They all found that fatties eat more than thinnies, but Wooley (1971, liquid meals), Adams (1973, sandwiches), Hill and McCutcheon (1975, conventional meals) and Adams, Ferguson, Stunkard and Agras (1978, sandwiches and desserts) found no difference between the intake of obese and non-obese subjects.
This confusion is reported in 'Type of test meal affects palatability and eating style' by Guy- Grand, Lehnert and Doassans, in the current issue of Appetite. Suspecting the results were affected by the type of food used, they fed seven female students of normal weight three different types of meal: conventional, semi-liquid and sandwiches, each designed to be either of low or high palatability. The subjects' eating was recorded and analysed for bite-rate, average bite duration, and length of pauses between bites.
The results showed that the more palatable a conventional meal, the more subjects will eat, and the longer they will spend eating it. For sandwiches and yoghurt, the effect was considerably less pronounced. Relatively unpalatable meals are also eaten with more frequent but smaller bites. Sandwiches are eaten in smaller bites and at a slower bite-rate.
The conclusion is that palatability affects meal-size only above a minimum threshold. So psychologists will probably do well to feed their subjects properly.