Hacker News Reader

Comment by spravin | Hacker News Reader

The positioning of the dropdown clearly has a lot of influence on the results. People are horrible random number generators! In this case, people are unlikely to choose the ends (1/9) or the center (5) leading to a bimodal distribution with two modes at 4 and 7.

View on HN

The number picker was random, some people got a drop down, some got a list of numbers, some got a slider etc etc etc

Note however that different types of input systems were used. Refreshes show the following different types:

  * A slider that need to be dragged to choose a value.
  * A list of numbers shown with highlighting on hover.
  * An input cursor that on hover changes to a list of numbers (within a curved border this time).
  * A text field (enclosed in a circular border), allowing manual input.

I'm quite interested in how results varied depending upon input type (possibly more interesting than seeing how they vary by referrer).

Well, why don't you have a look? :) http://nfrom1to10.appspot.com/results/

In thinking back, that's exactly why I picked 4.

I just clicked whatever was closest because I wasn't even thinking about the exercise.

Reply on news.ycombinator.com