|
Sensory Computer Systems Forums Welcome to the Sensory Computer Systems Online Forums!
|
View previous topic :: View next topic |
Author |
Message |
Mary
Joined: Dec 21 2005 Posts: 8 Location: North America
|
Posted: Tue Jun 27, 2006 8:52 am Post subject: Q. How do I analyze a multiple sample ranking question? |
|
|
Q. How do I analyze an incomplete block multiple sample ranking question?
I have an incomplete, balanced experimental design, each consumer evaluates 5 out of 10 samples, n = 150. At the end of the questionnaire, we asked consumers to rank the five products tested. How do I analysis this data ?
Thank you very much for your help,
Mary |
|
Back to top |
|
|
Karen
Joined: May 26 2006 Posts: 3
|
Posted: Tue Jun 27, 2006 8:56 am Post subject: SIMS 2000 advanced analysis options |
|
|
(Editor: this answer focuses for incomplete blocks only. Try searching this forum on 'Ranking' for related topics)
SIMS 2000 advanced analysis reporting options provide Means, T statistics, P-Values, Significiance, and the Mean Separations. Improved reporting for Ranking Incomplete Blocks was added to SIMS 2000 during February 2008.
Typically questionnaires with Ranking attributes are presented with a Complete Block experimental design type. For complete blocks, SIMS 2000 generates a nonparametric analysis using the Friedman statistic for ranking attributes. This recommended statistic requires that all the samples ranked must be presented with a Complete Block Experimental Design.
Incomplete Block and Ranking, and Pairwise Ranking, was added to SIMS 2000, March 2008. Each Panelist sees (k) samples which are part of the total (t) samples in the test. Try to design your test with enough panelists (n) so all samples are seen the same number of times, commonly referred to as a Balanced Incomplete Block (BIB).
Pairwise Ranking is one type of an incomplete block experimental design, when total samples >=3, and presenting only two samples to each panelist, such as 3C2. For 2C2 complete block experimental designs, also available is Affective/Discrimination questionnaires using Discrimination Paired Comparison (PC) question.
For more information on the use of Ranking attributes, you may want to read 'Sensory Evaluation Techniques 3rd Edition' by Meilgaard, Civille, and Carr, pages 106, 116, 292, and 295.
*****
As an alternative you can present the ranking question as a Hedonic attribute instead, for your incomplete block design, and perform the statistical analysis manually. Use the Block Number for each panelist to identify each actual sample selection. Since the question is Hedonic, only the top Ranked sample will be identified. The Chi-Square statistic can apply here. Try searching this forum on 'Ranking' for related topics.
If you are only concerned about which sample is ranked the highest, you can present a Hedonic question and use a Chi-Square analysis. See the post in this forum How do I analyze a multi-sample preference question?
Karen
|
|
Back to top |
|
|
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2005 phpBB Group
|