difficulty level. The survey questions were repeatedly examined and refined until
every question was unambiguous and each distractor was likely to make sense to
some of the pupils. The surveys were conducted by the class teacher in normal class
time. The teacher explained to the pupils that the survey results would not affect
their course grades. They were also asked to leave a question unanswered, rather
than guess, if they did not know the answer.
In addition, a representative sample of six class members was pre-interviewed 7
days before the intervention (figure 1) to gain a more detailed understanding of
pupils’ conceptions. The class teacher was asked to select the pupils to be
interviewed on the basis that they were articulate and reflected the range of abilities
in the class. The same pupils were post-interviewed 5 days after the intervention.
The interviews explored the pupils’ mental models of the Sun–Earth–Moon system
and associated conceptions such as days, months, years, seasons, eclipses, phases of
the Moon and tides. Their ideas about models in science were also elicited. Optional
prompt questions were used when necessary to explore pupils’ ideas further. Each
interview lasted about 20 min.
Changes in the pupils’ mental models of the Sun–Earth–Moon system and
their views of the role of models were monitored by comparing their responses to
three surveys and two interviews (figure 1).
One of the authors adopted the role of participant classroom observer
(Shepardson 1997), a role that was familiar and acceptable to the pupils and
teachers. This involved observing during both whole-class phases and small-group
phases. Each of the astronomy lessons was audio-taped and different groups were
recorded on different occasions.
The teacher was interviewed before and immediately after the intervention and
after 8 of the 11 lessons in order to ascertain his views about how those lessons went
in practice.
Results
Pupils’ Sun–Earth–Moon mental models
Data about class members’ Sun–Earth–Moon mental models were generated by
interviews and a pre-, post- and post-post-survey question (Jones and Lynch 1987),
which asked the pupils to indicate which one of five Sun–Earth–Moon systems
made most sense to them. The model closest to the scientists’ model was E. The
distribution of the pupils’ responses is shown in table 3.
The limited initial support for mental models A, C and D declined to zero by
the time of the post-post-survey. Support for mental model B increased slightly
from zero initially to 10 per cent by the time of the post-post-survey, although the
change was not statistically significant. On the other hand, 90 per cent of pupils
chose mental model E, the scientists’ model, in the post-post-survey.
Findings from the interviews clarified these survey data. Prior to the
intervention, five of the interviewed pupils understood that ‘the Earth goes around
the Sun’ (Caleb)1, while one pupil held very strongly to the view that ‘the Sun and
the Moon just stay still and the Earth moves’ (Leonie) around them in a figure-ofeight
pattern. Two pupils offered reasons for their views. For example, ‘the Sun is in
the middle of the solar system so it can’t go round more than one planet at once’
(Jack).