If I could try to replicate any study, it would be that of Joseph Novak, in the 1960’s and 70’s.

Novak is most famous for having “invented” concept maps, but the story of why he developed concept maps in the first place is little-known and particularly relevant to primary school teachers and science curriculum writers, even today.

The researchers, led by Novak, found that students who had learned about abstract science concepts, including energy and the particular nature of matter, in Years 1 and 2, had a better understanding of science in Year 12. Not only did these students hold more conventional conceptions in science than students who had not learned about science in Year 1 and 2, but they held less scientific misconceptions.

The 1960’s were a time of great debate in education. Piaget’s stages of cognitive development were challenged by Ausubel‘s assimilation theory of learning (1963), which recognised the prior knowledege of students in developing new understandings. Constructivism began to dominate the discussions of science educators. Novak wanted to know if young students could learn abstract concepts, such as the nature of matter and energy, and if such learning would influence later understandings of science. To do this, he had to undertake a 12 year longitudinal study.

What did Novak do?

Considering the progressive sequence of learning for students’ integration of new ideas to existing understandings, Novak and his students designed and field-tested a series of 28 lessons for Grade 1 students. Lessons mainly made use of audio-tapes, but also featured films and props. Many concepts were revisited in later lessons, using different contexts or resources. Novak and his students ensured that lessons began with objects and events familiar to students. It took 6 years to design and test the 28 lessons, and many of Novak’s students earned their PhDs from the research. Although this technology is now outdated, at the time of the study, audio-tutorials were considered state-of-the-art technology-aided lessons!

A representative sample of 5 schools (of 13) in Ithaca, New York, were chosen, and all Grade 1 classes were given the audio-tutorial instruction during the first year (1971), and all Year 2 classes in the second year (1972). Teachers did not play a part except to allow their students time to complete the tutorial lessons, and carried out their regular curriculum alongside the tutorials. A control group consisted of “uninstructed” students in the same classrooms with the same teachers the year following the “instructed” group (1972 and 1973). This method controlled many of the variables that could influence the outcome of the instruction, such as teacher, curriculum, setting, demographics, etc.

One at a time, students would enter an area of the classroom set up for the 20 minute lessons, and some students would repeat lessons [there’s no explanation in the research explaining which students or why; perhaps they enjoyed the lessons, or perhaps the teacher just needed the student out of the way at the time!]. Teachers reported excellent student responses to the audio-tutorials and many teachers continued to use the lessons beyond the research period. Very few of the teachers taught science in the curriculum, usually including topics such as seasons, clouds, and plant growth, but only in a descriptive manner and not including topics such as energy transformations and the particulate nature of matter. Formal science instruction did not begin until Year 7.

Novak and his students carried out interviews with the Instructed and Uninstructed students several times a year, on topics presented by the audio-tutorial lessons: matter, energy and energy transformations. As the number of interviews grew into the hundreds, it became apparent to Novak and his team that they needed a better way to record student conceptions. Looking at Ausubel’s assimiliation theory, Novak identified three ideas that were central to their thinking:

  1. the development of new meaning is built upon prior relevant concepts and propositions
  2. cognitive structure is organised hierarchically, with general, more inclusive concepts occupying higher levels in the hierarchy than specific, less inclusive concepts, which are subsumed by the more general concepts
  3. when meaningful learning occurs, relationships between concepts become more explicit, precise and better integrated with other concepts and propositions
The number of valid and invalid notions held by Instructed and Uninstructed students in grades 2, 7, 10 and 12 (Novak, 2005).
Figure 1: The number of valid and invalid notions held by Instructed and Uninstructed students in grades 2, 7, 10 and 12 (Novak, 2005).

These ideas developed into the invention of the tool we now know as the concept map. Novak and his team developed concept maps of each student’s cognitive structure, and using these, were able to follow specific changes in the student’s knowledge and understanding as s/he progressed through the grades.

Now comes the exciting bit! Using the concept maps that Novak had developed from interviews with students, the researchers tallied the valid and invalid notions (conceptions) that students held. It is clear from Figure 1 (right) that Instructed students held more valid notions (conventional conceptions) of science than Uninstructed students. Increases in the number of Instructed students’ valid notions were greater throughout high school than the Uninstructed students’s valid notions. Furthermore, the number of invalid notions (scientific misconceptions) decreased over the course of Instructed students’ education, while for the Uninstructed students, the number of misconceptions stayed relatively stable. Early childhood education in science may not prevent the formation of misconceptions, but it does aid for the deconstruction of misconceptions!

“This significant difference in performance over the years for the Instructed and Uninstructed groups led to a statistically significant interaction variance for years in school.” (Novak, 2005)

What does this mean?

Students who had developed some ideas about energy, energy transformation and the particulate nature of matter in the early years of their learning through meaningful teaching and learning (appropriate to the time of the study) were primed not only to develop a good understanding of science, but also to reduce their misconceptions of science.

The evidence also demonstrates that 6 to 8 year old students can develop sufficient understandings of basic, but abstract, science concepts that serves as a solid foundation for later science learning.

“Such remarkable results shout for replication, but to my knowledge, no one else has attempted a 12-year longitudinal study of children’s science concept development.” (Novak, 2005)

In the past, we underestimated the capability of young children to develop abstract ideas. It is my experience though, that there are still some in primary education who do not believe that early childhood students can develop abstract concepts. The Australian Curriculum: Science does not even use the term ‘energy’ in the primary years content descriptions, despite referring to heat, light, sound and electricity! It is time for us to recognise the expertise required by early childhood educators, the capacity of small children to develop abstract cognitive conceptions, and the value of a strong science program from the very first years in school.


Novak, J. (2005). Results and Implications of a 12-Year Longitudinal Study of Science Concept LearningResearch in Science Education, 35. 23-40.

Here’s an interesting article about students learning science from media before learning it at school, and how this leads to the development of misconceptions.