7-2 factorial experiments a significant interaction mask the significance of main effects....

46

Upload: rolf-everett-jacobs

Post on 19-Jan-2016

219 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects
Page 2: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

7-2 Factorial Experiments

Page 3: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

7-2 Factorial Experiments

A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects of the factors involved in the interaction may not have much meaning.

Page 4: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

7-7 Factorial Experiments with More than Two Levels

Page 5: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Two-Way Factorial Model

where

= A main effect, i = 1, = B main effect, i = 1,

= Interaction effect= Error term, which is assumed to be normally

distributed with constant variance

HypothesisA main Effect Hypothesis:

B main Effect Hypothesis:

7-7 Factorial Experiments with More than Two Levels

Page 6: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Interaction

Notice that the main effects are in terms of marginal means (means average over the other factor). It makes sense to do this only if the relationship between the means of one factor are the same for all levels of the other factor. If this is true, the factors are said not to interact. Before we can interpret the main effect tests, we must verify that the factors do not interact. That is test

Two-way Interaction test:

A graphical means of assessing interaction is to make an interaction (profile) plot. This consists plotting one of the factors along the horizontal axis and the the vertical axis. The points corresponding to the same level of the other factor are connected by a line. No interaction implies that the lines will be parallel.

7-7 Factorial Experiments with More than Two Levels

Page 7: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

7-7 Factorial Experiments with More than Two Levels

Page 8: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

7-7 Factorial Experiments with More than Two Levels

Sum of Squares partition:

Degrees of freedom partition:

Page 9: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

7-7 Factorial Experiments with More than Two Levels

Mean Squares:

Page 10: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

𝑆𝑆𝑀𝑂𝐷=𝑆𝑆𝐴+𝑆𝑆𝐵+𝑆𝑆𝐴𝐵 h𝑤𝑖𝑡 𝑑𝑓 =𝑎𝑏−1

7-7 Factorial Experiments with More than Two Levels

Page 11: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Multiple comparison and contrasts follow the same formulas as in the one-way ANOVA. The difference is that the comparisons are made on the marginal means for factors A & B. The ni are replaced by the number of observations used in calculating the sample mean. Also, they are only meaningful if there is no interaction.

7-7 Factorial Experiments with More than Two Levels

Page 12: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

7-7 Factorial Experiments with More than Two Levels

Page 13: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

7-7 Factorial Experiments with More than Two Levels

Page 14: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

7-7 Factorial Experiments with More than Two Levels

Page 15: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

7-7 Factorial Experiments with More than Two Levels

Page 16: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Model Adequacy

7-7 Factorial Experiments with More than Two Levels

Page 17: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Model Adequacy

7-7 Factorial Experiments with More than Two Levels

Page 18: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Model Adequacy

7-7 Factorial Experiments with More than Two Levels

Page 19: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Computer Output

7-7 Factorial Experiments with More than Two Levels

Page 20: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Example 7-11

7-7 Factorial Experiments with More than Two Levels

OPTIONS NOOVP NODATE NONUMBER LS=80;DATA ex711;DO obs= 1 to 3; DO type=1 to 3; DO method='Dipping', 'Spraying'; INPUT force @@; OUTPUT; END; END;END;CARDS;4 5.4 5.6 5.8 3.8 5.5 4.5 4.9 4.96.1 3.7 5 4.3 5.6 5.4 6.3 4 5ods graphics on;PROC GLM DATA=ex711 plots=all; CLASS type method; MODEL force= type method type*method; MEANS type method type*method/snk; OUTPUT out=new r=resid;TITLE 'Two-way ANOVA';PROC PLOT DATA=NEW; PLOT RESID*TYPE; PLOT RESID*METHOD;run; QUIT;

Page 21: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Two-way ANOVA The GLM Procedure

Dependent Variable: force Sum of Source DF Squares Mean Square F Value Pr > F Model 5 9.73111111 1.94622222 23.67 <.0001 Error 12 0.98666667 0.08222222 Corrected Total 17 10.71777778

R-Square Coeff Var Root MSE force Mean 0.907941 5.747656 0.286744 4.988889

Source DF Type I SS Mean Square F Value Pr > F type 2 4.58111111 2.29055556 27.86 <.0001 method 1 4.90888889 4.90888889 59.70 <.0001 type*method 2 0.24111111 0.12055556 1.47 0.2693

Source DF Type III SS Mean Square F Value Pr > F type 2 4.58111111 2.29055556 27.86 <.0001 method 1 4.90888889 4.90888889 59.70 <.0001 type*method 2 0.24111111 0.12055556 1.47 0.2693

7-7 Factorial Experiments with More than Two Levels

Page 22: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

7-7 Factorial Experiments with More than Two Levels

Page 23: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

7-7 Factorial Experiments with More than Two Levels

Page 24: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

7-7 Factorial Experiments with More than Two Levels

Page 25: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

7-7 Factorial Experiments with More than Two Levels

Page 26: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Two-way ANOVA The GLM Procedure

Student-Newman-Keuls Test for force NOTE: This test controls the Type I experimentwise error rate under the complete null hypothesis but not under partial null hypotheses.

Alpha 0.05 Error Degrees of Freedom 12 Error Mean Square 0.082222

Number of Means 2 3 Critical Range 0.3607064 0.4416697

Means with the same letter are not significantly different.

SNK Grouping Mean N type A 5.6833 6 2 B 4.7833 6 1 B 4.5000 6 3

Two-way ANOVA The GLM Procedure

Student-Newman-Keuls Test for force NOTE: This test controls the Type I experimentwise error rate under the complete null hypothesis but not under partial null hypotheses.

Alpha 0.05 Error Degrees of Freedom 12 Error Mean Square 0.082222

Number of Means 2 Critical Range 0.2945156

Means with the same letter are not significantly different.

SNK Grouping Mean N method A 5.5111 9 Sprayin B 4.4667 9 Dipping

7-7 Factorial Experiments with More than Two Levels

Page 27: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

7-7 Factorial Experiments with More than Two Levels

Page 28: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Two-way ANOVA

The GLM Procedure

Level of Level of ------------force------------ type method N Mean Std Dev

1 Dipping 3 4.26666667 0.25166115 1 Sprayin 3 5.30000000 0.36055513 2 Dipping 3 5.30000000 0.36055513 2 Sprayin 3 6.06666667 0.25166115 3 Dipping 3 3.83333333 0.15275252 3 Sprayin 3 5.16666667 0.28867513

7-7 Factorial Experiments with More than Two Levels

Page 29: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Residual Plot

resid*type 도표 . 범례 : A = 1 관측치 , B = 2 관측치 , 등 .

resid |0.35 + | A0.30 +A A |0.25 + | A A0.20 + | A 0.15 + | 0.10 +A A | 0.05 + | A A0.00 + | A-0.05 + |-0.10 + | A-0.15 + | B-0.20 + |-0.25 + | A A-0.30 + |-0.35 + |-0.40 +A A | -+-----------------------------------+-----------------------------------+ 1 2 3

type

Residual Plot

resid*method 도표 . 범례 : A = 1 관측치 , B = 2 관측치 , 등 .

resid | 0.35 + | A 0.30 + A A | 0.25 + | A A 0.20 + | A 0.15 + | 0.10 + A A | 0.05 + | A A 0.00 + | A -0.05 + | -0.10 + | A -0.15 + | B -0.20 + | -0.25 + | A A -0.30 + | -0.35 + | -0.40 + A A | ---+------------------------------------------------+-- Dipping Sprayin

method

7-7 Factorial Experiments with More than Two Levels

Page 30: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Speed

Material Feed 100 220 475 715 870

B10 0.004 122110

10885

10860

6650

8060

0.008 332330

276310

248295

248275

276310

0.014 640500

612500

543450

612610

696610

V10 0.004 192170

136130

12285

10875

13675

0.008 386365

333330

318330

473350

499390

0.014 810725

779670

810750

893890

1820890

Thrust Forces in Drilling (3-Way Factorial)

7-7 Factorial Experiments with More than Two Levels

Page 31: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

OPTIONS NOOVP NODATE NONUMBER LS=80;DATA threeway;INFILE 'C:\Users\user\Documents\Teaching\ 학부과목 \imen214-stats\ch07\sas\threeway.txt';INPUT material$ feed speed thrust @@;ods graphics on;PROC glm data=threeway plots=(diagnostics); CLASS material feed speed; MODEL thrust = material | feed | speed; MEANS material | feed | speed/snk;TITLE 'Three-way ANOVA';

DATA means1; INPUT b10 v10 feed @@; CARDS; 84.9 122.9 0.004 290 377.3 0.008 577.3 903.7 0.014 PROC PLOT data=means1; PLOT b10*feed='B' v10*feed='V'/overlay;TITLE 'Interaction Plot for Material*Feed';

DATA means2; INPUT b10 v10 speed @@; CARDS;339 441.3 100 315.2 396.3 220 284 402.5 475 310.2 464.7 715 338.7 635 870PROC PLOT data=means2; PLOT b10*speed='B' v10*speed='V'/overlay;TITLE 'Interaction Plot for Speed*Material';

DATA means3; INPUT fd1 fd2 fd3 speed @@; CARDS;148.5 353.3 668.3 100 114.8 312.3 640.3 220 93.8 297.8 638.3 47574.8 336.3 751.3 715 87.8 368.8 1004 870PROC PLOT data=means3; PLOT fd1*speed='1' fd2*speed='2' fd3*speed='3'/overlay;TITLE 'Interaction Plot for Feed*Speed';RUN; QUIT;

7-7 Factorial Experiments with More than Two Levels

B10 .004 100 122 B10 .004 100 110 B10 .004 220 108 B10 .004 220 85 B10 .004 475 108 B10 .004 475 60 B10 .004 715 66 B10 .004 715 50B10 .004 870 80 B10 .004 870 60 B10 .008 100 332 B10 .008 100 330

Page 32: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Three-way ANOVA The ANOVA ProcedureDependent Variable: thrust Sum of Source DF Squares Mean Square F Value Pr > F Model 29 5343804.483 184269.120 11.30 <.0001 Error 30 489376.500 16312.550 Corrected Total 59 5833180.983

R-Square Coeff Var Root MSE thrust Mean 0.916105 32.52509 127.7206 392.6833

Source DF Anova SS Mean Square F Value Pr > F material 1 340054.817 340054.817 20.85 <.0001 feed 2 4157143.633 2078571.817 127.42 <.0001 material*feed 2 237956.433 118978.217 7.29 0.0026 speed 4 152472.233 38118.058 2.34 0.0781 material*speed 4 88303.433 22075.858 1.35 0.2735 feed*speed 8 255092.367 31886.546 1.95 0.0880 material*feed*speed 8 112781.567 14097.696 0.86 0.5565

Three-way ANOVA The ANOVA Procedure Class Level Information Class Levels Values material 2 B10 V10 feed 3 0.004 0.008 0.014 speed 5 100 220 475 715 870

Number of Observations Read 60 Number of Observations Used 60-------------------------------------------------------------------------------

7-7 Factorial Experiments with More than Two Levels

Page 33: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

7-7 Factorial Experiments with More than Two Levels

Page 34: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Three-way ANOVA The ANOVA Procedure

Student-Newman-Keuls Test for thrust

NOTE: This test controls the Type I experimentwise error rate under the complete null hypothesis but not under partial null hypotheses.

Alpha 0.05 Error Degrees of Freedom 30 Error Mean Square 16312.55

Number of Means 2 Critical Range 67.348663

Means with the same letter are not significantly different.

SNK Grouping Mean N material A 467.97 30 V10 B 317.40 30 B10

Three-way ANOVA The ANOVA Procedure

Student-Newman-Keuls Test for thrust

NOTE: This test controls the Type I experimentwise error rate under the complete null hypothesis but not under partial null hypotheses.

Alpha 0.05 Error Degrees of Freedom 30 Error Mean Square 16312.55

Number of Means 2 3 Critical Range 82.48493 99.569343

Means with the same letter are not significantly different.

SNK Grouping Mean N feed A 740.50 20 0.014 B 333.65 20 0.008 C 103.90 20 0.004

7-7 Factorial Experiments with More than Two Levels

Page 35: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Three-way ANOVA The ANOVA Procedure

Level of Level of ------------thrust----------- material feed N Mean Std Dev B10 0.004 10 84.900000 25.631794 B10 0.008 10 290.000000 30.430248 B10 0.014 10 577.300000 75.938645 V10 0.004 10 122.900000 38.922858 V10 0.008 10 377.300000 62.241555 V10 0.014 10 903.700000 330.551408-----------------------------------------------------------------------------

Three-way ANOVA The ANOVA Procedure

Student-Newman-Keuls Test for thrust NOTE: This test controls the Type I experimentwise error rate under the complete null hypothesis but not under partial null hypotheses.

Alpha 0.05 Error Degrees of Freedom 30 Error Mean Square 16312.55

Number of Means 2 3 4 5 Critical Range 106.48759 128.54347 141.77902 151.24267

Means with the same letter are not significantly different.

SNK Grouping Mean N speed A 486.83 12 870 A 390.17 12 100 A 387.42 12 715 A 355.75 12 220 A 343.25 12 475

7-7 Factorial Experiments with More than Two Levels

Page 36: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Three-way ANOVA The ANOVA Procedure

Level of Level of ------------thrust----------- material speed N Mean Std Dev

B10 100 6 339.000000 207.933643 B10 220 6 315.166667 209.632456 B10 475 6 284.000000 188.258333 B10 715 6 310.166667 250.363269 B10 870 6 338.666667 264.809869 V10 100 6 441.333333 268.722658 V10 220 6 396.333333 271.457302 V10 475 6 402.500000 309.417356 V10 715 6 464.666667 362.389661 V10 870 6 635.000000 649.672533

Level of Level of ------------thrust----------- feed speed N Mean Std Dev

0.004 100 4 148.50000 38.897301 0.004 220 4 114.75000 23.200216 0.004 475 4 93.75000 27.183022 0.004 715 4 74.75000 24.459150 0.004 870 4 87.75000 33.270357 0.008 100 4 353.25000 27.097048 0.008 220 4 312.25000 26.234519 0.008 475 4 297.75000 36.206583 0.008 715 4 336.25000 100.260910 0.008 870 4 368.75000 99.113992 0.014 100 4 668.75000 132.185161 0.014 220 4 640.25000 116.339661 0.014 475 4 638.25000 169.800618 0.014 715 4 751.25000 161.953440 0.014 870 4 1004.00000 556.462637

Level of Level of Level of ------------thrust-----------

material feed speed N Mean Std Dev

B10 0.004 100 2 116.00000 8.485281 B10 0.004 220 2 96.50000 16.263456 B10 0.004 475 2 84.00000 33.941125 B10 0.004 715 2 58.00000 11.313708 B10 0.004 870 2 70.00000 14.142136 B10 0.008 100 2 331.00000 1.414214 B10 0.008 220 2 293.00000 24.041631 B10 0.008 475 2 271.50000 33.234019 B10 0.008 715 2 261.50000 19.091883 B10 0.008 870 2 293.00000 24.041631 B10 0.014 100 2 570.00000 98.994949 B10 0.014 220 2 556.00000 79.195959 B10 0.014 475 2 496.50000 65.760931 B10 0.014 715 2 611.00000 1.414214 B10 0.014 870 2 653.00000 60.811183 V10 0.004 100 2 181.00000 15.556349 V10 0.004 220 2 133.00000 4.242641 V10 0.004 475 2 103.50000 26.162951 V10 0.004 715 2 91.50000 23.334524 V10 0.004 870 2 105.50000 43.133514 V10 0.008 100 2 375.50000 14.849242 V10 0.008 220 2 331.50000 2.121320 V10 0.008 475 2 324.00000 8.485281 V10 0.008 715 2 411.00000 86.267027 V10 0.008 870 2 444.50000 77.074639 V10 0.014 100 2 767.50000 60.104076 V10 0.014 220 2 724.50000 77.074639 V10 0.014 475 2 780.00000 42.426407 V10 0.014 715 2 891.50000 2.121320 V10 0.014 870 2 1355.00000 657.609307

7-7 Factorial Experiments with More than Two Levels

Page 37: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Interaction Plot for Material*Feed b10*feed 도표 . 사용된 기호 : 'B'. v10*feed 도표 . 사용된 기호 : 'V'.b10 |900 + V | | |800 + | | |700 + | | |600 + | B | |500 + | | |400 + | V | |300 + | B | |200 + | | | V100 + | B | | 0 + | ---+-------------+-------------+-------------+-------------+-------------+-- 0.004 0.006 0.008 0.010 0.012 0.014

feed

Interaction Plot for Speed*Material b10*speed 도표 . 사용된 기호 : 'B'. v10*speed 도표 . 사용된 기호 : 'V'.b10 |650 + | V | |600 + | | |550 + | | |500 + | | | V |450 + | V | | |400 + V | V | | |350 + | B B | | B | B300 + | | B | |250 + ---+-------------+-------------+-------------+-------------+-------------+-- 0 200 400 600 800 1000

speed

Interaction Plot for Feed*Speed fd1*speed 도표 . 사용된 기호 : '1'. fd2*speed 도표 . 사용된 기호 : '2'. fd3*speed 도표 . 사용된 기호 : '3'.

fd1 | |1000 + 3 | | 900 + | | 800 + | | 3700 + | 3 | 3 3600 + | | 500 + | | 400 + | 2 2 | 2 | 2 300 + 2 | | 200 + | 1 | 1 100 + 1 | 1 1 | | | 0 + | --+-------------+-------------+-------------+-------------+-------------+-- 0 200 400 600 800 1000

speed

7-7 Factorial Experiments with More than Two Levels

Page 38: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

The Latin square design is used to eliminate two nuisance sources of variability; that is, it systemically allows blocking two dimensions. Thus, the rows and columns actually represent two restrictions on randomization. In a Latin square design, there are p treatments and p levels of each of the two blocking variables. Each treatment level appears in each row and column once. The arrangement should be randomly selected from all possible arrangement. For instance, there is only 1 3x3, 4 4x4, 56 5x5, and 9408 6x6 Latin squares.

The Latin Square Design

Page 39: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

The Latin Square Design

Page 40: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

OPTIONS NOOVP NODATE NONUMBER LS=80;DATA latin;INPUT operator batch treat$ force @@;CARDS;1 1 A -1 1 2 B -8 1 3 C -7 1 4 D 1 1 5 E -32 1 B -5 2 2 C -1 2 3 D 13 2 4 E 6 2 5 A 53 1 C -6 3 2 D 5 3 3 E 1 3 4 A 1 3 5 B -54 1 D -1 4 2 E 2 4 3 A 2 4 4 B -2 4 5 C 4 5 1 E -1 5 2 A 11 5 3 B -4 5 4 C -3 5 5 D 6ods graphics on;PROC GLM data=latin plots=(diagnostics); CLASS operator batch treat; MODEL force = operator batch treat; MEANS operator batch treat/snk; output out=new p=phat r=resid;TITLE 'Latin Square Design';proc plot data=new; plot resid*(operator batch treat)/vaxis= -3.5 to 5.0 by 0.5; Title 'Residual plot'; RUN; ods graphics off; QUIT;

The Latin Square Design

Example 5-4

Page 41: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Latin Square Design

The GLM Procedure Class Level Information

Class Levels Values operator 5 1 2 3 4 5 batch 5 1 2 3 4 5 treat 5 A B C D E

Number of Observations Read 25 Number of Observations Used 25

Latin Square Design The GLM Procedure

Dependent Variable: force Sum of Source DF Squares Mean Square F Value Pr > F Model 12 548.0000000 45.6666667 4.28 0.0089 Error 12 128.0000000 10.6666667 Corrected Total 24 676.0000000

R-Square Coeff Var Root MSE force Mean 0.810651 816.4966 3.265986 0.400000

Source DF Type I SS Mean Square F Value Pr > F operator 4 150.0000000 37.5000000 3.52 0.0404 batch 4 68.0000000 17.0000000 1.59 0.2391 treat 4 330.0000000 82.5000000 7.73 0.0025

Source DF Type III SS Mean Square F Value Pr > F operator 4 150.0000000 37.5000000 3.52 0.0404 batch 4 68.0000000 17.0000000 1.59 0.2391 treat 4 330.0000000 82.5000000 7.73 0.0025

The Latin Square Design

Page 42: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

7-7 Factorial Experiments with More than Two Levels

Page 43: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Latin Square Design The GLM Procedure Student-Newman-Keuls Test for force

NOTE: This test controls the Type I experimentwise error rate under the complete null hypothesis but not under partial null hypotheses.

Alpha 0.05 Error Degrees of Freedom 12 Error Mean Square 10.66667

Number of Means 2 3 4 5 Critical Range 4.5005364 5.5107155 6.1325358 6.5839317

Means with the same letter are not significantly different.

SNK Grouping Mean N operator A 3.600 5 2 B A 1.800 5 5 B A 1.000 5 4 B A -0.800 5 3 B -3.600 5 1

The Latin Square Design

Latin Square Design The GLM Procedure Student-Newman-Keuls Test for force

NOTE: This test controls the Type I experimentwise error rate under the complete null hypothesis but not under partial null hypotheses.

Alpha 0.05 Error Degrees of Freedom 12 Error Mean Square 10.66667

Number of Means 2 3 4 5 Critical Range 4.5005364 5.5107155 6.1325358 6.5839317

Means with the same letter are not significantly different.

SNK Grouping Mean N batch A 1.800 5 2 A 1.400 5 5 A 1.000 5 3 A 0.600 5 4 A -2.800 5 1

Page 44: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

7-7 Factorial Experiments with More than Two Levels

Page 45: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Latin Square Design The GLM Procedure Student-Newman-Keuls Test for force

NOTE: This test controls the Type I experimentwise error rate under the complete null hypothesis but not under partial null hypotheses.

Alpha 0.05 Error Degrees of Freedom 12 Error Mean Square 10.66667

Number of Means 2 3 4 5 Critical Range 4.5005364 5.5107155 6.1325358 6.5839317

Means with the same letter are not significantly different.

SNK Grouping Mean N treat A 4.800 5 D A 3.600 5 A B A 1.000 5 E B C -2.600 5 C C -4.800 5 B

The Latin Square Design

Page 46: 7-2 Factorial Experiments A significant interaction mask the significance of main effects. Consequently, when interaction is present, the main effects

Residual plot resid*batch 도표 . 범례 : A = 1 관측치 , B = 2 관측치 , 등 .

resid | 5.0 + A | A 4.5 + | A 4.0 + | 3.5 + | 3.0 + | A 2.5 + | 2.0 + A | A 1.5 + | 1.0 + A | A 0.5 + | 0.0 + A A A | B -0.5 + | A -1.0 + A A A | A A -1.5 + | A -2.0 + A | -2.5 + | A A -3.0 + A | A -3.5 + ---+--------------+--------------+--------------+--------------+-- 1 2 3 4 5 batch

Residual plot resid*treat 도표 . 범례 : A = 1 관측치 , B = 2 관측치 , 등 .

resid | 5.0 + A | A 4.5 + | A 4.0 + | 3.5 + | 3.0 + | A 2.5 + | 2.0 + A | A 1.5 + | 1.0 + A | A 0.5 + | 0.0 + A B | A A -0.5 + | A -1.0 + A B | A A -1.5 + | A -2.0 + A | -2.5 + | B -3.0 + A | A -3.5 + | ---+--------------+--------------+--------------+--------------+-- A B C D E treat

The Latin Square Design

Residual plot resid*operator 도표 . 범례 : A = 1 관측치 , B = 2 관측치 , 등 . resid | 5.0 + A | A 4.5 + | A 4.0 + | 3.5 + | 3.0 + | A 2.5 + | 2.0 + A | A 1.5 + | 1.0 + A | A 0.5 + | 0.0 + A B | A A -0.5 + | A -1.0 + B A | B -1.5 + | A -2.0 + A | -2.5 + | A A -3.0 + A | A -3.5 + ---+--------------+--------------+--------------+--------------+-- 1 2 3 4 5

operator