Stimulus Selection and Sequencing
Imagine you wanted to design an FPVS experiment to study the differences between processing living objects and non-living objects. If you included images of birds, horses, dogs, cats, etc as living objects, and then images of cars, trucks, buildings, etc as non-living objects, this would technically be a valid FPVS oddball paradigm. However, there are many things could that influence the oddball responses.
For example, living objects like animals tend to have smooth edges, curves, and may appear in natural scenes with lots of green colors. Non-living objects like cars and buildings tend to have sharp edges, straight lines, and may feature a different color scheme than the base images. As such, these low-level visual differences could be influencing the oddball responses, and it would be difficult to determine whether the oddball responses were due to the living vs non-living distinction, due to low-level visual differences between the two categories, or some combination of both.
Previous work has shown that even low-level visual differences such as this, like a slight change in vertical line orientation between base and oddball stimuli, can drastically influence oddball responses, so it is important to carefully consider the stimuli used in an FPVS experiment and to control for low-level visual differences between base and oddball stimuli.