Except it's about meaning, uncertainty, probability and possibility. The meaning of observation: to determine, to objectively measure (i.e. without involving belief or meaning). There's a problem with the "standard" view, because it's actually impossible to arrange any sort of observation, without first assigning meaning, or while observing, or after. There is no "objective moment" to sit in, and see things happen, so sorry.
Belief is, like I keep saying, a process of assignation. Meaning is involved, in order to classify, find order, file under "unexplained", or "anomalous"; we observers are bound to do this, or how else do we learn (anything)?
It's about what an observation is, and what it isn't, and it's about what meaning is and isn't.
An observation is a pattern, but it's also the representation of the pattern (a map of it). Patterns have labels and things in them that we "see", and assign meaning to (a valid result, a meaningful outcome in an experiment). We would not be able to use the technology we do--or build it--if we didn't assign meaning to it.
Belief is the assignation of meaning (status) to any observation (pattern). Not seeing anything (meaningful) is discarding the observation, or believing it has no relevant meaning.
We can believe or disbelieve our, or someone else's, explanation of why some pattern exists. Not believing something to be false is an expectation, so is believing it to be true (or the inverse). Something is possible, or it might be believed impossible (zero expectation). Something is believed because it explains an observation, possibility is due to expectation of further observation.