Knowledge restructuring in the development of children's cosmologies

E. Blown, T.G.K. Bryce

Research output: Contribution to journalArticlepeer-review

33 Citations (Scopus)


The development of children's cosmologies was investigated over a 13-year period, using multi-modal, in-depth interviews with 686 children (217 boys, 227 girls from New Zealand and 129 boys, 113 girls from China), aged 2-18. Children were interviewed while they observed the apparent motion of the Sun and Moon, and other features of the Earth; drew their ideas of the shape and motion of the Earth, Moon and Sun, and the causes of daytime and night-time; then modelled them using play-dough; which led into discussion of related ideas. These interviews revealed that children's cosmologies were far richer than previously thought and surprisingly similar in developmental trends across the two cultures. There was persuasive evidence of three types of conceptual change: a long-term process (over years) similar to weak restructuring; a medium-term process (over months) akin to radical restructuring; and a dynamic form of conceptual crystallisation (often in seconds) whereby previously unconnected/conflicting concepts gel to bring new meaning to previously isolated ideas. The interview technique enabled the researchers to ascertain children's concepts from intuitive, cultural, and scientific levels. The evidence supports the argument that children have coherent cosmologies that they actively create to make sense of the world rather than fragmented, incoherent "knowledge-in-pieces".
Original languageEnglish
Pages (from-to)1411-1462
Number of pages51
JournalInternational Journal of Science Education
Issue number12
Publication statusPublished - 2006


  • astronomy
  • longitudinal studies
  • cross cultural studies
  • elementary secondary education
  • science education
  • piagetian theory
  • statistical analysis


Dive into the research topics of 'Knowledge restructuring in the development of children's cosmologies'. Together they form a unique fingerprint.

Cite this