Thank you, that is a very clear explanation. And "construct validity", from its name and from a quick look on wikipedia is exactly my concern.
Neither. I am aware though that whole fields of research periodically go through upheavals and abandon previous methodological orthodoxies, and that happens only because there is a debate on those methodological orthodoxies. For example, if I understand correctly, experiments with rats in a maze used to be the mainstay of psychology experiments in the past, but they are now not so much.
Is my criticism really that hurtful?
Edit: I'm also aware of the fact that whole fields can be stuck in a rut and continue work that pays off now, in terms of publications, but is later dismissed. In machine learning we talk of "low-hanging fruit". It means that, you can publish a paper now that pushes the state-of-the-art a bit further but does nothing to address the limitations of whatever technique you are using. The result is noise.
What they mean by sensitivity, is " is someone who experiences acute physical, mental, or emotional responses to stimuli." This isn't very vague or abstract compared to things we measure all the time like "GDP".
>> The cited studies are claimed to have been conducted cross-culturally, but I really doubt that different cultures would really assign the same meaning to these words.
It really doesn't matter if they assign the same meaning to those words, or even have words to describe the concept. You're not measuring whether they are sensitive relative to their own meaning of the word, but if they are sensitive relative to a specific trait that psychology professors are interested in measuring. A baby has no concept of weight but I can still weigh them.
>> For instance, if I like flowers, am I "sensitive"? If I cry at weddings? If I cry at Game of Thrones? At MMA fights?
What would make you sensitive is if you are "someone who experiences acute physical, mental, or emotional responses to stimuli." Of course people are different. Of course they find different things emotionally salient.
Image we ran an experiment. And we asked two people, Person S(ensitive), and person R(ock) if they found 10 situations emotionally overwhelming. S tells you every time they were extremely overwhelmed.(Weddings, deaths, criticism, something unfair happens) R tells you they didn't feel overwhelmed at all.
Now if we ask S and R if they felt overwhelmed about an 11th scenario what are they chances they say yes? Previous studies have said S is much more likely to answer yes than R. This correlation is the thing they are measuring.
Btw the criticism you are making is called construct validity and is a something that psychology professors study and worry about all the time.