues, and organs [56, 57]. Students find it difficult to understand how mechanisms and interactions at the molecular (genes, proteins) and micro-levels (cells) bring about effects at the macro-level (organism, population) [56]. Moreover, they find it difficult to understand the mechanisms of gene expression, and how environmental factors interact with genes [22]. Within public health, genetic determinism is thought to have a negative impact on people’s understanding of health and disease [32, 33, 43, 51]. Genetic deterministic views can, for example, lead people to devalue the role of environmental or experiential factors in a number of traits such as mental illness, cancer, obesity, diabetes, which in turn can hamper efforts to prevent diseases [51]. In general, it is believed, thus, that genetic determinism poses a threat to genetic and health literacy.

Public attitudes towards modern genetics and genomics.

The concept of attitudes has been borrowed from the social psychology field into other areas such as public understanding of science and science education research. Fazio [58] gives a general definition of the concept: “An attitude is viewed as an association in memory between a given object and one's evaluation of that object. This definition implies that the strength of an attitude, like any construct based on associative learning, can vary”. Fazio’s model assumes that attitudes are important factors to predict behaviors. Even if the level of consistency of attitudes and behaviors is still discussed among experts, some empirical studies have shown a strong relationship between them [e.g. 59, 60].

Whilst public knowledge about modern genetics seems to be low, studies on public attitudes towards genetics and gene technologies indicate that the public are generally positive towards genetic applications, particularly if there are clear medical benefits [14, 24, 25]. There are, however, some specific genetic technologies that the public are more negative towards, such as cloning of human cells, human embryo research, and using genetic diagnoses to decide whether or not to continue a pregnancy [12, 24, 25]. People tend to be ambivalent towards genetically modified foods [14, 24, 25]. On the whole, we can say that people’s attitudes tend to be more favorable for uses of genetics that maintain a perceived “natural order” and more unfavorable for uses that are perceived to change that order.

Very few studies have investigated the possible links between attitudes towards genetics and belief in genetic determinism and/or genetic knowledge. A study about attitudes towards genetic testing among the Dutch public found no association between people’s attitudes towards genetic testing and their level of genetic knowledge or belief in genetic determinism [61]. A study by Rose et al. [62], in turn, found that in a U.S. sample, after adjustment for sociodemographic characteristics and family cancer history, higher knowledge about genetic testing for cancer risk was correlated with more positive attitudes towards genetic testing. However, this study did not include genetic determinism as a variable in their questionnaire. It is clear, thus, that more studies are needed, considering that belief in genetic determinism may be an intervening factor in the relation between knowledge about genetics and attitudes towards genetic applications.

Identification of core ideas and topics.

In the first step of the process, we identified a set of core ideas about genetics and genomics. Our set of core ideas covers two main knowledge areas: gene-environment interaction (section three of the questionnaire), and knowledge about modern genetics and genomics (section four). These core ideas are like “knowledge benchmarks” that the science education and genetics communities consider important for high school education and general public knowledge [64, 71]. We searched and identified core ideas from the literature about genetics literacy and genetics education, in addition to using our own expertise within genetics and genetics education. A complete list of the core ideas with descriptions, corresponding items, and sources can be found in the S2 Table.

The core ideas about gene-environment interaction (Section 3) are about whether and how genes and environment interact in the development of phenotypic traits, which can in turn influence their degree of commitment to genetic deterministic thinking. They reflect the transition from a deterministic to a probabilistic understanding in genetics, as shown in Fig 1. To identify the core ideas for this section we consulted literature on genetics, epigenetics, philosophy, psychiatry, and genetics education [1, 2, 42, 72, 73].

The core ideas in modern genetics and genomics (Section 4) are divided into three sub-areas: (1) characteristics of the genome, (2) gene function and expression, and (3) epigenetics. These sub-areas represent key conceptual domains, currently deemed essential to genetics literacy [64, 71], as well as newly emerging topics such as epigenetics, which may or may not yet be a part of high school or university genetics curricula.

To define the core ideas, we drew heavily on several curriculum benchmarks in genetics that outline what high school students ought to know and what content should be taught [71, 74], on literature about the current status of genetics education [7, 44, 52, 75], on developments within the scientific and philosophical literature [72, 73], and on genetics textbooks [76]. Please see S2 Table for a full description of the core ideas.

The attitude section of the questionnaire (Section 5) covers four topics within modern genetic and genomic-based technologies: gene therapy; genetic testing; prenatal genetic diagnosis; personalized medicine and pharmacogenomics. These topics were chosen on the basis of what is considered important for the public to know about such technologies now and in the near future [8, 9]. Some of these genetic applications are already in use or are fairly well-known, such as genetic testing and prenatal genetic testing. Others are currently emerging and are likely to become relevant for people’s lives in the near future, such as gene therapy, personalized medicine and pharmacogenomics.

Development and selection of questionnaire items.

In the next step, we developed a set of questions (items) that would test the core ideas. Where it was possible to relate our core ideas and topics to existing items in other questionnaire instruments, we gave priority to these items since they had already been validated and tested. Hence, most of the items in our questionnaire are based on items that have been used in previous studies, but we have modified them in order to match our core ideas. We also used the combined expertise in genetics and genetics education of the authors to help select and refine the items. Some of the items were also developed by the authors themselves.

Section two (belief in genetic determinism) consists of a task in which students were asked to indicate the relative importance of genes or environments for various traits presented in a table (herein called “the table of traits”). For building the table of traits task, we browsed several high school biology textbooks published in Brazil, which were submitted by publishing companies to a Brazilian National Program that distributes textbooks for public high schools across the country. We searched these textbooks to find examples of traits and diseases that are commonly mentioned in genetics education, since we wanted respondents to have heard of them. The table of traits task was designed with a Likert-scale response format, from “only environmental differences contribute to the traits” to “only genetic differences contribute to the traits”. The answers were originally scored on a ten-point scale, but later reduced to a five-point scale in the revision process (see section 4.1.1. above). We assume answers to this task are based more on beliefs about genetic determinism than on knowledge.

Section three (knowledge about gene-environment interaction) consists of a set of True/False/Don’t know items designed to measure deterministic understandings of genes and traits (more knowledge about gene-environment interaction indicating less determinism, and vice-versa). We included a “Don’t know” option in order to uncover topics or concepts that respondents potentially have not heard of before. We consulted previous determinism scales [16, 48, 51, 61, 77], but found that most of these previous items were either too general (e.g., “When people know their genetic make-up they will not be able to lead their own lives”) [61] or were too closely connected to sociopolitical issues such as racism, or were based on specific topics that students may or may not have much prior knowledge about. We therefore drew on ideas from these previous items, but made them less topic-specific and more closely related to our core ideas (focusing on the transition between deterministic and probabilistic understandings of genes and traits).

The items about modern genetics and genomics (Section 3) were largely based on items from previous questionnaires [63, 64] as well as curriculum benchmarks in genetics education [71, 74]. We also consulted the scientific literature on genetics and genomics [9, 72, 73], especially for the items on the genome and epigenetics, since there were not many previous items to draw from.

To develop the attitude items, we reviewed items used in previous questionnaires about attitudes towards genetics and gene technologies [13, 24, 61, 68–70]. We used a four-point Likert scale response format (strongly disagree, disagree, agree and strongly agree). In the first pilot we also included a “Don’t know” option, to see if some of the items were too difficult, but we removed it after the pilot studies.

Our initial questionnaire, used in the first pilot, had 51 items and 20 core ideas (See the S3 Table).

Evaluation by experts.

To test the content validity of the core ideas and items, we invited international experts from a range of fields related to genetics to evaluate the initial questionnaire and the core ideas. The experts were personally invited by email; thirty-two were invited and seventeen responded and evaluated the questionnaire. Of the seventeen experts who responded, four were working as biologists or geneticists, eight were working in genetics education, four in the field of sociology of biology, and two in the history and philosophy of genetics (one of them was both geneticist and educator). In addition, four of them had extensive expertise in designing assessment instruments in science education.

The experts were given a feedback sheet that they were requested to fill in, which contained the questionnaire items and the core ideas to which they were related. For the core ideas we asked: “Please indicate if you have any comments to the core ideas, for instance if you think some are unnecessary, if some should be altered, or if other core ideas related to modern genetics and genomics are missing.” For the section on Background information they were asked to comment on whether they thought some of the questions in the section should be changed in any way, and why. For the table of traits task we asked them to provide any helpful comments on the table as a whole. For the knowledge items on gene-environment interaction and on modern genetics and genomics they were asked to answer the following three questions with “yes”, “no” or “maybe” for each item: (1) Does the question test the core idea?; (2) Is this a quality question?; (3) Is it redundant? For the attitude items they were asked to answer with “yes”, “no” or “maybe” whether or not each item adequately probed for positive or negative attitudes towards the topic in question. They were also asked to provide any useful comments to any aspects of the questionnaire as a whole (for example, suggestions for improvements in wording or content).

In the first revision of the PUGGS questionnaire, the experts’ comments were analyzed for themes, for example, if several of the experts made similar comments to a particular item, this was especially considered. Also, items with a lot of negative feedback were particularly taken into account in the revision process, as well as the experts’ corresponding suggestions for improvements. Many of the experts provided useful comments on how the items could be improved, and some also suggested new items. The entire questionnaire and the core ideas were altered in the first revision process.

Interviews with students.

During the first pilot study, the questionnaire was administered to seven different classes on different occasions. At each of these sessions, after completing the questionnaire, two or three students were asked to participate in a short 3–5 minutes interview about their comprehension of the questionnaire items. We asked them three questions: 1) Did you find any of the questions difficult to understand?; 2) Why is the question difficult?; 3) Are there any words you didn’t understand?

Evaluation by experts: First revision of the PUGGS questionnaire.

Following the advice from the experts, we firstly simplified the core ideas by removing references to specific diseases or examples, to represent principles rather than detailed knowledge. Some of the core ideas were removed because they were considered too difficult by many of the experts, in the sense that they related to specific scientific knowledge that would either not be relevant from a public understanding perspective, or would be beyond the grasp of lay people or even basic education students. We also added a core idea about the levels of organization involved in the gene-phenotype relationship, as suggested by several of the experts. We kept two items per core idea for the section on genetic determinism, and in the section on modern genetics and genomics we increased to three items per core idea, to allow for future analyses of associations between the subsections of that section.

Some very minor adjustments were made to the background information (section 1) of the questionnaire. For section two of the questionnaire (table of traits), our initial version had a list of traits and disorders where respondents were asked to mark, on a scale from 0 to 10, the relative importance of environment and genes for each trait (from 0: Environment is most important to 10: Genes are most important). Several of the experts pointed out problems with this scale, mostly because it reflected a heritability scale (0 to 100% heritability). They advised us to reduce the number of points on the scale, and to use simpler categories. We therefore reduced the scale from a 10-point to a 5-point Likert format (from 1: Only environmental differences contribute to the trait, to 5: Only genetic differences contribute to the trait).

For section three (gene-environment interaction), the general advice from the experts was to make the items shorter, simpler, and less specific, so that detailed knowledge about specific traits or diseases would not be required from the respondents. For instance, the original item 4 read as follows: “A gene can be associated with several different traits—for example, the breast cancer gene mutation (BRCA2 gene) is associated with prostate cancer and melanoma”. In the revised version this item became: “A single gene can influence several different health problems”. On the basis of the experts’ responses, a majority of the items were modified, some were removed altogether, and some new items were added, improving content validity.

The items in section 4 (Knowledge about modern genetics and genomics) and section 5 (Attitudes) received fewer critical comments by the experts, and were generally considered satisfactory.

Evaluation through students’ interviews during the pilot testing.

Of the nineteen students that were interviewed about their comprehension of the questionnaire items, ten said they did not understand the items on epigenetics, because they did not know what epigenetics is. We also asked them if there were any words they did not understand, and ten of the students said they had never heard of epigenetics, two said they had not heard of Tourette’s syndrome (one of the diseases in the table of traits) and one said he did not know what pharmacogenomics is. Consequently, we removed the item “Tourette’s syndrome” from the table of traits task, but we kept the items on epigenetics and pharmacogenomics, because even though students had not heard of them, they understood what these items were asking. Moreover, their knowledge on epigenetics and attitudes towards pharmacogenomics are part of the data the questionnaire intends to gather. Apart from these remarks, all the other questions were understood. Seven of the students did not report any problems at all.

After the first revision of the PUGGS questionnaire.

The results of the internal consistency of the PUGGS questionnaire after the first revision are shown in Table 1. We tested the effect that removing certain items would have on the Cronbach alpha for most of the sections, and used these predictions as a guide to help us decide which items to revise or remove in the second pilot. The Cronbach alpha for the items on gene-environment interaction in (Section three) was poor (0.52), indicating that the items in this section were not all measuring what they were supposed to measure. Further revision on this section was therefore necessary, and, due to the number of changes, we recalculated the Cronbach alpha for this section based on the results from the second pilot. The reliability measures for the other sections were considered acceptable, although removing certain items improved the alpha slightly (see Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Cronbach alpha results of the first pilot test and decision-making about revision of the PUGGS questionnaire.

https://doi.org/10.1371/journal.pone.0169808.t001

Second revision of the PUGGS questionnaire.

On the basis of the Cronbach alpha results, as well as the students’ feedback in the interviews, the questionnaire was revised a second time. We also explicitly attended to items with a high frequency of “don’t know” responses, and assessed whether the wording of these items were confusing, thus making them difficult to comprehend, or if the content was simply too difficult. We also used the student interviews to address these issues, since the students had the possibility to address all questions and their difficulties. We did not explicitly consider the number of correct and incorrect answers to each item during the revision process, because we felt that the students’ feedback in the interviews and the number of “don’t’ know” responses addressed the issue of item difficulty more adequately. However, in most cases the number of respondents answering “correct” on an item was lower if the item had a high amount of “don’t know” responses. Examples of this are given below.

For the table of traits (section 2) we removed the items “coronary heart disease”, “haemophilia”, and “Tourette´s syndrome” because over 15% of the respondents answered “Don’t know” on these. Moreover, in the interviews two students said they had never heard of Tourette’s syndrome.

The original thirteen items on gene-environment interaction (Section 3) underwent the most extensive revision, since they had the lowest estimate of reliability (alpha 0.52). In order to avoid unnecessary confusion, we removed double negatives (items with a negative statement that probed for a “disagree” answer in order to remain positive). For instance, the item “Traits and diseases caused by a single gene are not very common” was modified to “Most human traits and diseases are caused by a single gene”. This item also had 26% “Don’t know” answers and only 19.3% correct answers in the first pilot, indicating that it was difficult to understand—but this increased to 78.8% correct answers in the second pilot. We also removed potentially difficult examples in items, e.g., diabetes, Alzheimer’s disease, personality, intelligence, either by substituting the item with a more general statement, or by substituting a specific example with a more common trait. For instance, an item on Alzheimer’s disease had the highest amount of “Don’t know” answers (36%), and only 37.2% correct answers, so we changed it from “Alzheimer’s disease is influenced by one gene only” to “A person’s height is influenced by one gene only”. The number of correct answers in the second pilot increased to 56%.

We also tried to improve the readability and clarity of the items in Section three by making them simpler and shorter. We deleted altogether four items that were either too complicated, or potentially misleading. This included an item on levels of biological organization, because it was too long and complicated (with 18.8% “Don’t know” answers); an item on diabetes, because it was too specific and also redundant; and two items on mutations, because they were too difficult for the intended respondents (we also removed the corresponding core idea). For the remaining items the amount of “Don’t know” answers was fairly low (ranging from 0.5% to 9.7%).

The Cronbach alpha in Section 4 (Modern Genetics and Genomics) was good (0.69) and it was not necessary to make many changes in the corresponding items. However, we removed two items due to potential inaccuracy: for instance, the item “Humans have about the same amount of genes as a fruit fly”. We made some minor changes in the formulation of some of the items to improve clarity, for instance by removing double negatives (e.g. “Epigenetic changes are not influenced by environmental factors” was modified to “Epigenetic changes are influenced by environmental factors”). Respondents frequently answered “Don’t know” to many of the items in this section. In particular, between 43% and 58% respondents answered “Don’t know” on the items on epigenetics. Despite this, we decided to keep these items since they represent newly established concepts in science and society, and answering “Don’t know” to newer concepts can give an indication as to whether people have heard about them or not. This is important in the context of this instrument since an understanding of epigenetics (involving an idea on how environmental factors can influence gene activity) could be a possible factor inhibiting genetic deterministic beliefs.

In the Attitudes section, very few students answered “Don’t know” (ranging from 0% to 5.8%), indicating that they understood what was being asked in these items. This was also confirmed in the students’ interviews. Also, the overall Cronbach alpha for the Attitudes section was very good (0.85). We therefore did not change much in this section, except for some of the wording—in particular, we modified double negatives, since they are potentially confusing. For example, in the subsection of gene therapy the item “I would be worried about gene therapy being used to modify or enhance physical attributes such as athletic performance” was modified to “I think gene therapy should be used to modify or enhance physical attributes such as athletic performance”. There were also two items in the subsection on genetic testing that had double negatives and this sub-section had the lowest Cronbach alpha score (0.54) of the four attitude topics. These items were subsequently modified so that they would be easier to comprehend. Although removing these items would have had a positive effect on the Cronbach alpha for each of these subsections, we opted to modify instead of removing them because we think that the items cover important issues within genetic testing and gene therapy.

We also changed some of the answer options: Since the questions on gene-environment interaction (Section 3) and modern genetics and genomics (section 4) are predominantly knowledge questions, we changed the modalities for these from a Likert to a True/False response format: In the first pilot, we used a Likert scale for all items, with the options “strongly disagree/disagree/agree/strongly agree/don’t know”. This was changed to a True/False (and “Don’t know”) format for sections 3 and 4 in the second pilot. We decided to include the “Don’t know” option because it gives valuable information about the topics and items students know less about. We removed it from the Attitudes section (Section 4), since very few students marked this option. We also decided to remove the “Don’t know” option from the table of traits task (Section 2) since this task is capturing belief in determinism, rather than pure knowledge, and therefore assume that most students can answer this task regardless of how much knowledge they have.

After the second revision.

After the questionnaire had been revised a second time, another pilot test among a new sample of students (n = 78) was conducted and the Cronbach alpha was measured a second time for the parts of the questionnaire that had undergone substantial revision (Section 3 on gene-environment interaction). The new alpha score for the items on gene-environment interaction was 0.67, which was acceptable (Table 1). This indicates that the second revision had significantly improved the reliability of these items and we decided that no more revision was necessary. The final version of the PUGGS questionnaire contained 45 items and 11 core ideas.