A short aside on Predicmática


On the heels of YVPolis’s latest, just a quick note on no-name, no-track-record “pollster” Predicmática. Although this new firm is apparently not entirely a Maletín-based undertaking – good sources tell me they actually do real field polling – their presentations fail catastrophically to demonstrate even basic competence.

Besides the criminally misdesigned PowerPoint slides, the way you can tell these people are hacks is from the way they frame their questions:

I’m sorry, but if I’m teaching an Introduction to Research Methods course and a student hands in a survey with that question on it, he automatically fails the semester.

It’s a blatantly leading question, designed to elicit the answer the pollster wants to hear. If you don’t understand why that’s bad practice, you really have no business conducting a poll.

But there’s more.

Predicmática brilliantly cuts through the thicket of methodological problems involved in segmenting their sample by Socio-Economic Stratum by…just asking them!

Oh brother…

Bottom line: Faketie fake-fake with a cherry on top.