6 replies to this topic

[DonRocks]

Del Ray Dreamery has chosen "That Custard Place In Del Ray" as their new name (click here for details). What do you think of the new name?

Instead of a simple yes or no, there are ten choices. The square root of the arithmetic mean of the squares of the deviations from the arithmetic mean (i.e., the standard deviation) will reveal more than the mode (most-selected response).

Cheers,
Rocks

[CrescentFresh]

Instead of a simple yes or no, there are ten choices.  The square root of the arithmetic mean of the squares of the deviations from the arithmetic mean (i.e., the standard deviation) will reveal more than the mode (most-selected response).

Cheers,
Rocks

Who you callin' a deviant?

And, Rocks, some of the categories seem too close. Is "pretty sucky" more like a 2 or a 3?

Edited by CrescentFresh, 24 October 2005 - 04:12 PM.

[Tweaked]

This here's the darndest naming contversy we've had round these parts since the "daSto" incident of '04...

[JPW]

Del Ray Dreamery has chosen "That Custard Place In Del Ray" as their new name (click here for details).  What do you think of the new name?

Instead of a simple yes or no, there are ten choices.  The square root of the arithmetic mean of the squares of the deviations from the arithmetic mean (i.e., the standard deviation) will reveal more than the mode (most-selected response).

Cheers,
Rocks

Really it depends on what you are looking for. It also relies upon the assumption of a relatively normal distribution. Standard Deviations are also not entirely reliable with bounded/censored data or discrete scales. However, with a 10 point scale you may be able to get away with it, assuming of course that there are enough responses to nudge things towards that normal distribution.

Cheers.

Your friendly neighborhood statistician.

Edit to add -- In addition, preference questions are more reliable when performed on scales with between 4 and 6 options. Debate goes back and forth over which performs the best. My personal preference is 5 as I think that it is good to give respondents a completely neutral option. The argument against this is that it can open the door to the scale capturing a second dimension of salience -- i. e. instead of leaving the response blank due to a lack of information the respondent will merely respond with the neutral value, hence skewing the results.

Edited by JPW, 25 October 2005 - 07:12 AM.

[Josh Radigan]

thats del ray for ya! I like the name, and for a place that has picnic tables and childrens art on the walls, its a match.

[tripewriter]

You'd think for October, they would have taken this opportunity to go with an interim seasonal name: Del Ray Screamery.

[tenunda]

You'd think for October, they would have taken this opportunity to go with an interim seasonal name: Del Ray Screamery.

You think, for Del Ray, they would have taken this opporunity to go with something that's not completely unoriginal and painfully boring.