scan00012

among groups within groups

Step 3 -/ANOVA table

So^rce of vaĄation	r	55	MS	Fs	Expected MS
i	a-l	^^among	ss„ a-l	MS.^	a‘ + n„o;
Y-Y '		SSw|th,n	ss.^		o^2
Y-Y ^		SStoui

total

among groups within groups

Step 3 -/ANOVA table

Ąurceof Ąr/atńn j	dr	SS	MS	Fs
Y-f /	3	1807.73	602.58	5.26
Y-Y '	33	3778.00	114.48
Y-Y	36	5585.73

total

| Step 4 - Conclusion		i Step 5 - Estimation of variance 1 components (model II)
JP		3p Since sample size differs among groups, no single
■ The corresponding p-value for calculated F statistic		value of n is appropriate. We therefore use an
with [dfsmong/ dfw(thin] degrees of freedom is equal to		„average" n - this is not the arithmetic mean of the
0.0045 < 0.05.		n.'s, but is l n„ *-r-	[t.-ŁSl.
■ We reject the H0 hypothesis at p < 0.005.		^# a-l	i-t V‘ n, V. i-*-1 ^1 y
■ There is a significant added variance component		^1 [[8 + 10 + 13+6] ^8	+ 10^ł + 13^2 +6^2)
among hosts for width of scutum in larval ticks		4-1 ^1	8 + 10 + 13 + 6 J 9.009

i Step 5 - Estimation of variance		i Step 5 - Estimation of variance
components (model II)		1 components (model II)
We may estimate the variance component as follows		Percent of variation among groups
ił, MS_-MS^ 602.58-114.48		ww. 54.179
^A n„ 9.009 because		- - x 100 =-x 100 = 32.1% s^2+*i 114.48 + 54.179
^e(^ms*m)- «' +“.<*! ^E(^MS.«u.)=0^J so ^Ms— MSw-kto=,^2		The coefficient of interclass correlation rj=0.321

2