Waffle Warning
dichotomy in the universe of closed questions
a ‘closed question’ is a question that has a specific answer, answers like:
Lets suppose that in the universe of possible questions there are an infinitie number of closed questions.
What is the dichotomy in the universe of closed questions? The dichotomy is between questions that can be answered ‘no’ such as ‘Wendy, do you live in an igloo?’, and questions that can be answered ‘yes’ such as “Wendy do you live in a wooden house?” Tonight’s beer-induced Wendy-epiphany is that this dichotomy of closed questions may not be equally populated. I suspect that there are more possible questions to which the answer is ‘no’ than there are questions to which the answer is ‘yes’. This suspicion is based on the following preliminary analysis:
Take this question structure as an example: “Wendy, do you live in a [insert word here]?”
If the inserted word is a physical home-type without counting all possible insertions I am estimating that the answer is more often No than Yes.
Example physical home-type: house, bungalo, igloo, TeePee, tent, hotel, skyscraper, apartment, condominium, flat, tree, bath, lake, road
If the inserted word is some other plausible descriptor of living conditions I suspect there is still an obvious weighting towards no over yes.
Example plausible descriptor: mess, illusion, happy place, circus, bubble, dream, fantasy
If the inserted word is not plausible the answer is most likely to be no
Example not-plausible words: pin, parrot, toe-nail, bling, 43
There are more no than yes answers in the range of possible answers. People tend to produce ‘yes’ answers, it’s been studied by psychologists so that they can create and understand the results of questionnaires. Since people tend (bias?) to agree, to provide ‘yes’ answers, the tendancy has been given the fancy name of ‘acquiescence bias’.
People, not psychonlogists, use skill and prior knowledge to help raise the baseline for the production of ‘yes’ answers above that which would be predicted by either a
- model that assumes the answers produced are a proportionally representative subset of all possible answers (More ‘No’ responses), or
- counter-balanced (half no, half yes) answers approach normally used in questionnaire design to ‘control’ the bias.
Some people, and psychologists, are so cunning they minimize asking questions that can be answered no and can effectively use this acquiescence bias to move towards, and gain, a concensus. People are wonderfully clever like that; giving each other the opportunity to say yes.
I really like questions where the answer is ‘yes’, I’ll leave you with this example:
“Wendy would you like another beer?”
Waffle warning over