Conversation

I made a really really long survey and instructed people not to try finishing it (though some ppl did in order to defy me) People, on average, completed roughly 175 questions; Idk what this means but it might be some good indication for survey exhaustion limits.

135

Replying to

You instructed us to give up when we got bored. A histogram of qs completed would be good?

Aella

@Aella_Girl

Replying to

@allyrooney

they were randomized tho? or maybe i dont understand your suggestion

10:54 PM · Jan 14, 2023

829

Views

Replying to

and

the implication would be that some questions are more likely to cause someone to leave the survey?

Replying to

and

I don’t know what he wants, but his request makes me wonder if the people who completed more questions are more or less like each other. And it would be nice to see the modal range of number of questions competed.

Replying to

Sorry was opaque - what’s the distribution of how many qs each participant completed. Histo with ‘n of Qs answered’ on the X axis and ‘n of sample’ on Y.. more granularity than mean completion (and might show interesting subgroups tho randomisation complicates interp.)

Replying to

and

All three responses to this question so far are good further pointers.