Blood clotting activity (PCA) is measured for 158 Norway rats from two locations just before (baseline) and four days after injection of an anticoagulant (bromadiolone). Normally this would cause reduced blood clotting after 4 days compared to the baseline, but these rats are known to possess anticoagulent resistence to varying extent. The purpose is to relate anticoagulent resistence to gender and location and perhaps weight. Dose of injection is, however, admistered according to weight and gender.
A data frame with 158 observations on the following 6 variables.
a numeric vector
a
factor with levels Loc1 Loc2
a factor with
levels F M
a numeric vector
a numeric vector with percent blood clotting activity at baseline
a numeric vector with percent blood clotting activity on day 4
Ann-Charlotte Heiberg, project at The Royal Veterinary and
Agricultural University, 1999.
Added by Ib M. Skovgaard <ims@life.ku.dk>
data(clotting)
dim(clotting)
#> [1] 158 6
head(clotting)
#> rat locality sex weight PCA0 PCA4
#> 1 1 Loc1 F 284 78.6 73.2
#> 2 2 Loc1 F 274 65.2 67.8
#> 3 3 Loc1 F 276 78.2 91.9
#> 4 4 Loc1 F 298 62.9 76.3
#> 5 5 Loc1 F 284 55.3 53.1
#> 6 6 Loc1 F 266 77.7 80.1
day0= transform(clotting, day=0, pca=PCA0)
day4= transform(clotting, day=4, pca=PCA4)
day.both= rbind(day0,day4)
m1= lm(pca ~ rat + day*locality + day*sex, data=day.both)
anova(m1)
#> Analysis of Variance Table
#>
#> Response: pca
#> Df Sum Sq Mean Sq F value Pr(>F)
#> rat 1 3090 3090 1.5599 0.21263
#> day 1 12083 12083 6.0996 0.01406 *
#> locality 1 86364 86364 43.5985 1.757e-10 ***
#> sex 1 679 679 0.3430 0.55852
#> day:locality 1 309 309 0.1561 0.69301
#> day:sex 1 5339 5339 2.6953 0.10166
#> Residuals 309 612094 1981
#> ---
#> Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
summary(m1)
#>
#> Call:
#> lm(formula = pca ~ rat + day * locality + day * sex, data = day.both)
#>
#> Residuals:
#> Min 1Q Median 3Q Max
#> -102.075 -25.065 -5.473 12.944 307.721
#>
#> Coefficients:
#> Estimate Std. Error t value Pr(>|t|)
#> (Intercept) 78.2248 7.5826 10.316 < 2e-16 ***
#> rat 0.4522 0.2120 2.133 0.03371 *
#> day -0.8559 2.0795 -0.412 0.68093
#> localityLoc2 -55.9472 18.0586 -3.098 0.00213 **
#> sexM 14.0545 11.0268 1.275 0.20342
#> day:localityLoc2 -0.9210 2.5049 -0.368 0.71337
#> day:sexM -4.1593 2.5335 -1.642 0.10166
#> ---
#> Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
#>
#> Residual standard error: 44.51 on 309 degrees of freedom
#> Multiple R-squared: 0.1498, Adjusted R-squared: 0.1333
#> F-statistic: 9.075 on 6 and 309 DF, p-value: 3.78e-09
#>
m2= lm(pca ~ rat + day, data=day.both)
anova(m2)
#> Analysis of Variance Table
#>
#> Response: pca
#> Df Sum Sq Mean Sq F value Pr(>F)
#> rat 1 3090 3089.9 1.3723 0.24232
#> day 1 12083 12082.6 5.3660 0.02118 *
#> Residuals 313 704786 2251.7
#> ---
#> Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
## Log transformation suggested.
## Random effect of rat.
## maybe str(clotting) ; plot(clotting) ...