We discuss the results of 1/f noise measurements made on films of poly
crystalline indium oxide and ZnO accumulation layers. In these systems
, the amount of static disorder can be readily and reversibly changed
by fine-tuning the stoichiometry, which gives one a convenient and uni
que method to study the sensitivity of the noise to changes in some tr
ansport parameters. We present detailed experimental evidence that rul
es out classical percolation phenomena as an explantation for the high
noise level observed in these materials. We elaborate on a qualitativ
e model given by Cohen et al. that ascribes the noise characteristics
of these systems to an impending metal-insulator transition and dwell
on some of its implications.