Incomplete block designs for the minimisation of order and carry-over effects in sensory analysis

被引:25
作者
Ball, RD
机构
关键词
D O I
10.1016/S0950-3293(96)00026-2
中图分类号
TS2 [食品工业];
学科分类号
0832 ;
摘要
Two important considerations for experimental designs for sensory evaluation are order and carry-over effects. Ideally each sample would be presented equally often in each order and would be preceded equally often by every other sample. Williams (1949) gives a number of designs which achieve this but only for complete block designs. We describe methods for generating incomplete block designs which are approximately, balanced for carry-oner effects. Using the cyclic structure of cyclic designs as in Davis and Hall (1969), we obtain designs balanced for order effects and nearly balanced for carry-over effects. We describe a computer intensive discrete optimisation method which can be used when a cyclic design is not possible. (C) 1997 Elsevier Science Ltd.
引用
收藏
页码:111 / 118
页数:8
相关论文
共 39 条
[1]  
AARTS EHL, 1985, PHILIPS J RES, V40, P193
[2]  
AFSARINEJAD K, 1983, BIOMETRIKA, V70, P199
[3]  
[Anonymous], [No title captured]
[4]  
BALL RD, 1991, 30 DSIR
[5]  
Becker RA, 1998, WADSWORTH BROOKSCOLE
[6]  
BOSE RC, 1954, N CAROLINA AGR EXPT, V107
[7]  
BOX GEP, 1978, STATISTICS EXPT
[8]  
BREIMAN L, 1985, J AM STAT ASSOC, V80, P580, DOI 10.2307/2288473
[9]  
CLATWORTHY WH, 1973, NBS APPLIED MATH SER, V63
[10]  
Cochran W.G. G.M. Cox., 1957, Experimental Design