<p>Preference is a highly individualized thing. A website can’t capture it despite them coming with Harvard as number 1 as evidence of the methodology being perfect.</p>
<p>Self reported method does not derive this correctly. Only non-biased polling can.</p>