Older children entering school in Scotland appear to be starting from a lower level of development than children in other countries.
Researchers commissioned by the Scottish Executive to examine the cognitive development of children entering primary also found little connection between the amount of pre-school experience and children's starting points at school - a finding which raises important questions about the effectiveness of pre-school education in Scotland.
The team from the curriculum, evaluation and management centre based at Durham University, with Mike Cowie from Aberdeen University, have made an urgent plea for further research into the differences in approach to pre-school between Scotland and the three other systems they studied - in England, Western Australia and New Zealand.
The authors of the report, Peter Tymms, Paul Jones, Christine Merrell, Brian Henderson and Mr Cowie, state: "In calling for this work, we are aware that it may be taken as an implied criticism of current provision.
But this would be a misreading of our intentions.
"The research has thrown up some puzzling findings that need further investigation. We really do not know why we have found what we have found.
"In addition to investigating aspects of cognitive development, it would be interesting to look at the personal, social and emotional development of children in relation to age of starting school."
The researchers stress that their work has not concluded that Scotland should move from the status quo in terms of the optimum age for starting school. In any case, they add, such a strategy would have to be "evidence-based".
Different approaches could be tried out in different parts of Scotland. The authors believe that "this would be a world first and could parallel the enormously important work from Tennessee on class size".
The data used came from the Performance Indicators in Primary Schools (PIPS) project, which includes early predictors of reading, phonological awareness, maths and vocabulary.
The researchers found that girls were slightly ahead of boys in early reading and phonics and vocabulary, but not at all in maths.
The authors stated: "Little connection was found between the amount of pre-school experience and children's starting points at school. Perhaps this was because most children had spent a considerable period in a pre-school setting, but that does not appear to be the full explanation."
They did find, however, that the amount of pre-school experience was positively related to the starting point in mathematics, but to a very small extent.
In terms of comparing levels of development in Scotland against those of pupils in England, Western Australia and New Zealand, the research found considerable similarities in early maths. However, the mathematical development of children aged five-and-a-quarter to five-and-a-half appeared to be a little behind those in Western Australia of similar ages.
"For reading, the children in Scotland were generally a little behind those in other countries age for age, except for those who were just four-and-a-half years old."
Comparisons with New Zealand showed that youngsters there were considerably ahead. After taking age into account, pupils in Scotland could identify fewer letters of the alphabet and had lower scores on the "concepts about print" section of the assessment than pupils from other countries. But the proportion of unusual children who were already reading sentences and doing other work was just the same as elsewhere.
A spokeswoman for the Executive said: "We have seen the research and are not planning to make any change to the age at which children start school."
Of the report's findings on the weak link between pre-school education and cognitive abilities at the start of primary, she said. "We have a much softer curriculum in pre-school and are not sure how fair that comparison was. We think that pre-school education in Scotland does give children important developmental opportunities."