Introduction

Genetics has been one of the fastest-evolving branches of Biology in the last decades. This discipline is present in many aspects of the daily life of citizens: the media are responsible for the popularization of terms such as DNA, genes and genetic tests, and journals, news and social media contribute to disseminating scientific advancements and contribute to the debate in society (genetically modified aliments, gene therapies, cloning, vaccination, etc.). This requires that the population is scientifically literate, for citizens to be able to participate and make informed decisions in a scientifically and technologically highly developed society (Abril & Muela, 2015).

Despite the personal relevance of possessing some basic knowledge about genetics, genetics literacy in the general population is usually low (Chapman et al., 2019). The formal instruction that could help improve this literacy takes place for the first time in secondary in many national curricula. Spain is one of the cases of late introduction. There, pupils begin to delve into the more detailed study of these topics towards the end of secondary school (15–16 years in the Spanish educational system; grade 10). During primary school, the curriculum focuses on the fundamental characteristics of living beings, nutrition, and perpetuation, although without establishing an explicit connection with inheritance or with the transmission of genetic traits. In the early years of secondary school, the study of the cell and its division is addressed (12–13 years). However, it is not until the last year of secondary school (15–16 years) that the contents related to biological inheritance are explicitly incorporated into the curriculum.

This way, most of the educational research on the teaching of genetics conducted during the last two decades has been focused at the secondary school level, confirming that pupils at the end of Middle School have serious issues understanding and integrating genetic concepts (Banet & Ayuso, 2000; Duncan & Hmelo‐Silver, 2009). The latest pieces of research confirm these same difficulties (Haskel-Ittah & Yarden, 2021). Research shows that the topic is hard to grasp because it is complex and abstract (Duncan & Reiser, 2007; Lewis & Wood-Robinson, 2000), and also because it requires pupils to deal with phenomena at different levels of biological organization (molecules/ cell/ organism) (Duncan & Reiser, 2007; van Mil et al., 2013). Moreover, different studies show that certain simplifications can lead pupils to consider that a gene is always responsible for a trait or that a gene with a mutation always causes a disease, representing a kind of ‘genetic determinism’ (Carver et al., 2017; Gericke & Mc Ewen, 2023; van Mil et al., 2013).

Concerning the research on primary school pupils’ knowledge of biological inheritance, the research shows that pupils formulate the first theories on kinship since preschool (K, 5–6 years), when pupils become aware of the similarity between parents and their siblings (Solomon & Johnson, 2000; Springer & Keil, 1989). Regarding the mechanisms that explain this similarity, some authors report that a large proportion of pupils over 10 recognize that both parents contribute to descendants (Cisterna et al., 2013; Smith & Williams, 2007), although they can attribute higher or lower resemblance to unbalanced contribution from each parent. I.e., younger pupils think that daughters receive more genetic information from their mothers, while sons get more from their fathers. So, if the child was a girl, then a larger percentage of the genes would be inherited from the mother, while if the child was a boy, the majority of the genes would come from the father (Chin & Teou, 2010).

Furthermore, most primary school pupils accept that children may look like grandparents. This idea is mainly based on observation as evidenced by the fact that children living in rural environments or in frequent contact with pets, and thus allegedly with better direct knowledge of the variability in animals and plants, considered more often the possibility of traits jumping a generation (Williams & Smith, 2006). However, in general terms Primary school pupils cannot explain the mechanism underlying this resemblance to the grandparents, even after a 4-week intervention that dealt with the appearance/ disappearance of certain traits in plants (Ibourk et al., 2018), with some notable (but rare) exceptions, where single pupils argument that “the genes transmitted by their parents were independent of those of their grandparents” or that “a child could may not have the parents’ genes but it has the grandparents’ genes (Chin & Teou, 2010).

Another basic idea necessary to build foundational knowledge about genetics is the comprehension that some traits are inherited, while others depend on the interaction with the environment. Strictly speaking, and as shown by recent advances in epigenetics, this distinction should be nuanced, and understood as a gradient and not as a dichotomy (Gericke & Mc Ewen, 2023) However, this simplification into biologically inherited and acquired traits may be valid as a first step in genetics learning progression. This core idea, the heritability only of certain traits, is the main construct analysed in educational research focused on the preschool context, where publications are scarce and date back to the 80 s and 90 s. Research showed that by the age of 7 children can differentiate biological inheritance from cultural transmission (Solomon et al., 1996), and build a differentiated model. Indeed, very young children tend to think that traits with a biological function (i.e. big eyes to see the enemies) are heritable (Springer & Keil, 1989), while social (language, food preferences, etc.) or psychological functions are not inherited. Regarding kinship theories among elementary school pupils, it is worth noting the work of Venville et al. (2005), which confirms the conclusions of this research at preschool and shows that most pupils aged 9 years and older differentiate between socially and genetically inherited characteristics.

However, they hold some doubts about the heritability of "accidents" generated by the environment (for example, losing a part of the body). With differences depending on the type of organism (plant, human or animal), it is from the age of 11 when most of them predict that the offspring do not inherit this "accident" (Kargbo et al., 1980).

If we focus on the molecular structures of heredity, pupils do not have a clear notion of the location or function of genes, although they may have heard of them. Several pieces of research show that pupils aged 10–12 have encountered concepts such as DNA or genes thanks to TV, TV series, films, video games, etc. (Donovan & Venville, 2012). However, these terms can have different meanings for elementary school pupils. A study involving pupils aged 9 to 15 revealed that pupils link genes to the transmission model (i.e., they are the entities that are passed on from parents to descendants, or carriers of traits), while DNA is linked to the model of "identity" (i.e., it is "what we are", and serves to identify a criminal, a parent, etc.). Thus, the attention to DNA is driven by the particular technological use that humans made of it, more than by its role in biological processes. Moreover, DNA is conceived as a "fingerprint" that makes individuals different from others, while genes make them alike other people, which suggests, at least in part, opposite considerations (Venville et al., 2005).

Lastly, considering the idea of evolution, which encompasses all the previous models (basic inheritance, interaction with the environment and molecular genetics), research shows that primary school pupils can generate adequate ideas to explain the phenomenon of evolution, with a large number and variety of explanations (Vázquez-Ben & Bugallo-Rodríguez, 2022). The systematic review by Bruckermann et al. (2021) suggests that the youngest pupils possess incipient knowledge not only of heredity and variation of traits, but also of natural selection. Even more, at the primary school level, pupils are aware of variability and may understand that variable traits may confer adaptative advantages, and thus impact the chances of the individual or population surviving (Gormley et al., 2022).

In summary, although the works focused on the comprehension of heredity and kinship at primary school are still scarce if compared to High School, research on knowledge of biological inheritance shows that elementary and middle school pupils can elaborate explanations about some aspects of heredity and kinship, although with a certain level of confusion, and without a detailed knowledge of the underlying biological mechanisms (Ibourk et al., 2018; Solomon & Johnson, 2000). These limitations can become an obstacle at later levels (Secondary school), since knowledge that is built on an unstable basis can hinder the understanding of more advanced concepts (Duncan et al., 2009; Ibourk et al., 2018). As such, it seems desirable to begin the systematic introduction of precursor concepts from initial levels (Primary School) (Russell & McGuigan, 2015) to guarantee the construction of a solid foundational model of inheritance.

Objectives

Most of the research done in elementary school, including the works referred to in the previous sections, have been conducted in countries where curricula include references to genetics concepts. That is not the case in Spain, where genetics is first introduced formally in grade 10. Therefore, with the ultimate aim of making a proposal that allow for a progressive introduction of precursory genetic concepts in earlier levels, our initial objectives have been:

  1. A)

    To characterize the incipient heredity model of knowledge of elementary school pupils, i.e., before formal instruction

  2. B)

    To evaluate the progression of the heredity model with age (i.e., cross-sectional study) to identify the most appropriate time to introduce these basic ideas.

We have aimed to characterize two main core domains:

  • Which traits are biologically inherited, and which are acquired (result of the interaction with the environment? (Hereafter WHAT is inherited?)

  • The mechanism by which descendants resemble their parents (HOW the traits are inherited?)

Methodology

Participants

In Spain, compulsory education starts at the age of 6 (Grade 1) and lasts until 16 (G10); this period comprises primary education, G1 to G6, and secondary education G7 to G10. In these periods optativity is very limited, implying that most pupils follow the same itinerary, at least until G9. All the schools must follow the same prescriptive curriculum, irrespective of their nature (public, subsidized or private).

To the study, three schools from two Spanish regions were selected. The three schools were public urban schools, mixed classes (51.1% male, 48.9% female), with an average number of 25 pupils per class (as defined by law), and 2–3 classes per grade. Most pupils are from middle to low-income families, and there were no pupils with special education needs.

In this context, 535 primary school pupils took part in the study (6–12 years) (Table 1). The selection criterion for the survey was convenience sampling and all the pupils of each of the selected schools took part in the study.

Table 1 Summary of the sample
Full size table

All the pupils answered a paper-and-pencil questionnaire. In addition, a sample of thirty participants from the whole sample were selected to undergo interviews. The selection of informants proceeded by purposive sampling, i.e., the researchers selected the pupils that could provide the most informative answers and hence the most nuanced depictions of the heredity models across ages (objective 2) (Etikan & Bala, 2017). For that purpose, the researchers identified the pupils with more extensive discourse in response to the open-ended questions. The selection was irrespective of the accuracy of their answers, so as to try to avoid bias and capture most diversity in the underlying models. A sample of 1–2 pupils per class (2–11 per academic year) was found to saturate the variation of responses (Zapata-Barrero & Sánchez-Montijano, 2011). However, it must be noted that the cross-sectional analysis was based on the answers to the questionnaire, and that the objective of the interviews was to develop a more nuanced understanding of these answers.

Research Instrument

Questionnaire

Before this study, and as a part of the validation process, a preliminary 22-question survey was answered by 300 pupils in grades 1–6, from two different schools. Some of the questions were modified from Ibourk et al. (2018), and the rest were created de novo.

The final version of the questionnaire was modified according to these preliminary results and the suggestions made by several elementary school teachers: unclear sentences or ambiguous words were rephrased, some questions were modified to refer to contexts closer to the pupils (kittens instead of crows), and redundant questions removed, with the ultimate goal of improving feasibility, readability and consistency (face validity). The final version included 12 questions, selected to cover all the relevant connections within the two main constructs considered (content validation): (1) Which traits are inherited? (WHAT) and (2) How are they inherited? (HOW) (Fig. 1 and Table 2 and 3).

Fig. 1
figure 1

Concept map (created with CmapTools) summarizing the main core concepts that structure basic knowledge about genetics. Concepts in the WHAT domain, coloured in orange; HOW, in purple (see Tables 2 and 3)

Full size image
Table 2 Summary of the questions included in the survey
Full size table
Table 3 Statements of the questions included in the survey, and the core concepts assessed
Full size table

Considering the limited reading and writing abilities of younger pupils, resulting in earlier fatigue, as well as cognitive limitations (López & Ridao, 2012), the questions 7 to 11 were only answered by pupils in grades 4–6. The first 6 questions were identical for all the courses, except for question 1, where the term “inherit” was replaced by “receive from the parents” for grades 1–3. Moreover, only the students in grade 6 answered the question 12.

Table 2 shows the distribution of questions by grade, domain and type of question. Table 3 shows the statements of the questions. The complete questionnaire is available as supplementary material.

The pupils were allowed one class session for answering the questionnaires. Not to condition their answers, especially regarding the comprehension of the concept of "inheritance", teachers and researchers provided no help.

Interviews

During the interviews, the researcher and the student revised the questions and answers given one by one. The researcher read the question and the given answer and asked the student to provide a justification. No further questions were asked beyond those necessary to request the pupils to justify their claims. Some examples of the questions formulated were:

  • Q: Why have you chosen this option?

  • Q: Why have you colored one parent red feathered and the other blue feathered?

  • Q: How do you explain that the guinea pig will have this type of hair?

The interviews provided a closer insight into the level of comprehension of certain key ideas and illustrated the reasoning behind the given answers. They were only intended to illustrate the answers, and hence didn’t seek a complete representativity (for example, gender balance).

The interviews also served to ensure the reliability of the questionnaire: 85% (range 70%—100%) of the answers obtained from the questionnaire would have been rated similarly as based on the interview, including those answers that do not provide different information but also additional, more nuanced answers that allowed the researchers to do a finer grained categorization of the answers to the questionnaire.

Analysis

For the purposes of the analysis, the answers to the multiple-choice questions (Q1, Q2, Q8, Q9 and Q11), in addition to the table in Q6, were coded as dichotomous (correct/incorrect).

The answers to the open-ended (Q5, Q7, Q10, Q12) and colouring questions (Q3 and Q4) were classified into inductively defined categories. To do this, initially, each researcher independently analysed 20 questionnaires and proposed some categories for analysis. These categories were contrasted and refined to define the final coding scheme, using also recent studies as a reference (Ibourk et al., 2018). From this point, researchers worked in pairs to first code individually and then reach a consensus on the categorization of the open questions. Although the initial intercoder reliability was high (90%), discussions continued until a consensus was reached for all the questions.

The analysis of the questionnaire was conducted using the software RStudio. On top of the descriptive analysis of frequencies, Sommers’ d statistic was used to measure the intensity of the association between grades (1–6) or stages (1–3, 4–6) and categories of answers, and X2 to assess the independence of variables. Somers’ d is a measure of association for ordinal variables, which ranges from -1 to 1. Values closer to ± 1 denotes high association, while values close to 0 indicates lack of dependence. Confidence intervals are calculated with a significance level of 0.05.

Results

In the following lines, we will review the main results separately for the two domains considered (WHAT/HOW).

A. Which Traits are Biologically Inherited, and which are the Result of the Interaction with the Environment? (Fig. 1, Orange)

A1. Resemblance Parents- Descendants

Most of the primary school pupils, of all levels, are aware of the similarity between parents and their descendants, but not only in human examples but also among plants. Indeed, 92% of the pupils of all levels clearly identified the most alike flower as the most likely descendant of a mature flower (Q2).

A2. What is Inherited or Acquired?

The results show a progression in the proportion of right answers across grades, when the primary school pupils are asked whether different traits are inherited or not when referring to exclusively inherited traits (Q1, 11), acquired by accident or cultural transmission (Q5, Q7) or a combination of both types (Q6) (Tables 4 and 5).

Table 4 Percentage (%) of right answers to Q1, Q5, Q7 and Q11 and Somers’d statistic by stage and grade
Full size table
Table 5 Percentage (%) of right answers to items of Q6 and Somers’d statistic by stage and grade
Full size table

In the first place, pupils perceive that physical resemblance is inherited (Q1 and “fur colour” in Q6), with higher frequency the higher the grade. It is also true for plants (Q11), where the pupils at higher grades (G4-G6) clearly identified (> 75%; Table 4) that the distinctive shape of the leaves of a tree depends on the species, and not on other ambient factors (wind speed, water or nutrients availability).

The following excerpt shows that the pupil understands the difference between traits that are genetically inherited or acquired during socialization.

QFootnote 1: Which of these features come from its parents?

A:I think that the tricks it does come from its parents. It cannot do these just because “It’s like that”, I think its parents must have taught it. And then the fur colour, I don’t think its parents have taught him the colour, it’s born like that. (student77, G2)

In the second place, pupils have more difficulties in identifying physiological traits (hearing well or living long) as inheritable (Table 5), although the success rate is higher the higher the grade.

However, when pupils are faced with the possibility of inheriting a trait that has been acquired accidentally (Q5, a dog gets its tail cut), most pupils give the right answer, from the earliest grades (78% stage 1), and improving with age (88% stage 2) (Table 4). More in detail, the younger pupils (G1- G3) provided no or vague justifications, being the categories “with tail” and “with tail – others” (see Table 6 for definitions of the categories) the most frequent (38,5% and 31,1%, respectively, and decreasing with grade). However, the most frequent category among older pupils (G4 – G6) was “with tail- H” (Table 6) (39,4%, as compared to 17,8% in the previous stage). Moreover, 13,1% of the pupils alluded to essentialist reasons (category “with tail-dogs”; Table 6) to explain why the acquired trait (get the tail cut) was not inherited, and this was more frequent among younger pupils.

Table 6 Examples of answers to Q5, by inductively defined category
Full size table

Last, when we consider culturally and socially transmitted traits such as speaking a language (Q7), the answers vary according to the grade, although overall a high percentage of the pupils say this trait is learnt (Table 4).

B. How are the Traits Inherited? (Fig. 1, Purple)

Variation of Traits and Parents’ Contribution

The second key concept assessed in the questionnaire was the variation of traits between parents and descendants, and the mechanism that explains this difference.

To that end, the answers to multiple choice questions were categorized as true/ false, and open and colouring questions (Q3, Q4, Q10, Q12) were categorized as shown in Fig. 2 and 3 and in Table 7.

Fig. 2
figure 2

Categories defined on the basis of the colour patterns in the pupils’ productions for Q3 and Q4. Any other colour pattern was categorized as “other” in both questions

Full size image
Fig. 3
figure 3

Examples of pupils’ drawings for the “same sex” category (a) and “Both “ category (b) (Q4)

Full size image
Table 7 Categories, description and examples for questions 10 and 12
Full size table

The pattern of answers to Q4 suggests that youngest pupils often think that siblings are identical to one parent (Fig. 4a; Somers’d (stage) = -0.217 [-0.221, -0.215]); frequently, the parent of the same sex. From grade 3, the idea that siblings express traits of both parents becomes more prevalent (Fig. 4a). Moreover, the frequency of new variants, even suggesting new phenotypes (new traits, not present in the parents—e.g., green eyes- or mixture of previous traits – e.g., mestizo faces—increase in older pupils (Fig. 4b; Somers’d (stage) = 0.277 [0.274, 0.280]).).

Fig. 4
figure 4

Analysis of drawings in Q4. a Percentage of drawings showing children identical to one parent vs. children with traits from both. b Drawings coloring children with new or intermediate traits vs. children with traits similar to those of the parents

Full size image

With the aim of characterizing this model of inheritance, we proposed a question (Q3) that refers to the same idea, but thinking “backwards”; i.e., it requires the pupils to infer the parents’ phenotype from the appearance of descendants. Like the previous question (Q4), most pupils predict that parents will have the same colours that siblings. However, older pupils make alternative predictions. Namely, there is a clear increment in the number of pupils that predict that parents will have simultaneously both colours found in the descendance (red, blue) (6.9% in G1 to 16.18% in G6), or a new (sometimes intermediate) colour (8.3% in G1 to 15% in G6).

Supporting these results, when older pupils (G4-G6) are explicitly asked how these traits are inherited (Q8) through a multiple-choice question related to the fur type of guinea pig, most of them (74–77%) recognize that descendants receive information from both parents even though the descendant only looks like one of them, without significant differences between grades (Fig. 6a). If we look at the distractors, the younger pupils frequently select the idea that siblings receive information only (or mostly) from the parent they resemble more (16.7% in G4, 13.8% in G5 and 11.2% in G6).

However, when an intermediate character appears (a cat with grey fur, descendant of a white and a black cat) (Q9), the percentage of right answers is lower (53–63%) (Fig. 5b), and only a 20% (18–22%) of pupils in G4- G6 choose that only grey cats inherit the fur color from both parents.

Fig. 5
figure 5

Percentage of right answers for question 8 (a) and 9 (b)

Full size image

The next step in this learning progression depends on understanding how these new characters can appear in the descendants, without being present in the ancestors. Like this, in Q10 – is it possible to obtain white flowers from purple parents – are mostly G6 pupils who recognize that a trait can skip a generation, justifying that this white color comes from grandparents (65% vs. 42–33%) (X2 = 1.282, gl = 1; p = 0.2575) (Fig. 6).

Fig. 6
figure 6

Percentage of answers for question 10

Full size image

Although in this statement it is implicit that the trait is also present in the parents (albeit hidden or inactive), very few pupils signal to the parents for the origin of the white color (see excerpt of the interview to student327-G5, when the student expresses also an incipient idea of dominance), and they even deny that parents may have inherited it (interview to student242-G4).

Q: Is it possible that two purple flowers can have two descendants, one purple and the other white?

A: Although the white colour is not visible [in the parents] they may have inherited it in some part of the flower, and then, when they have children, they [the children] can be white… or also, although both of them are purple [the parents], they can have inherited something from the grandparent, and also from the mother and the father. (student327-G5)

A: I think it is possible; if the grandparents are white, you can inherit it, even if your parents haven’t, because it runs in the family. (student242-G4)

Q: But the parents are purple… How can it be that one of the daughter flowers is white?

A: I don’t know. (student242-G4)

Amongst the most advanced explanations, only 4 pupils used the molecular model of inheritance, in reference to genes and DNA.

Q: And how has it passed from grandparents to grandchildren?

A: This one [points to a purple parent] may have the white in the genes, but it is not visible, and it has passed it to the siblings. (student461-G6)

Q: What do you mean by “it is not visible”?

A: It may not be physically apparent but be in the genes. (student461-G6)

Knowledge about the Molecular Mechanisms of Heredity: Relationships between Traits, Gene, DNA and Chromosomes

Although in the individual interviews, pupils used terms such as gen or DNA, and even more specific words such as dominance, only the questionnaire for G6 included a question in which the pupils were prompted to use the words DNA, chromosomes and genes to explain the inheritance of a physical trait.

In this respect, most pupils in G6 did not make any link between physical traits and genetic structures (DNA, chromosome and gene) (55.2%). Regardless of whether they expressed the relationship, the words “gene” (23.3%) and “DNA” (21.6%) are much more common than chromosome, that is less known (10.3%) (Table 8).

Table 8 Percentage of pupils that made reference to DNA, chromosome and gene in Q12
Full size table

Only a fifth of the participants (19.8%) explicitly related traits with some of the mentioned genetic structures (Table 9). In these cases, the most cited is the relationship trait- gene (9.48% of the total of pupils in G6, student465), and only a small fraction cites simultaneously the three structures (4.31%, student439).

Table 9 Percentage of answers to Q12 considering the relationship between traits, genes, DNA and chromosomes
Full size table

Student465-G6: Example of relationship gene – trait:

“The genes of the parents have come together and have created those ears”.

Student439-G6: Example of relationship gene – DNA – chromosome:

“Because the mother had in the DNA these genes, and when she has got a baby, the DNA of the mother has been transferred in the chromosomes”.

In the interviews conducted it became clear that the concept of gene is often associated with “something” that is inherited from parents to children, while DNA associates with a unique identifier that defines the identity of individuals.

Q: Have you ever heard the word “gene”?

A: Yes, I know it in my way. It’s like a part that is how a person is… and when a child is born it takes genes from the father and from the mother… and when they come together, they have a baby, with an aspect that is different but similar. (student24-G4)

A: A gene is something that is inherited that the humans have. (student389-G5)

Q: And do you know what is the DNA?

A: Yes, it’s like a IDcard that has the blood for identifying you. (student263-G4)

A: Yes, I’ve heard it. It’s something like the saliva, for analysing it and saying this or that, and for knowing who is right. (student298-G4)

A: In the news… and sometimes in the school. They are things that the children inherit from the parents… The DNA is something that they extract for analysing it. And gene what the son has got form the father. Chromosomes? No, I have never heard of. (student402-G6)

Discussion

The objectives of this work were to know the previous ideas of primary school pupils about inheritance and to evaluate the progression of their model to make educational recommendations to introduce foundational ideas about inheritance at primary education. The research was conducted in a context with no specific instruction on this topic and without presence of inheritance in the curriculum of science at primary education. Specifically, we aimed to assess their knowledge pertaining the inheritance of hereditary and acquired characters, the possibility of variation of traits between parents and children, and their knowledge about the mechanisms and patterns of heritance. Generally speaking, the results show that elementary pupils have pre-instructional models of inheritance close to normative genetics ideas, even at very early age.

In coherence with previous studies (Ergazaki et al., 2015; Solomon & Johnson, 2000; Springer & Keil, 1989), since the age of 6 (G1) pupils are aware of the resemblance between parents and children.

Looking first at the distinction between biologically inherited and acquired traits, our results showed that, similar to previous studies (Kargbo et al., 1980), younger pupils correctly identified the behaviours that cubs learnt from their parents, although a significant percentage identifies as biologically inherited traits that are socially acquired. The cross-sectional study showed that pupils in G4-G6 were able to discriminate better than younger pupils both types of traits. Based on the pupils’ answers and specially on the new insights gained in the interviews, the following progression in the model of inheritance can be depicted (Fig. 7).

Fig. 7
figure 7

Identifiable progression in the inheritance model distinguishing between inherited and acquired traits

Full size image

Regarding the acquired traits, we can observe that the pupils use different models of inheritance for traits that are learnt (i.e., to speak a new language) or acquired “by accident” (i.e., a traumatic event or an accident that affects one parent), with a higher ratio of right answers for the latter. For acquired traits, even the younger pupils have an intuitive incipient model, including most pupils in G1- G3, who identified that the cubs won’t inherit the missing tail. This figure is coherent with percentages reported in previous studies, with partially overlapping questions but much smaller sample sizes (Kargbo et al., 1980). Despite having this intuitive pre-instructional knowledge, pupils have difficulties in justifying why this character should not be inherited. In fact, few of them gave explanations using terms related with genes, inheritance or said that an accident or disease are not passed on to the next generation. The evolution of this model was clear, as can be noted when comparing the percentage of responses in the category “With tail-H” between the youngest and oldest pupils. Previous research works showed that, even after instruction, pupils at G5 have difficulties for building a coherent model for explaining that an “accident” is not inherited (Ibourk et al., 2018), and that only 31% of the students give partially coherent explanations. Consequently, the inheritance of environmental physical experiences seems to be one of the key leverage points for reconstructing the model of inheritance.

One of the results we should bear in mind is the percentage of essentialist answers (it will have tail because all the dogs have one), which remained constant across all the courses (except G1). Psychological essentialism is defined as the belief that organisms have fixed characteristics (essences), which makes them belong to a group of similar organisms (Stern et al., 2022). This essentialism is presented from early years (Gelman, 2003) as it could be detected when children learn words (i.e. they are able to generalize using the word “dog” to any animal with similar traits, despite their differences). In turn, genetic essentialism refers to the model that deems inherited traits inheritable (it’s in the genes, so it cannot change), and which considers that individuals that share more genetic load have a resemblance that is invariable (families, taxonomic groups, races, etc.). This genetic essentialism, which implies mastering the idea of gene, has been widely studied among older pupils (high school and university) (Stern et al., 2020), and hampers comprehension of inheritance and evolution.

Regarding the idea of how inheritance works, our research showed that the idea of the contribution from both parents is present from G1 although it is more frequent among learners over 10 (G4 and older), similarly to previous publications (Cisterna et al., 2013; Kargbo et al., 1980; Terwogt et al., 2003). Distinctively, introducing in our questionnaire a set of colouring questions allowed us to delve into the conceptions of the youngest pupils, without the limitations associated with translating mental models into words, and giving them the opportunity to freely suggest new variants of the traits, unlike previous works with closed options (Ibourk et al., 2018; Terwogt et al., 2003; Williams, 2012). This allows us to detect that older students not only consider a combination of parents’ traits in descendants, but also, the possibility of the mixture of traits or the presence of new variants.

Like Terwogt et al. (2003) we consider that the pupils’ model of inheritance to explain family resemblance is not generated just by observation, but is also a consequence of verbal interaction in the social environment. The inconsistence of this model in pupils under 10 could be related to the lack of the development of abilities of probabilistic reasoning and hypothetical deductive reasoning by this age (Williams, 2012).

The next step to understand inheritance mechanisms implies the knowledge that genetic information received from parents could be expressed or silent in the descendants (Fig. 1). Thus, traits in the descendants are the result not only of the combination of the parents’ genetic information, but also of the regulation of the gene expression. Q10 was included in the questionnaire to evaluate this concept. Results showed that most students justified the presence of a new trait in the descendants because it was present in the grandparents. Although this fact implies that “something” has passed on from grandparents (P) to parents (F1), and then to the third generation (grandchildren, F2), very few pupils pointed out the parents as the origin of this new trait.

Educational Recommendations

This cross-sectional study showed that the model of inheritance that allows for correctly distinguishing inherited and acquired traits is progressively developed (Fig. 8), and that younger pupils may have difficulties in dealing with the inheritance of physical experiences. One strategy potentially useful would be proposing activities that show the variability and evolution of living beings within the groups traditionally studied at the school (for instance, https://www.nsta.org/biological-diversity).

Fig. 8
figure 8

Concept maps summarizing the main core concepts that might structure basic knowledge about genetics for grades 1–3 (a) and grades 4–6 (b). Grey circles show the different concepts introduced for grades 4–6 with respect to grades 1–3

Full size image

The second weak point in the intuitive model of inheritance of the pupils, that should be the objective of future intervention, is the idea of dominance- recessiveness, that explains the reason why some traits may remain hidden or skip one generation. To the best of our knowledge, no intervention has been published to introduce the idea of dominance – recessiveness in Primary.

This reconstruction of the model should be approached without specifically describing the DNA molecule, genes or chromosomes, given its (sub)microscopic nature. However, it cannot be ignored that the pupils have heard these terms in their social environment (family, media, videogames, etc.), as evidenced previously (Donovan & Venville, 2012, 2014) and we noted in our study (I have heard the word gene from many people, from my parents talking about…they say: You have the same genes than me, student492-G6). In consequence, it would be necessary to employ analogies that allowed for reconstructing the genetic model of inheritance, as proposed by several authors (Ergazaki et al., 2015; Solomon & Johnson, 2000; Springer & Keil, 1989). The aim would be to help them move from a “theory of kinship” to building a solid basis for a “theory of genetics” (Elmesky, 2013).

With this aim, it would be necessary not only to work the concepts discussed in this study, but also to build different constructs in parallel. Following the learning model of Castro-Faix et al. (2021), it would be convenient to first mention that there is “something” that carries information (construct A) and explains how we are (construct B). This information goes from parents to children (constructs E1 and E2) and, sometimes, part of this information is not seen externally (construct F). Last, there are traits that do not only depend on the inheritance, but on the interaction with the environment (construct H).

Based on the initial model we presented in this introduction (Fig. 1), we would like to suggest two different theoretical inheritance models for youngest and oldest primary school pupils (Fig. 8).

Limitations

This study has also some limitations that should be acknowledged. Despite the extensive pilots, some problems with the interpretation of symbols (i.e. interpretation of pedigrees) and the language persist. For example, in the interviews, the youngest pupils (G1-G3) did not seem to identify the expression “receive from the parents” (used in the questionnaire) with inheritance of traits, but with something that is passed on by cultural transmission at home, so answers to some of the questions (Q1, Q6) could be better than found. This should be revised in future implementations of the questionnaire. Second, only three schools were included in the study. However, the comparability of the results among the different scenarios, the lack of salient features, and the total sample size reached (number of pupils) make us think that the results can be considered representative and transposable. And lastly, the fact that some questions were no answered by all the courses (i.e., younger pupils weren’t asked about DNA, gene and chromosome concepts, and “hidden traits”) implies that we cannot be certain as to whether these concepts are already known and can be advanced to earlier levels. Future research would be advisable to gain further insight into these concepts.

Conclusion

Our study has allowed us to characterize the incipient inheritance model of primary pupils, which, generally speaking, is close to normative genetics ideas. The size of our sample and the cross-sectional nature of the study allow us not only to confirm the previously published research but also to define better the progression of this knowledge. Therefore, we have suggested as necessary point of intervention, first, to work on the difference between biologically inherited and environmentally acquired traits. Learning instruction of this point have to pay special attention to the language, especially with the youngest pupils that confound the terms, inherit, receive, pass on from, learnt or acquire. Second, to propose activities showing intraspecific variability and evolution, in order to overcome the natural essentialism detected. Thirdly, reinforce the idea of the contribution of both parents in the genetic endowment of the new individual and, therefore, the possibility of receiving two different pieces of information for the same trait. And finally, using analogies to the concepts of DNA, gene or chromosome, to introduce the dominance- recessiveness construct for G4-G6 pupils.