An excellent, thoughtful article, Marilia.
“Research is not an exact science” is a sentiment any scientist would subscribe to. Yet, as you point out, too many people assume quant is king and therefore infallible, despite the interpretation and “modelling” to which many quantitative research results are subjected.
I like Doug Hubbard’s definition of measurement (from his book How to Measure Anything): measurement is something that reduces uncertainty, but is unlikely to reduce uncertainty to zero. “No matter how fuzzy the measurement is, it’s still a [valid] measurement if it tells you more than you knew before.”