HEMACYTOMETER CELL COUNT DISTRIBUTIONS - IMPLICATIONS OF NON-POISSON BEHAVIOR

被引:70
作者
NIELSEN, LK [1 ]
SMYTH, GK [1 ]
GREENFIELD, PF [1 ]
机构
[1] UNIV QUEENSLAND,DEPT MATH,ST LUCIA,QLD 4072,AUSTRALIA
关键词
D O I
10.1021/bp00012a600
中图分类号
Q81 [生物工程学(生物技术)]; Q93 [微生物学];
学科分类号
071005 ; 0836 ; 090102 ; 100705 ;
摘要
The variance of observed cell counts using a hemacytometer was analyzed. The variance was found to be greater than expected under the normally assumed Poisson distribution because of variations in the volume of the counting area. Furthermore, counts from the two sides of the hemacytometer were found to be correlated. Total cell count on a hemacytometer was better described using a normal distribution with quadratic variance, N(lambda, lambda + sigma-lambda-2), with sigma = 0.012. Because hemacytometer cell counts are not Poisson distributed, modified counting protocols are recommended.
引用
收藏
页码:560 / 563
页数:4
相关论文
共 5 条
[1]  
ANSCOMBE FJ, 1948, BIOMETRIKA, V35, P246, DOI 10.1093/biomet/35.3-4.246
[2]  
HASTINGS NA, 1975, STAT DISTRIBUTIONS, P112
[3]  
JOHNSON NL, 1970, CONTINUOUS UNIVARIAT, V1, P181
[4]  
MCCULLAGH P, 1983, GENERALIZED LINEAR M, P25
[5]  
ROHATGI VK, 1984, STATISTICAL INFERENC, P370