I am trying to recreate an analysis that has been done by another group
(in SAS I believe). I'm stuck on one part, I think because my stats
knowledge is lacking, and while it's OT, I'm hoping someone here can
help.
Given this dataframe;
foo*<-*structure(list(OBS = structure(1:18, .Label = c("1",
"2", "3",
"4", "5", "6", "7", "8",
"9", "10", "11", "12", "13",
"14", "15",
"16", "17", "18", "19", "20",
"21", "22", "23", "24", "25",
"26",
"27", "28", "29", "30", "31",
"32", "33", "34", "35", "36",
"37",
"38", "39", "40", "41", "42",
"43", "44", "45", "46", "47",
"48",
"49", "50", "51", "52", "53",
"54"), class = "factor"), NOM =
structure(c(1L,
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L,
1L), .Label = c("0.05", "0.1", "1"), class =
"factor"), RUN =
structure(c(1L,
1L, 1L, 2L, 2L, 2L, 3L, 3L, 3L, 4L, 4L, 4L, 5L, 5L, 5L, 6L, 6L,
6L), .Label = c("1", "2", "3", "4",
"5", "6"), class = "factor"),
CALC = c(0.04989, 0.04872, 0.04544, 0.05645, 0.06516, 0.0622,
0.04868, 0.05006, 0.04746, 0.05574, 0.04442, 0.04742, 0.05508,
0.0593, 0.04898, 0.06373, 0.05537, 0.04674)), .Names = c("OBS",
"NOM", "RUN", "CALC"), row.names = c(NA, 18L),
class = "data.frame")
I want to perform an anova on CALC~RUN, and based on that calculate the
95% confidence interval. However the interval produced by the earlier
analysis is [0.04741, 0.05824]. Is there some way to calculate a
confidence interval based on an ANOVa that I'm completely missing ?
> nrow(foo)
[1] 18
> mean(foo$CALC)
[1] 0.05282444
> fooaov<-aov(CALC~RUN,data=foo)
> print(fooaov)
Call:
aov(formula = CALC ~ RUN, data = foo)
Terms:
RUN Residuals
Sum of Squares 0.0003991420 0.0003202277
Deg. of Freedom 5 12
Residual standard error: 0.005165814
Estimated effects may be unbalanced
> print(summary(fooaov))
Df Sum Sq Mean Sq F value Pr(>F)
RUN 5 0.00039914 7.9828e-05 2.9914 0.05565 .
Residuals 12 0.00032023 2.6686e-05
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05
'.' 0.1 ' ' 1
> model.tables(fooaov,type="means",se=TRUE)
Tables of means
Grand mean
0.05282444
RUN
RUN
1 2 3 4 5 6
0.04802 0.06127 0.04873 0.04919 0.05445 0.05528
Standard errors for differences of means
RUN
0.004218
replic. 3
> t.test(foo$CALC,conf.level=0.95)
One Sample t-test
data: foo$CALC
t = 34.4524, df = 17, p-value < 2.2e-16
alternative hypothesis: true mean is not equal to 0
95 percent confidence interval:
0.04958955 0.05605934
sample estimates:
mean of x
0.05282444
Thanks
Paul.
[[alternative HTML version deleted]]