Some debate about sense-data is done by C.W. Savage from the University of Minnesota. Sense-data are only in a relative way and to some degree certain, immediate and precise. Sensation is although a fundamentally relational occurence in which a subject is aware of an object. The critique of William James that there is no relational character of sensation is heared and followed. The issue is confronted and resolved.
Since we have acquaintance with anything of which we are directly aware, without the intermediary of any process of inference or any knowledge of truth doesn't mean that sense-data are always simple. They can also be complex, which means they interact on the ground level when striving for a certain goal or reason. In this way they come closer to the subject, and may play a role into the derived behavior.
So stated Bertrand Russel that "presentation (or acquaintance) is a two-term relation of a subject, or (better) an act, to a single (simple or complex) object, while judgment is a multiple relation of a subject or act to the several objects concerned in the judgment."
An interactive process this complex is of course prone to error. But even how they are named sensible presentations, and are given quite of truth having. There has been stated that perception is in fact synonymous with this kind of sensation. There is some kind of logic available when this frame is used to regard at them. The role of judgment has been given, as stated by Russel, prime value to this. Which is a belief that evaluation is strictly accompanied with sensation this kind. They speak of a judgment of sensation (perception).
Judgment is always based upon a complex. Therefore are sense-data used as the grounds and tests for all our rational, conscious empirical judgments. This explanation is not linear but rather interdependent and symbiotic. Pure sensation is not cognitive, it is a cause of cognition. Where the image of the sensation is called sense-datum. Although a sensation which is remembered becomes an object of consciousness as soon as it begins to be remembered. Even if you can't grasp it in consciousness. Also it is difficult to draw the boundaries between perceiver and the perceived object. Here memory can interfere.
From this moment on, a duality has to be re-introduced. You can only know when something is noticed. So are sense-data replaced by experience. With the understanding that perception plays more role in relation with the past versus the present. So there has to be stated that sensations are unvarnished experiences, facts, which cannot in themselves be mistaken, and have a kind of incorrigibility. They do not constitute knowledge, for there is no distinction between knowing of a sensation and a sensation known. Perceptions, unlike sensations, involve interpretations. But since sensation is not a cognition but rather a cause of cognition. They have a sort of independent credibility. There is although no reason to suppose they are inferentially purer. Even when personal sense is hard to refute, from a judgmental perspective.