UA MATH571B Things to know about statistical model of experimental design

Experimental with one treatment factor

Simple Experiment (Randomized Complete Design)

Balanced

Statistical Model is
$$y_{ij}=\mu +{\tau }_i+{ϵ}_{ij},i=1,\cdots ,a,j=1,\cdots ,n$$

$\mu$ is the grand mean, ${\tau }_i$ is the treatment effect of factor level $i$.

Assumptions:

1. Fixed Effect: ${\sum }_{i=1}^a{\tau }_i=0$
2. Homogenous normality: ${ϵ}_{ij}{\sim }_{iid}N(0,{\sigma }^2)$

ANOVA table is ($N=na$)

Hypothesises of ANOVA F test is
$$\begin{gathered} H_0:{\tau }_1=\cdots ={\tau }_a=0 \\ {H}_a:atleastoneisnotzero \end{gathered}$$

And the rejection rule is $F_0>F_{1-\alpha ,a-1,N-a}$.

OLS estimators for the parameters are

Unbalanced

Statistical Model is
$$y_{ij}=\mu +{\tau }_i+{ϵ}_{ij},i=1,\cdots ,a,j=1,\cdots ,n_i$$

$\mu$ is the grand mean, ${\tau }_i$ is the treatment effect of factor level $i$.

Assumptions:

1. Fixed Effect: ${\sum }_{i=1}^a{\tau }_i=0$
2. Homogenous normality: ${ϵ}_{ij}{\sim }_{iid}N(0,{\sigma }^2)$

RCBD

$$\begin{gathered} y_{ij}=\mu +{\tau }_i+{\beta }_j+{ϵ}_{ij} \\ {ϵ}_{ij}{\sim }_{iid}N(0,{\sigma }^2);i=1,\cdots ,a;j=1,\cdots ,b \end{gathered}$$

where ${\tau }_i$ means treatment effect and ${\beta }_j$ means blocking effect.

Assumptions:

1. Fixed Effect: ${\sum }_{i=1}^a{\tau }_i=0,{\sum }_{i=1}^b{\beta }_j=0$
2. Homogenous normality: ${ϵ}_{ij}{\sim }_{iid}N(0,{\sigma }^2)$

ANOVA table is

Source	SS	df	MS	F
Treatment	$SSM$	a-1	$MSM=\frac{SSM}{d{f}_M}$	$F_M=MSM/MSE$
Blocking	$SSB$	b-1	$MSB=\frac{SSB}{d{f}_B}$	$F_B=MSB/MSE$
Residuals	$SSE$	(a-1)(b-1)	$MSE=\frac{SSE}{d{f}_E}$
Total	$SST$	N-1	$MST=\frac{SST}{d{f}_T}$

$$\begin{gathered} \sum _{i=1}^a\sum _{j=1}^b(y_{ij}+{\bar{y}}_{..}-{\bar{y}}_{i.}-{\bar{y}}_{.j}{)}^2=\sum _{i=1}^a\sum _{j=1}^b{e}_{ij}^2=SSE \\ \sum _{i=1}^a\sum _{j=1}^b({\bar{y}}_{.j}-{\bar{y}}_{..}{)}^2=\sum _{i=1}^a\sum _{j=1}^b{\hat{\beta }}_j^2=SSB \\ \sum _{i=1}^a\sum _{j=1}^b({\bar{y}}_{i.}-{\bar{y}}_{..}{)}^2=\sum _{i=1}^a\sum _{j=1}^b{\hat{\tau }}_i^2=SSM \\ SST=SSM+SSB+SSE \end{gathered}$$

Parameter estimation:
$$\begin{gathered} y_{i.}=\sum _{j=1}^b{y}_{ij},{\bar{y}}_{i.}=\frac{y_{i.}}{b} \\ {y}_{.j}=\sum _{i=1}^a{y}_{ij},{\bar{y}}_{.j}=\frac{y_{.j}}{a} \\ {y}_{..}=\sum _{i=1}^a{y}_{i.}=\sum _{j=1}^b{y}_{.j},{\bar{y}}_{..}=\frac{y_{..}}{N},N=ab \\ {\hat{\tau }}_i={\bar{y}}_{i.}-{\bar{y}}_{..},{\hat{\beta }}_j={\bar{y}}_{.j}-{\bar{y}}_{.j},e_{ij}=y_{ij}-{\bar{y}}_{i.}-{\bar{y}}_{.j}+{\bar{y}}_{..} \end{gathered}$$

Latin Square Design

Graeco-Latin Square Design

BIBD

Experimental with multiple treatment factors

Factorial design with fixed factors

Factorial design with random factors

Factorial design with mixed factors

Nested design

Split-plot design