Everyone hyped on surveys or polling should actually read Rensis Likert's PhD dissertation in which he describes the method we've reduced to "the Likert scale".
Here, I googled it for you:
legacy.voteview.com/pdf/Likert_193…
Here, I googled it for you:
legacy.voteview.com/pdf/Likert_193…
Oh, I should add content warning: racism.
Not his so much as the attitudes he is studying—although he does use the 1932 vernacular.
Not his so much as the attitudes he is studying—although he does use the 1932 vernacular.
I shudder to think of giving the same survey in today's America.
"It cannot be sufficiently emphasized that the reactions on an attitude test are no more meaningful than the situation in which the attitude test was given."—Rensis Likert
Context is everything.
Context is everything.
"The danger of not having the full cooperation of the subject cannot be overemphasized in the present promiscuous use of attitude tests."
Right there with me on the danger of surveys.
Right there with me on the danger of surveys.
"In any discussion of the validity of attitude scales of the
kind presented here it might be well to emphasize that at present we are dealing only with verbal behavior and claim nothing more than the importance of the verbal reactions."
No claim on predicting other behavior.
kind presented here it might be well to emphasize that at present we are dealing only with verbal behavior and claim nothing more than the importance of the verbal reactions."
No claim on predicting other behavior.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
