Previously, these categories were thought to be relatively stable traits indicative of individual differences, with links to later cognitive development. However, the research revealed that babies often move between these two categories over the timescale studied. "The literature talks about the short-looking and long-looking categories, and links to later abilities are suggested. Unusually, we looked at this longitudinally, so we were able to pick up that these categories weren't stable" says Dr Farran. "So these differences can't be indicative of differential brain development, or predictive of later abilities."
Some of the research was designed to test whether infants are able to organize visual stimuli into groups based on similar attributes: brightness, shape, and proximity. To take part in the visual grouping experiments, each baby was placed in a car seat facing a screen onto which images were projected. Overhead cameras recorded how long each infant looked at images on the screen. The infants were shown an array where the stimuli were arranged by similarity in either horizontal lines or vertical columns. For example, for grouping by shape, an array of horizontal lines (or vertical columns) made up of squares and circles was used, constructed so that shape discrimination would be needed to 'see' lines or columns in the array. To test if visual grouping had occurred, images of plain bars of horizontal lines and vertical columns were then shown. If the infants looked for a longer time at either the line or column bars on average, this would indicate an effect of the earlier lines or columns made up of squares and circles, indicating grouping.
The results showed that grouping by brightness emerges first: it was observed at two months, in line with previous observations that this ability is present in newborn babies. At four months, two further grouping abilities emerged: grouping by shape and by proximity. Proximity grouping had not been tested in infants prior to this research, and grouping by shape had previously been seen only at six or seven months.
Dr Farran argues that it's important to understand the development of low-level processes such as attention from early on in order to understand how higher-level processes such as object recognition which requires grouping by several different visual characteristics develop. But the researchers also have a further interest: having established this pattern of development in normal infants, they intend to turn their attention to what happens in infants affected by developmental disorders. The research group already has a new ESRC-funded project under way with Williams syndrome infants a condition where attention and visual perception are particularly affected.
Visual grouping research, Dr Farran argues, is essential to providing a proper starting point for the new research. "In many atypical disorders, people look at what's happening in adults and assume that the same patterns of performance would be observed in children", says Dr Farran. "But often there are quite different patterns. In cognitive development, small differences can cascade over time, so it's very important to get a grip on what's going on at the start and how it develops longitudinally, so we can see how the developmental trajectory grows."
According to Dr Farran, until recently Williams syndrome children were rarely diagnosed at an early enough stage. Now diagnosis is often earlier, plus there is a genetic test but interventions are still something for the future. "If we can find out about cognitive development from infancy onwards in today's Williams syndrome children, the next generation will benefit from this", she predicts.