John
Joined: Jun 13 2005 Posts: 23
|
Posted: Wed Jun 22, 2005 3:11 pm Post subject: Triangle Tests and Rotated Blinding Codes |
|
|
Triangle Tests and Rotated Blinding Codes
Your Blinding Codes are already rotated if your Triangle Experimental Design is *not* set for Same Blinding Codes For All Panelists, so that all panelists get presented different and random blinding codes. This is the common way to eliminate risk of any order bias amounst panelists.
If you prefer Same Blinding Codes For All Panelists, where all panelists get presented the same blinding codes, SIMS 2000 assumes that order bias is not a concern. 1stA 2ndA 1stB 2ndB, each has it's own Blinding Code for all panelists. 1stA and 1stB is always presented, along with only one of the 2nds.
Same Blinding Codes For All Panelists works well in most all studies, except in rare cases where the two samples have some other variable, such as room location. Example: Fragrance scents, 4 seperate rooms, each panelist visits three rooms. When using Same Blinding Codes For All Panelists, a problem you may observe is that the 1st rooms were always seen, and the 2nd ones less.
Recommended solution: Uncheck Same Blinding Codes For All Panelists, Export (button) all the codes into Excel, re-rotate the 1st 8 Blinding codes as desired, then copy-n-paste these 8 cells over all the remain cells. Return to SIMS 2000 and Import (button) your new grid. It's easy, only takes a minute.
More Information
http://www.sims2000.com/triangle
Sensory Computer Systems |
|