삼원배치법 (모수모형) (반복있음)

데이터 구조

요인 $A$는 모수인자

요인 $B$는 모수인자

요인 $C$는 모수인자

$$ y_{ijkp} = \mu + a_{i} + b_{j} + c_{k} + (ab)_{ij} + (ac)_{ik} + (bc)_{jk} + (abc)_{ijk} + e_{ijkp} $$

  • $y_{ijkp}$ : $A_{i}$ 와 $B_{j}$ 그리고 $C_{k}$ 에서 얻은 $p$ 번째 측정값
  • $\mu$ : 실험전체의 모평균
  • $a_{i}$ : $A_{i}$가 주는 효과
  • $b_{j}$ : $B_{j}$가 주는 효과
  • $c_{k}$ : $C_{k}$가 주는 효과
  • $(ab)_{ij}$ : $A_{i}$와 $B_{j}$의 교호작용 효과
  • $(ac)_{ik}$ : $A_{i}$와 $C_{k}$의 교호작용 효과
  • $(bc)_{jk}$ : $B_{j}$와 $C_{k}$의 교호작용 효과
  • $(abc)_{ijk}$ : $A_{i}$와 $B_{J}$ 그리고 $C_{k}$의 교호작용 효과
  • $e_{ijkp}$ : $A_{i}$와 $B_{j}$ 그리고 $C_{k}$에서 얻은 $p$번째 측정값오차 ($e_{ijkp} \sim N(0, \sigma_{E}^{ \ 2})$이고 서로 독립)
  • $i$ : 인자 $A$의 수준 수 $( i = 1,2, \cdots ,l )$
  • $j$ : 인자 $B$의 수준 수 $( j = 1,2, \cdots ,m )$
  • $k$ : 인자 $C$의 수준 수 $( k = 1,2, \cdots ,n )$
  • $p$ : 실험의 반복 수 $( p = 1,2, \cdots ,r )$

자료의 구조

인자
$B$
인자
$C$
인자 $A$
$$A_{1}$$ $$A_{2}$$ $$\cdots$$ $$A_{l}$$
$$B_{1}$$ $$C_{1}$$ $$y_{1111}$$ $$y_{2111}$$ $$\cdots$$ $$y_{l111}$$
$$\vdots$$ $$\vdots$$ $$\vdots$$ $$\vdots$$
$$y_{111r}$$ $$y_{211r}$$ $$\cdots$$ $$y_{l11r}$$
$$C_{2}$$ $$y_{1121}$$ $$y_{2121}$$ $$\cdots$$ $$y_{l121}$$
$$\vdots$$ $$\vdots$$ $$\vdots$$ $$\vdots$$
$$y_{112r}$$ $$y_{212r}$$ $$\cdots$$ $$y_{l12r}$$
$$\vdots$$ $$\vdots$$ $$\vdots$$ $$\vdots$$
$$C_{n}$$ $$y_{11n1}$$ $$y_{21n1}$$ $$\cdots$$ $$y_{l1n1}$$
$$\vdots$$ $$\vdots$$ $$\vdots$$ $$\vdots$$
$$y_{11nr}$$ $$y_{21nr}$$ $$\cdots$$ $$y_{l1nr}$$
$$B_{2}$$ $$C_{1}$$ $$y_{1211}$$ $$y_{2211}$$ $$\cdots$$ $$y_{l211}$$
$$\vdots$$ $$\vdots$$ $$\vdots$$ $$\vdots$$
$$y_{121r}$$ $$y_{221r}$$ $$\cdots$$ $$y_{l21r}$$
$$C_{2}$$ $$y_{1221}$$ $$y_{2221}$$ $$\cdots$$ $$y_{l221}$$
$$\vdots$$ $$\vdots$$ $$\vdots$$ $$\vdots$$
$$y_{122r}$$ $$y_{222r}$$ $$\cdots$$ $$y_{l22r}$$
$$\vdots$$ $$\vdots$$ $$\vdots$$ $$\vdots$$
$$C_{n}$$ $$y_{12n1}$$ $$y_{22n1}$$ $$\cdots$$ $$y_{l2n1}$$
$$\vdots$$ $$\vdots$$ $$\vdots$$ $$\vdots$$
$$y_{12nr}$$ $$y_{22nr}$$ $$\cdots$$ $$y_{l2nr}$$
$$\vdots$$ $$\vdots$$
$$B_{m}$$ $$C_{1}$$ $$y_{1m11}$$ $$y_{2m11}$$ $$\cdots$$ $$y_{lm11}$$
$$\vdots$$ $$\vdots$$ $$\vdots$$ $$\vdots$$
$$y_{1m1r}$$ $$y_{2m1r}$$ $$\cdots$$ $$y_{lm1r}$$
$$C_{2}$$ $$y_{1m21}$$ $$y_{2m21}$$ $$\cdots$$ $$y_{lm21}$$
$$\vdots$$ $$\vdots$$ $$\vdots$$ $$\vdots$$
$$y_{1m2r}$$ $$y_{2m2r}$$ $$\cdots$$ $$y_{lm2r}$$
$$\vdots$$ $$\vdots$$ $$\vdots$$ $$\vdots$$
$$C_{n}$$ $$y_{1mn1}$$ $$y_{2mn1}$$ $$\cdots$$ $$y_{lmn1}$$
$$\vdots$$ $$\vdots$$ $$\vdots$$ $$\vdots$$
$$y_{1mnr}$$ $$y_{2mnr}$$ $$\cdots$$ $$y_{lmnr}$$

$AB$ 2원표

인자 $B$ 인자 $A$ 합계
$$A_{1}$$ $$A_{2}$$ $$\cdots$$ $$A_{l}$$
$$B_{1}$$ $$T_{11..}$$ $$T_{21..}$$ $$\cdots$$ $$T_{l1..}$$ $$T_{.1..}$$
$$B_{2}$$ $$T_{12..}$$ $$T_{22..}$$ $$\cdots$$ $$T_{l2..}$$ $$T_{.2..}$$
$$\vdots$$ $$\vdots$$ $$\vdots$$ $$\vdots$$ $$\vdots$$
$$B_{m}$$ $$T_{1m..}$$ $$T_{2m..}$$ $$\cdots$$ $$T_{lm..}$$ $$T_{.m..}$$
합계 $$T_{1...}$$ $$T_{2...}$$ $$\cdots$$ $$T_{l...}$$ $$T$$

$AC$ 2원표

인자 $C$ 인자 $A$ 합계
$$A_{1}$$ $$A_{2}$$ $$\cdots$$ $$A_{l}$$
$$C_{1}$$ $$T_{1.1.}$$ $$T_{2.1.}$$ $$\cdots$$ $$T_{l.1.}$$ $$T_{..1.}$$
$$C_{2}$$ $$T_{1.2.}$$ $$T_{2.2.}$$ $$\cdots$$ $$T_{l.2.}$$ $$T_{..2.}$$
$$\vdots$$ $$\vdots$$ $$\vdots$$ $$\vdots$$ $$\vdots$$
$$C_{n}$$ $$T_{1.n.}$$ $$T_{2.n.}$$ $$\cdots$$ $$T_{l.n.}$$ $$T_{..n.}$$
합계 $$T_{1...}$$ $$T_{2...}$$ $$\cdots$$ $$T_{l...}$$ $$T$$

$BC$ 2원표

인자 $C$ 인자 $B$ 합계
$$B_{1}$$ $$B_{2}$$ $$\cdots$$ $$B_{m}$$
$$C_{1}$$ $$T_{.11.}$$ $$T_{.21.}$$ $$\cdots$$ $$T_{.m1.}$$ $$T_{..1.}$$
$$C_{2}$$ $$T_{.12.}$$ $$T_{.22.}$$ $$\cdots$$ $$T_{.m2.}$$ $$T_{..2.}$$
$$\vdots$$ $$\vdots$$ $$\vdots$$ $$\vdots$$ $$\vdots$$
$$C_{n}$$ $$T_{.1n.}$$ $$T_{.2n.}$$ $$\cdots$$ $$T_{.mn.}$$ $$T_{..n.}$$
합계 $$T_{.1..}$$ $$T_{.2..}$$ $$\cdots$$ $$T_{.m..}$$ $$T$$
$$T_{i...} = \sum_{j=1}^{m} \sum_{k=1}^{n} \sum_{p=1}^{r} y_{ijkp}$$ $$\overline{y}_{i...} = \frac{T_{i...}}{mnr}$$
$$T_{.j..} = \sum_{i=1}^{l} \sum_{k=1}^{n} \sum_{p=1}^{r} y_{ijkp}$$ $$\overline{y}_{.j..} = \frac{T_{.j..}}{lnr}$$
$$T_{..k.} = \sum_{i=1}^{l} \sum_{j=1}^{m} \sum_{p=1}^{r} y_{ijkp}$$ $$\overline{y}_{..k.} = \frac{T_{..k.}}{lmr}$$
$$T_{ij..} = \sum_{k=1}^{n} \sum_{p=1}^{r} y_{ijkp}$$ $$\overline{y}_{ij..} = \frac{T_{ij..}}{nr}$$
$$T_{i.k.} = \sum_{j=1}^{m} \sum_{p=1}^{r} y_{ijkp}$$ $$\overline{y}_{i.k.} = \frac{T_{i.k.}}{mr}$$
$$T_{.jk.} = \sum_{i=1}^{l} \sum_{p=1}^{r} y_{ijkp}$$ $$\overline{y}_{.jk.} = \frac{T_{.jk.}}{lr}$$
$$T_{ijk.} = \sum_{p=1}^{r} y_{ijkp}$$ $$\overline{y}_{ijk.} = \frac{T_{ijk.}}{r}$$
$$T = \sum_{i=1}^{l} \sum_{j=1}^{m} \sum_{k=1}^{n} \sum_{p=1}^{r} y_{ijkp}$$ $$\overline{\overline{y}} = \frac{T}{lmnr} = \frac{T}{N}$$
$$N = lmnr$$ $$CT = \frac{T^{2}}{lmnr} = \frac{T^{2}}{N}$$

제곱합

개개의 데이터 $y_{ijkp}$와 총평균 $\overline{\overline{y}}$의 차이는 다음과 같이 8부분으로 나뉘어진다.

$$\begin{displaymath}\begin{split} (y_{ijkp}-\overline{\overline{y}}) &= (\overline{y}_{i...} - \overline{\overline{y}}) + (\overline{y}_{.j..} - \overline{\overline{y}}) + (\overline{y}_{..k.} - \overline{\overline{y}}) \\ &+ (\overline{y}_{ij..} - \overline{y}_{i...} - \overline{y}_{.j..} + \overline{\overline{y}}) + (\overline{y}_{i.k.} - \overline{y}_{i...} - \overline{y}_{..k.} + \overline{\overline{y}}) + (\overline{y}_{.jk.} - \overline{y}_{.j..} - \overline{y}_{..k.} + \overline{\overline{y}}) \\ &+ (y_{ijk.} - \overline{y}_{ij..} - \overline{y}_{i.k.} - \overline{y}_{.jk.} + \overline{y}_{i...} + \overline{y}_{.j..} + \overline{y}_{..k.} - \overline{\overline{y}}) \\ &+ (y_{ijkp}-\overline{y}_{ijk.}) \end{split}\end{displaymath}$$ 양변을 제곱한 후에 모든 $i, \ j, \ k, \ p$에 대하여 합하면 아래의 등식을 얻을 수 있다.

$$\begin{displaymath}\begin{split} \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(y_{ijkp}-\overline{\overline{y}})^{2} &= \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(\overline{y}_{i...} - \overline{\overline{y}})^{2} + \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(\overline{y}_{.j..} - \overline{\overline{y}})^{2} \\ &+ \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(\overline{y}_{..k.} - \overline{\overline{y}})^{2} \\ &+ \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(\overline{y}_{ij..} - \overline{y}_{i...} - \overline{y}_{.j..} + \overline{\overline{y}})^{2} \\ &+ \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(\overline{y}_{i.k.} - \overline{y}_{i...} - \overline{y}_{..k.} + \overline{\overline{y}})^{2} \\ &+ \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(\overline{y}_{.jk.} - \overline{y}_{.j..} - \overline{y}_{..k.} + \overline{\overline{y}})^{2} \\ &+ \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(y_{ijk.} - \overline{y}_{ij..} - \overline{y}_{i.k.} - \overline{y}_{.jk.} + \overline{y}_{i...} + \overline{y}_{.j..} + \overline{y}_{..k.} - \overline{\overline{y}})^{2} \\ &+ \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(y_{ijkp}-\overline{y}_{ijk.})^{2} \end{split}\end{displaymath}$$ 위 식에서 왼쪽 항은 총변동 $S_{T}$이고, 오른쪽 항은 차례대로 $A$의 [[변동]], $B$의 [[변동]], $C$의 [[변동]], $A, \ B$의 [[교호작용]]의 변동, $A, \ C$의 [[교호작용]]의 변동, $B, \ C$의 [[교호작용]]의 [[변동]], $A, \ B, \ C$의 [[교호작용]]의 변동, [[오차변동]]인 $S_{A}$, $S_{B}$, $S_{C}$, $S_{A \times B}$, $S_{A \times C}$, $S_{B \times C}$, $S_{A \times B \times C}$, $S_{E}$가 된다.

$$\begin{displaymath}\begin{split} S_{T} &= \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(y_{ijkp}-\overline{\overline{y}})^{2} \\ &= \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}y_{ijkp}^{ \ 2} - CT \end{split}\end{displaymath}$$ $$\begin{displaymath}\begin{split} S_{A} &= \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(y_{i...}-\overline{\overline{y}})^{2} \\ &= \sum_{i=1}^{l}\frac{T_{i...}^{ \ 2}}{mnr}-CT \end{split}\end{displaymath}$$ $$\begin{displaymath}\begin{split} S_{B} &= \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(y_{.j..}-\overline{\overline{y}})^{2} \\ &= \sum_{j=1}^{m}\frac{T_{.j..}^{ \ 2}}{lnr}-CT \end{split}\end{displaymath}$$ $$\begin{displaymath}\begin{split} S_{C} &= \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(y_{..k.}-\overline{\overline{y}})^{2} \\ &= \sum_{k=1}^{n}\frac{T_{..k.}^{ \ 2}}{lmr}-CT \end{split}\end{displaymath}$$ $$\begin{displaymath}\begin{split} S_{A \times B} &= \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(\overline{y}_{ij..}-\overline{y}_{i...}-\overline{y}_{.j..}+\overline{\overline{y}})^{2} \\ &= S_{AB} - S_{A} - S_{B} \end{split}\end{displaymath}$$ $$\begin{displaymath}\begin{split} S_{AB} &= \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(\overline{y}_{ij..}-\overline{\overline{y}})^{2} \\ &= \sum_{i=1}^{l}\sum_{j=1}^{m} \frac{T_{ij..}^{ \ 2}}{nr} -CT \end{split}\end{displaymath}$$ $$\begin{displaymath}\begin{split} S_{A \times C} &= \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(\overline{y}_{i.k.}-\overline{y}_{i...}-\overline{y}_{..k.}+\overline{\overline{y}})^{2} \\ &= S_{AC} - S_{A} - S_{C} \end{split}\end{displaymath}$$ $$\begin{displaymath}\begin{split} S_{AC} &= \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(\overline{y}_{i.k.}-\overline{\overline{y}})^{2} \\ &= \sum_{i=1}^{l}\sum_{k=1}^{n} \frac{T_{i.k.}^{ \ 2}}{mr} -CT \end{split}\end{displaymath}$$ $$\begin{displaymath}\begin{split} S_{B \times C} &= \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(\overline{y}_{.jk.}-\overline{y}_{.j..}-\overline{y}_{..k.}+\overline{\overline{y}})^{2} \\ &= S_{BC} - S_{B} - S_{C} \end{split}\end{displaymath}$$ $$\begin{displaymath}\begin{split} S_{BC} &= \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(\overline{y}_{.jk.}-\overline{\overline{y}})^{2} \\ &= \sum_{j=1}^{m}\sum_{k=1}^{n} \frac{T_{.jk.}^{ \ 2}}{lr} -CT \end{split}\end{displaymath}$$ $$\begin{displaymath}\begin{split} S_{A \times B \times C} &= \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(y_{ijk.}-\overline{y}_{ij..}-\overline{y}_{i.k.}-\overline{y}_{.jk.}+\overline{y}_{i...}+\overline{y}_{.j..}+\overline{y}_{..k.}-\overline{\overline{y}})^{2} \\ &= S_{ABC}-(S_{A}+S_{B}+S_{C}+S_{A \times B}+S_{A \times C}+S_{B \times C}) \end{split}\end{displaymath}$$ $$\begin{displaymath}\begin{split} S_{ABC} &= \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(y_{ijk.}-\overline{\overline{y}})^{2} \\ &= \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\frac{T_{ijk.}^{ \ 2}}{r} -CT \end{split}\end{displaymath}$$ $$\begin{displaymath}\begin{split} S_{E} &= \sum_{i=1}^{l}\sum_{j=1}^{m}\sum_{k=1}^{n}\sum_{p=1}^{r}(y_{ijkp}-\overline{\overline{y}})^{2} \\ &= S_{T} - S_{ABC} \end{split}\end{displaymath}$$ ===== 자유도 ===== $$\nu_{A}=l-1$$ $$\nu_{B}=m-1$$ $$\nu_{C}=n-1$$ $$\nu_{A \times B}=\nu_{A} \times \nu_{B}=(l-1)(m-1)$$ $$\nu_{A \times C}=\nu_{A} \times \nu_{C}=(l-1)(n-1)$$ $$\nu_{B \times C}=\nu_{B} \times \nu_{C}=(m-1)(n-1)$$ $$\nu_{A \times B \times C}=\nu_{A} \times \nu_{B} \times \nu_{C} =(l-1)(m-1)(n-1)$$ $$\nu_{E}=\nu_{T}-(\nu_{A}+\nu_{B}+\nu_{C}+\nu_{A \times B}+\nu_{A \times C}+\nu_{B \times C}+\nu_{A \times B \times C})=lmn(r-1)$$ $$\nu_{T}=lmnr-1=N-1$$ ===== 평균제곱 ===== $$V_{A}=\frac{S_{A}}{\nu_{A}}$$ $$V_{B}=\frac{S_{B}}{\nu_{B}}$$ $$V_{C}=\frac{S_{C}}{\nu_{C}}$$ $$V_{A \times B}=\frac{S_{A \times B}}{\nu_{A \times B}}$$ $$V_{A \times C}=\frac{S_{A \times C}}{\nu_{A \times C}}$$ $$V_{B \times C}=\frac{S_{B \times C}}{\nu_{B \times C}}$$ $$V_{A \times B \times C}=\frac{S_{A \times B \times C}}{\nu_{A \times B \times C}}$$ $$V_{E}=\frac{S_{E}}{\nu_{E}}$$ ===== 평균제곱의 기대값 ===== $$E(V_{A})=\sigma_{E}^{ \ 2} +mnr \sigma_{A}^{ \ 2}$$ $$E(V_{B})=\sigma_{E}^{ \ 2} +lnr \sigma_{B}^{ \ 2}$$ $$E(V_{C})=\sigma_{E}^{ \ 2} +lmr \sigma_{C}^{ \ 2}$$ $$E(V_{A \times B})=\sigma_{E}^{ \ 2} +nr \sigma_{A \times B}^{ \ 2}$$ $$E(V_{A \times C})=\sigma_{E}^{ \ 2} +mr \sigma_{A \times C}^{ \ 2}$$ $$E(V_{B \times C})=\sigma_{E}^{ \ 2} +lr \sigma_{A \times B}^{ \ 2}$$ $$E(V_{A \times B \times C})=\sigma_{E}^{ \ 2} +r \sigma_{A \times B \times C}^{ \ 2}$$ $$E(V_{E})=\sigma_{E}^{ \ 2}$$ ===== 분산분석표 ===== ^ [[요인]] ^ [[제곱합]]\\ $SS$ ^ [[자유도]]\\ $DF$ ^ [[평균제곱]]\\ $MS$ ^ $E(MS)$ ^ $F_{0}$ ^ [[기각치]] ^ [[순변동]]\\ $S\acute{}$ ^ [[기여율]]\\ $\rho$ |

$$A$$ $$S_{_{A}}$$ $$\nu_{_{A}}=l-1$$ $$V_{_{A}}=S_{_{A}}/\nu_{_{A}}$$ $$\sigma_{_{E}}^{ \ 2}+mnr \ \sigma_{_{A}}^{2}$$ $$V_{_{A}}/V_{_{E}}$$ $$F_{1-\alpha}(\nu_{_{A}} \ , \ \nu_{_{E}})$$ $$S_{_{A}}\acute{}$$ $$S_{_{A}}\acute{}/S_{_{T}}$$
$$B$$ $$S_{_{B}}$$ $$\nu_{_{B}}=m-1$$ $$V_{_{B}}=S_{_{B}}/\nu_{_{B}}$$ $$\sigma_{_{E}}^{ \ 2}+lnr \ \sigma_{_{B}}^{2}$$ $$V_{_{B}}/V_{_{E}}$$ $$F_{1-\alpha}(\nu_{_{B}} \ , \ \nu_{_{E}})$$ $$S_{_{B}}\acute{}$$ $$S_{_{B}}\acute{}/S_{_{T}}$$
$$C$$ $$S_{_{C}}$$ $$\nu_{_{C}}=n-1$$ $$V_{_{C}}=S_{_{C}}/\nu_{_{C}}$$ $$\sigma_{_{E}}^{ \ 2}+lmr \ \sigma_{_{C}}^{2}$$ $$V_{_{C}}/V_{_{E}}$$ $$F_{1-\alpha}(\nu_{_{C}} \ , \ \nu_{_{E}})$$ $$S_{_{C}}\acute{}$$ $$S_{_{C}}\acute{}/S_{_{T}}$$
$$A \times B$$ $$S_{_{A \times B}}$$ $$\nu_{_{A \times B}}=(l-1)(m-1)$$ $$V_{_{A \times B}}=S_{_{A \times B}}/\nu_{_{A \times B}}$$ $$\sigma_{_{E}}^{ \ 2}+nr \ \sigma_{_{A \times B}}^{2}$$ $$V_{_{A \times B}}/V_{_{E}}$$ $$F_{1-\alpha}(\nu_{_{A \times B}} \ , \ \nu_{_{E}})$$ $$S_{_{A \times B}}\acute{}$$ $$S_{_{A \times B}}\acute{}/S_{_{T}}$$
$$A \times C$$ $$S_{_{A \times C}}$$ $$\nu_{_{A \times C}}=(l-1)(n-1)$$ $$V_{_{A \times C}}=S_{_{A \times C}}/\nu_{_{A \times C}}$$ $$\sigma_{_{E}}^{ \ 2}+mr \ \sigma_{_{A \times C}}^{2}$$ $$V_{_{A \times C}}/V_{_{E}}$$ $$F_{1-\alpha}(\nu_{_{A \times C}} \ , \ \nu_{_{E}})$$ $$S_{_{A \times C}}\acute{}$$ $$S_{_{A \times C}}\acute{}/S_{_{T}}$$
$$B \times C$$ $$S_{_{B \times C}}$$ $$\nu_{_{B \times C}}=(m-1)(n-1)$$ $$V_{_{B \times C}}=S_{_{B \times C}}/\nu_{_{B \times C}}$$ $$\sigma_{_{E}}^{ \ 2}+lr \ \sigma_{_{B \times C}}^{2}$$ $$V_{_{B \times C}}/V_{_{E}}$$ $$F_{1-\alpha}(\nu_{_{B \times C}} \ , \ \nu_{_{E}})$$ $$S_{_{B \times C}}\acute{}$$ $$S_{_{B \times C}}\acute{}/S_{_{T}}$$
$$A \times B \times C$$ $$S_{_{A \times B \times C}}$$ $$\nu_{_{A \times B \times C}}=(l-1)(m-1)(n-1)$$ $$V_{_{A \times B \times C}}=S_{_{A \times B \times C}}/\nu_{_{A \times B \times C}}$$ $$\sigma_{_{E}}^{ \ 2}+r \ \sigma_{_{A \times B \times C}}^{ \ 2}$$ $$V_{_{A \times B \times C}}/V_{_{E}}$$ $$F_{1-\alpha}(\nu_{_{A \times B \times C}} \ , \ \nu_{_{E}})$$ $$S_{_{A \times B \times C}}\acute{}$$ $$S_{_{A \times B \times C}}\acute{}/S_{_{T}}$$
$$E$$ $$S_{_{E}}$$ $$\nu_{_{E}}=lmn(r-1)$$ $$V_{_{E}}=S_{_{E}}/\nu_{_{E}}$$ $$\sigma_{_{E}}^{ \ 2}$$ $$S_{_{E}}\acute{}$$ $$S_{_{E}}\acute{}/S_{_{T}}$$
$$T$$ $$S_{_{T}}$$ $$\nu_{_{T}}=lmnr-1$$ $$S_{_{T}}$$ $$1$$

분산분석

인자 $A$에 대한 분산분석

  • $$F_{0}=\frac{V_{_{A}}}{V_{_{E}}}$$

기각역 : $F_{0} > F_{1-\alpha}(\nu_{_{A}},\nu_{_{E}})$


인자 $B$에 대한 분산분석

  • $$F_{0}=\frac{V_{_{B}}}{V_{_{E}}}$$

기각역 : $F_{0} > F_{1-\alpha}(\nu_{_{B}},\nu_{_{E}})$


인자 $C$에 대한 분산분석

  • $$F_{0}=\frac{V_{_{C}}}{V_{_{E}}}$$

기각역 : $F_{0} > F_{1-\alpha}(\nu_{_{C}},\nu_{_{E}})$


인자 $A , \ B$의 교호작용에 대한 분산분석

  • $$F_{0}=\frac{V_{_{A \times B}}}{V_{E}}$$

기각역 : $F_{0} > F_{1-\alpha}(\nu_{_{A \times B}},\nu_{_{E}})$


인자 $A , \ C$의 교호작용에 대한 분산분석

  • $$F_{0}=\frac{V_{_{A \times C}}}{V_{E}}$$

기각역 : $F_{0} > F_{1-\alpha}(\nu_{_{A \times C}},\nu_{_{E}})$


인자 $B , \ C$의 교호작용에 대한 분산분석

  • $$F_{0}=\frac{V_{B \times C}}{V_{E}}$$

기각역 : $F_{0} > F_{1-\alpha}(\nu_{_{B \times C}},\nu_{_{E}})$


인자 $A , \ B , \ C$의 교호작용에 대한 분산분석

  • $$F_{0}=\frac{V_{A \times B \times C}}{V_{E}}}$$

기각역 : $F_{0} > F_{1-\alpha}(\nu_{A \times B \times C},\nu_{_{E}})$

각 수준의 모평균의 추정 (주효과만이 유의한 경우)

주효과인 인자 $A, B, C$만이 유의한 경우 교호작용들이 모두 오차항에 풀링되어 버린다.

(단, $S_{E}\acute{}=S_{E}+S_{A \times B}+S_{A \times C}+S_{B \times C}+S_{A \times B \times C}, \ \nu_{E}\acute{}=\nu_{E}+\nu_{A \times B}+\nu_{A \times C}+\nu_{B \times C}+\nu_{A \times B \times C}, \ V_{E}\acute{}=S_{E}\acute{}/\nu_{E}\acute{}$이다.)

인자 $A$의 모평균에 관한 추정

$i$ 수준에서의 모평균 $\mu(A_{i})$의 점추정

  • $$\hat{\mu}(A_{i})=\widehat{\mu + a_{i}} = \overline{y}_{i...}$$

$i$ 수준에서의 모평균 $\mu(A_{i})$의 $100(1-\alpha) \% $ 신뢰구간은 아래와 같다.

  • $$\hat{\mu}(A_{i})= \left( \overline{y}_{i...} - t_{\alpha/2}(\nu_{E}\acute{} \ ) \sqrt{\frac{V_{E}\acute{}}{mnr}} \ , \ \overline{y}_{i...} + t_{\alpha/2}(\nu_{E}\acute{} \ ) \sqrt{\frac{V_{E}\acute{}}{mnr}} \right)$$

인자 $B$의 모평균에 관한 추정

$j$ 수준에서의 모평균 $\mu(B_{j})$의 점추정

  • $$\hat{\mu}(B_{j})=\widehat{\mu + b_{j}} = \overline{y}_{.j..}$$

$j$ 수준에서의 모평균 $\mu(B_{j})$의 $100(1-\alpha) \%$ 신뢰구간은 아래와 같다.

  • $$\hat{\mu}(B_{j})= \left( \overline{y}_{.j..} - t_{\alpha/2}(\nu_{E}\acute{} \ ) \sqrt{\frac{V_{E}\acute{}}{lnr}} \ , \ \overline{y}_{.j..} + t_{\alpha/2}(\nu_{E}\acute{} \ ) \sqrt{\frac{V_{E}\acute{}}{lnr}} \right)$$

인자 $C$의 모평균에 관한 추정

$k$ 수준에서의 모평균 $\mu(C_{k})$의 점추정

  • $$\hat{\mu}(C_{k})=\widehat{\mu + c_{k}} = \overline{y}_{..k.}$$

$k$ 수준에서의 모평균 $\mu(C_{k})$의 $100(1-\alpha) \% $ 신뢰구간은 아래와 같다.

  • $$\hat{\mu}(C_{k})= \left( \overline{y}_{..k.} - t_{\alpha/2}(\nu_{E}\acute{} \ ) \sqrt{\frac{V_{E}\acute{}}{lmr}} \ , \ \overline{y}_{..k.} + t_{\alpha/2}(\nu_{E}\acute{} \ ) \sqrt{\frac{V_{E}\acute{}}{lmr}} \right)$$

인자 $A$와 $B$ 그리고 $C$의 모평균에 관한 추정

$A$ 인자의 $i$ 수준과 $B$ 인자의 $j$ 수준, $C$ 인자의 $k$ 수준에서의 모평균 $\mu(A_{i}B_{j}C_{k})$의 점추정

  • $$\hat{\mu}(A_{i}B_{j}C_{k})=\widehat{\mu+a_{i}+b_{j}+c_{k}}=\overline{y}_{i...} + \overline{y}_{.j..} + \overline{y}_{..k.} - 2 \overline{\overline{y}}$$

$A$ 인자의 $i$ 수준과 $B$ 인자의 $j$ 수준, $C$ 인자의 $k$ 수준에서의 모평균 $\mu(A_{i}B_{j}C_{k})$의 $100(1-\alpha) \% $ 신뢰구간은 아래와 같다.

  • $$\hat{\mu}(A_{i}B_{j}C_{k})= \left( (\overline{y}_{i...} + \overline{y}_{.j..} + \overline{y}_{..k.} - 2\overline{\overline{y}}) - t_{\alpha/2}(\nu_{E}\acute{} \ )\sqrt{\frac{V_{E}\acute{}}{n_{e}}} \ , \ (\overline{y}_{i...} + \overline{y}_{.j..} + \overline{y}_{..k.} - 2\overline{\overline{y}}) - t_{\alpha/2}(\nu_{E}\acute{} \ )\sqrt{\frac{V_{E}\acute{}}{n_{e}}} \right)$$

단, $n_{e}$는 유효반복수이고 $n_{e} = \frac{lmnr}{l+m+n-2}$이다.