PhEur... Two-dose multiple assay with completely randomised design An assay of corticotrophin by subcutaneous injection in rats 00 80 60 0 0 00 80 60 0 0 catterplot of multiple variables against dose PhEur_.sta v*0c 9.8667-.667*x T 0.-06.*x U 9.9-.9*x két dózis, tehát a linearitás nem vizsgálható 00 0. 0. 0. 0. 0.6 0.7 0.8 0.9.0. dose T U Parallel_példák 7 00 catterplot of multiple variables against PhEur.sta 6v*0c 8.-60.07*x T -7.67*x U 0-.7*x 80 60 0 0 00 80 60 0 0 00 -.6 -. -. -.0-0.8-0.6-0. -0. 0.0 0. T U Parallel_példák 8
* treatment treatment Total Univariate Tests of ignificance for y (PhEur.sta) igma-restricted parameterization Type I decomposition; td. of Estimate: 7.66897 Degr. of M F p 706800. 706800. 68.08 0.000000 66.6 8..086 0.08 680.8 680.8 8.77 0.000000 88. 09..67 0.00780 0.9 76.6 Univariate Tests of ignificance for y (PhEur.sta) igma-restricted parameterization ive hypothesis decomposition; td. of Estimate: 7.66897 Degr. of M F p 706800. 706800. 68.08 0.000000 780.7 66. 0.7 0.000000 0.9 76.6 párhuzamosság Univariate Results for Each DV (PhEur.sta) igma-restricted parameterization ive hypothesis decomposition Degr. of y y y y M F p 706800. 706800. 68.08 0.000000 780.7 66. 0.7 0.000000 0.9 76.6 9 966.6 Parallel_példák 9 Univariate Tests of ignificance for y (PhEur.sta) igma-restricted parameterization Type I decomposition; td. of Estimate: 7.760 Degr. of M F p 968. 968. 6.6 0.000000 90.6 90.6 0.9 0.7766 6680.6 6680.6 90.9 0.000000 *.. 0.06 0.8077 687. 6 78. Univariate Tests of ignificance for y (PhEur.sta) igma-restricted parameterization ive hypothesis decomposition; td. of Estimate: 7.760 Degr. of M F p treatment treatment Total 968. 968. 6.6 0.000000 67. 8. 0. 0.000000 687. 6 78. Univariate Results for Each DV (PhEur.sta) igma-restricted parameterization ive hypothesis decomposition Degr. of y y y y M F p 968. 968. 6.6 0.000000 67. 8. 0. 0.000000 6 687. 78. 9 98.8 csak és T Parallel_példák 0
Raw Residuals 70 60 0 0 0 0 0 0-0 -0-0 -0-0 Predicted vs. Residual Values Dependent variable: y (Analysis sample) -60. 0 0 0 60 70 80 90 00 0 0 0 0.0 Predicted Values 0. Expected Normal Value.0..0 0.0-0. -.0 -. -.0 -. Normal Prob. Plot; Raw Residuals Dependent variable: y (Analysis sample).99.9.7....0.0 -.0-60 -0-0 -0-0 -0 0 0 0 0 0 0 60 70 Residual Parallel_példák Parameter Estimates (PhEur.sta) (*Zeroed predictors failed tolerance check) Level of umn Comment y y y y (B/Z/P) Param. td.err t p 9.0 7.9.906 0.000000-8.970 6.87-9.6767 0.000000 T Biased -6.00 8.89-0.768 0.687 Zeroed* 0.0000 catterplot of y against PhEur.sta v*0c Function 9.-8.97098*x-6. Function 9.-8.97098*x y 00 80 60 0 0 00 80 60 0 0 00 -.6 -. -. -.0-0.8-0.6-0. -0. 0.0 0. Include '' Include 'T' Other Parallel_példák
a minta a std 6. ln xminta ln xstd 0.060 b 8.970 xminta exp x std potency ( 0.060) 0. 899 0.899. Parallel_példák Konfidencia-intervallum CM [ V ] + ( C ) C( M ) T ± T C tα sr M T ln x minta ln x std V d a dózisok száma, n az ismétlések száma b dn C tα s r 6680.6 6680.6 78..08.076 V b dn 6680.6 ( 8.97) 0 0.9609 Parallel_példák
M T CM ln xminta ln xstd 0.060 [ + ] ( C ) C( M ) ± V T T [.0076 0.060 + 0. ].076 0.060 ± 0.076 9609 0.± 0.0 c c 0 minta 0std ( 0.8,.) ( 0.9, 0.) Parallel_példák PhEur... Three-dose Latin square design Antibiotic agar diffusion assay using a rectangular tray 6 Row Row Row Row 6 Row T T T T T T T T T T T T T T T 6 T T T 7 Row6 6 Row Row Row Row 6 Row 7 Row6 6 60 78 87 7 9 9 0 7 70 9 6 9 7 6 9 9 8 99 60 6 7 76 8 0 0 6 9 66 6 86 98 8 y: inhibiciós zóna (mm) *0 Parallel_példák 6
6 7 8 9 0 Row Treatment higszer 6 dilution 7 8 y. -0.8 6 T T. -0.8 60 T T. -0.0 78 0 0.000 87. -0.0 7 6 T T 0 0.000 9 T T. -0.8 T T 0 0.000 9. -0.8 0. -0.0 7 T T. -0.0 70 6 0 0.000 9 T T. -0.0 6 0 0.000 9. -0.0 7 dilution.^higszer log(/dilution) a Treatment egy szintje egy készítmény*dózis kombináció Parallel_példák 7 0 catterplot of y against PhEur.sta 8v*6c 00 90 80 y 70 60 0 0-0.9-0.8-0.7-0.6-0. -0. -0. -0. -0. 0.0 0. Include '' Include 'T' Other Parallel_példák 8
Main effects ANOVA Row Treatment Univariate Tests of ignificance for y (PhEur.sta) igma-restricted parameterization ive hypothesis decomposition; td. of Estimate:.70 Degr. of M F p 6.00 6.00 698.6 0.000000.00 8.0.97 0.08 8.67.7. 0.0699 80.00 70.00 8.96 0.000000. 0 0.77 A sorok közötti különbség szignifikáns, megérte! Az előnye ennek a vizsgálatnak, hogy nem az illesztett egyenesekhez képest nézi az eltéréseket, tehát azok hibája nem a reziduumba (az sorba) kerül. Parallel_példák 9.0 Normal Prob. Plot; Raw Residuals Dependent variable: y (Analysis sample) Expected Normal Value..0..0 0. 0.0-0. -.0 -. -.0 -..99.9.7....0.0 -.0 - -0-8 -6 - - 0 6 8 Residual Parallel_példák 60
General linear models> General linear models Parallel_példák 6 Row * Univariate Tests of ignificance for y (PhEur.sta) Type I decomposition; td. of Estimate:.70 Degr. of M F p 6.00 6 800.067 0.000000.00 8.08 0.00696 8.67.86 0.0806. 0.8 0.0 87.0 87.08 0.000000 8.8 8 0.96 0.7678 0.8 9 Type III, részben mások az eredmények! * Row Univariate Tests of ignificance for y (PhEur.sta) Type III decomposition; td. of Estimate:.70 Degr. of M F p 690.6 690.6 86.67 0.000000.7. 0.08 0.787 87.0 87.0.08 0.000000 8.8 8. 0.96 0.7678.00 8.. 0.00696 8.67.7.9 0.0806 0.8 9. Type I: ebben a sorrendben építjük föl a modellt a párhuzamosság teljesül Type III: ha már az összes többi tag benne van, mennyit hoz, ha ezt is bevesszük Parallel_példák 6
Parallel_példák 6 a linearitás teljesül, a másodfokú tag nem szignifikáns Row * *^ Univariate Tests of ignificance for y (PhEur.sta) Type I decomposition; td. of Estimate:.70 Degr. of M F p 6.00 6.00 698.6 0.000000.00 8.0.97 0.08 8.67.7. 0.0699.. 0. 0.7977 87.0 87.0 08. 0.000000 8.7 8.7 0.88 0.80.7.7 0. 0.8770. 0 0.77 Parallel_példák 6
Row Treatment Row * nonlin Univariate Tests of ignificance for y (PhEur.sta) igma-restricted parameterization ive hypothesis decomposition; td. of Estimate:.70 Degr. of M F p 6.00 6.00 698.6 0.000000.00 8.0.97 0.08 8.67.7. 0.0699 80.00 70.00 8.96 0.000000. 0 0.77 Univariate Tests of ignificance for y (PhEur.sta) Type I decomposition; td. of Estimate:.70 Degr. of M F p 6.00 6.00 800.07 0.000000.00 8.0. 0.00696 8.67.7.9 0.0806.. 0.8 0.0 87.0 87.0.08 0.000000 8.8 8.8 0.96 0.7678 0.8 9. treatment * másik módszer a linearitás ellenőrzésére 80.0. 87.0 8.8.7 ν--- Parallel_példák 6 Analysis of covariance, mert a párhuzamosság teljesült meredekség Row Row Row Row Row Row Parameter Estimates (PhEur.sta) (*Zeroed predictors failed tolerance check) Level of umn Comment y y (B/Z/P) Param. td.err 96.76.679 Biased -.8.88 Biased -9.8.88 Biased -8..88 Biased -.000.88 6 Biased -.000.88 6 7 Zeroed* 0.0000 8 Biased -0.6667.88 9 Biased 6.667.88 0 Biased.6667.88 Biased.000.88 Biased..88 6 Zeroed* 0.0000 Biased..687 T Zeroed* 0.0000 6 6.60.99878 Parallel_példák 66
ln x minta ln x std a minta a b std. 0.0 6.6 xminta exp x std ( 0.0) 0. 976 Parallel_példák 67 Konfidencia-intervallum CM [ V ] + ( C ) C( M ) T ± T C tα sr M T ln x minta ln x std V d a dózisok száma, n az ismétlések száma b dn C tα s r 87.0 87.0 0.7667.086.008 V b dn 87.0 ( 6.6) 6 0.9 Parallel_példák 68
M T ln x ln x minta std 0.0 CM [ + ] ( C ) C( M ) ± V T T [ + 0. ] ( 0.0) ± 0.008.008 ( 0.0).008 9 0.0 ± 0.0688 c c 0 minta 0std ( 0.9,.08) ( 0.098, 0.08) a különbség oka az eltérő modell Parallel_példák 69 y 0 0 00 9 90 8 80 7 70 6 60 0 Row; Weighted Means Current effect: F(, )., p.0067 Type III decomposition Vertical bars denote 0.9 confidence intervals 6 Row Parallel_példák 70
Kérdezhetjük, hogy a sor és oszlop hatása mint ingadozás, mekkora: Row Components of Variance (PhEur.sta) Type III decomposition y 0.087.06 9.098 Row Univariate Tests of ignificance for y (PhEur.sta) Type III decomposition; td. of Estimate:.69760 Degr. of M Den.yn. Den.yn. F p (F/R) df M Fixed 690.6 690.6 0..79 0067.8 0.000000 Random.0 8..00000 9.098. 0.0067 Random 8.7.7.00000 9.098.9 0.07960 Fixed...00000 9.098 0.8 0. Fixed 87.0 87.0.00000 9.098.8 0.000000 9. 9. Parallel_példák 7 Ha nem törődnénk a sorokkal és oszlopokkal: Univariate Tests of ignificance for y (PhEur.sta) Type III decomposition; td. of Estimate:.698 Degr. of M F p 690.6 690.6 68.7 0.000000 87.0 87.0 6. 0.000000.. 0. 0.6 069.8. Összehasonlításul a helyes földolgozás: Row * df nagyobb ugyan, de az M-be bekerül a sor és oszlop okozta különbség is, p nagyobb Univariate Tests of ignificance for y (PhEur.sta) Type I decomposition; td. of Estimate:.70 Degr. of M F p 6.00 6.00 800.07 0.000000.00 8.0. 0.00696 8.67.7.9 0.0806.. 0.8 0.0 87.0 87.0.08 0.000000 8.8 8.8 0.96 0.7678 0.8 9. Parallel_példák 7
Parameter Estimates (PhEur.sta) (*Zeroed predictors failed tolerance check) Level of umn Comment y y (B/Z/P) Param. td.err 96.76.679 Row Biased -.8.88 Parameter Estimates (PhEur.sta) Row Biased -9.8.88 (*Zeroed predictors failed tolerance Rowcheck) Biased -8..88 Row Biased -.000.88 Level of umn Comment Row y y 6 Biased -.000.88 (B/Z/P) Row Param. td.err6 7 Zeroed* 0.0000 9.6.776 8 Biased -0.6667.88 6.60.866 9 Biased 6.667.88 Biased..8979 0 Biased.6667.88 T Zeroed* 0.0000 Biased.000.88 Biased..88 6 Zeroed* 0.0000 Biased..687 T Zeroed* 0.0000 6 6.60.99878 a becsült paraméterek nem különböznek Parallel_példák 7 Konfidencia-intervallum CM [ V ] + ( C ) C( M ) T ± T C tα sr M T ln x minta ln x std V d a dózisok száma, n az ismétlések száma b dn C tα s r 87.0 87.0..0.06 az.008 helyett V b dn 87.0 ( 6.6) 6 0.9 Parallel_példák 7
ln x ln x M T minta std ( ) ( T ± C C M T ) + CM 0.0 [ ] V [ + 0. ] ( 0.0) ± 0.008.008 ( 0.0) 0.0688 ( 0.098, 0.08).008 9 0.0 ± c c 0 minta 0std ( 0.0) ± 0.06.06 ( 0.0) ( 0.898,.06) [ + 0. ] 0.98 9 0.06 ± 0.088 ( 0.07, 0.060) ( 0.9,.08) helyett Parallel_példák 7 PhEur... Four-dose randomised block design Antibiotic turbidimetric assay A csöveket vízfürdőbe teszik, nyolcasával, az egy blokk. Block 6 T 7 T 8 T 9 T 07 68 06 6 9 0 87 07 6 97 0 7 9 6 6 97 8 0 0 07 08 9 9 06 07 0 98 86 6 9 Parallel_példák 76
6 7 8 9 0 Block higszer dozis_no dilution 6 Treatment 7 8 y.7 -.6.7 -.6 9.7 -.6 7.7 -.6 0.7 -.6. -0.809 07. -0.809 0. -0.809 9. -0.809 07. -0.809 07. -0.07 68. -0.07 87. -0.07 6. -0.07. -0.07 0 Parallel_példák 77 60 catterplot of multiple variables against PhEur.sta 8v*0c 0 0 00 80 60 0 0 00 80 -. -. -.0-0.8-0.6-0. -0. 0.0 0. T Parallel_példák 78
Main effects ANOVA Block Treatment Univariate Tests of ignificance for y (PhEur.sta) igma-restricted parameterization ive hypothesis decomposition; td. of Estimate: 7.76 Degr. of M F p 780.6 780.6 98.7 0.000000 876.7 9.9.07 0.00099 066.97 7 666.00 7.0 0.000000 09.6 8.9 A blokk hatása szignifikáns. Parallel_példák 79 0 Predicted vs. Residual Values Dependent variable: y (Analysis sample) 0 Raw Residuals 0 0 -.0 Normal Prob. Plot; Raw Residuals Dependent variable: y (Analysis sample) -0. -.0. -0 80 00 0 0 60 80 00 0 0 60 80.0 Predicted Values 0. 0.0-0. -.0 -. -.0 -. Expected Normal Value.99.9.7....0.0 -.0-0 - -0-0 0 0 0 Residual Parallel_példák 80
General linear models> General linear models a párhuzamosság teljesül Block * Univariate Tests of ignificance for y (PhEur.sta) Type I decomposition; td. of Estimate: 7.69 Degr. of M F p 780.6 780.6 8. 0.000000 876.7 9.9.97 0.0009 6.0 6.0. 0.0096 07.6 07.6 80.7 0.000000.0.0 0.6 0.0 768.79.7 Parallel_példák 8 Block Treatment Block * Univariate Tests of ignificance for y (PhEur.sta) igma-restricted parameterization ive hypothesis decomposition; td. of Estimate: 7.76 Degr. of M F p 780.6 780.6 98.7 0.000000 876.7 9.9.07 0.00099 066.97 7 666.00 7.0 0.000000 09.6 8.9 Univariate Tests of ignificance for y (PhEur.sta) Type I decomposition; td. of Estimate: 7.69 Degr. of M F p 780.6 780.6 8. 0.000000 876.7 9.9.97 0.0009 6.0 6.0. 0.0096 07.6 07.6 80.7 0.000000.0.0 0.6 0.0 768.79.7 nonlin treatment * 066.97 6.0 07.6. 9. ν7--- Parallel_példák 8
Block * *^ *^ Univariate Tests of ignificance for y (PhEur.sta Type I decomposition; td. of Estimate: 7.76 Degr. of M F p 780.6 780.6 98.7 0.0000 876.7 9.9.07 0.000 6.0 6.0.7 0.009 07.6 07.6 887. 0.0000.0.0 0.7 0.997 8. 9..69 0.0 76.69 8. 0.7 0.997 09.6 8.9 a görbeség nem szignifikáns, sem másodfokú, sem harmadfokú tag nem kell 8.+76.699. Parallel_példák 8 Analysis of covariance Block Block Block Block Block Parameter Estimates (PhEur.sta) (*Zeroed predictors failed tolerance check) Level of umn Comment y y (B/Z/P) Param. td.err 9.7.899 Biased.70.68680 Biased.6.68680 Biased 8.6.68680 Biased 8.00.68680 6 Zeroed* 0.000 7 Biased 7.90.98 T 8 Zeroed* 0.000 9 -..7670 a minta a std 7.9 ln xminta ln xstd 0.07 b. xminta exp x std ( 0.07). 07 Parallel_példák 8
Konfidencia-intervallum CM [ V ] + ( C ) C( M ) T ± T C tα sr M T ln x minta ln x std V d a dózisok száma, n az ismétlések száma b dn C tα s r 07.6 07.6.9.08.00 V b dn 07.6 (.) 0. Parallel_példák 8 M T CM ln xminta ln xstd 0.07 [ + ] ( C ) C( M ) ± V T T [.00 0.07 + 0. ].00 0.07 ± 0.00 0.07 ± 0.07 c c 0 minta 0std (.09,.) ( 0.088, 0.) Parallel_példák 86
A blokk figyelembe vétele nélkül Treatment Univariate Tests of ignificance for y (PhEur.sta) igma-restricted parameterization ive hypothesis decomposition; td. of Estimate: 8.9606 Degr. of M F p 6777. 6777. 7.7 0.000000 07.6 07.6 09.6 0.000000 6.0 6.0 8.76 0.00 670.7 7 7. Univariate Tests of ignificance for y (PhEur.sta) igma-restricted parameterization ive hypothesis decomposition; td. of Estimate: 8.668 Degr. of M F p 780.6 780.6 69.09 0.00 066.97 7 666.00 96.66 0.00 86.0 7.7 Parallel_példák 87 C tα s r 07.6 07.6 7.7.0.0007 CM [ + ] ( C ) C( M ) ± V T T [.0007 0.07 + 0. ].0007 0.07 ± 0.0007 0.077 ± 0.00 ( 0.0, 0.) c c 0 minta 0std (.0,.0) (.09,.) helyett Parallel_példák 88
PhEur... Five-dose multiple assay with completely randomised design: An in-vitro assay of three hepatitis B vaccines against a standard dilution :6000 0.0 0.097 0.086 0.08 :8000 0.09 0.67 0.7 0. :000 0.9 0.7 0.77 0.8 :000 0.8 0.0 0.86 0. :000 0..0 0.97.07 6 :6000 0.0 0.097 0.07 0.08 7 :8000 0.099 0.7 0.6 0. 8 :000 0. 0. 0.68 0.06 9 :000 0.9 0.66 0.89 0. 0 :000 0..86 0.866.09 :6000 0.0 0.09 0.07 0.086 :8000 0.08 0.78 0. 0.7 :000 0.66 0. 0.69 0.6 :000 0.6 0.76 0.6 0.6 :000 0..0.0.068 T U V mindegyik készítményből független hígítási sor, abszorbanciát mérnek Föltételezzük, hogy a hígítás hibamentes! Parallel_példák 89.6 catterplot of multiple variables against PhEur.sta v*c.6+0.76*x T.9+0.76*x U.989+0.*x V.8+0.9*x...0 0.8 0.6 0. 0. 0.0-0. -0.0-9. -9.0-8. -8.0-7. -7.0-6. T U V Parallel_példák 90
Treatment Univariate Tests of ignificance for y (PhEur.sta) igma-restricted parameterization ive hypothesis decomposition; td. of Estimate:.0 Degr. of M F p 8.8007 8.8007 78.086 0.00 6.796 9 0.77.67 0.00 0.080 0 0.006 * Univariate Tests of ignificance for y (PhEur.sta Type I decomposition; td. of Estimate:.0 Degr. of M F p 8.8007 8.8007 8.90 0.000000 0.8 0.779.688 0.000006.7908.7908.66 0.000000 0. 0.08 7.88 0.00066 0.78808 0.060 nem párhuzamosak, de nem is lineárisak Parallel_példák 9 ha logaritmáljuk az adatokat, párhuzamosak lesznek a vonalak 0. catterplot of multiple variables against PhEur.sta v*c ln.879+0.88*x lnt 6.78+0.907*x lnu 6.868+0.979*x lnv 6.+0.9*x 0.0-0. -.0 -. -.0 -. -.0 -. -0.0-9. -9.0-8. -8.0-7. -7.0-6. ln lnt lnu lnv Parallel_példák 9
* Univariate Tests of ignificance for lny (PhEur.sta Type I decomposition; td. of Estimate:.08098 Degr. of M F p.. 6987.9 0.000000.7.97 7. 0.000000 7.8 7.8 79.0 0.000000 0.087 0.006 0.9 0.87 0. 0.0066 Treatment nonlin Univariate Tests of ignificance for lny (PhEur.sta) igma-restricted parameterization ive hypothesis decomposition; td. of Estimate:.08770 Degr. of M F p.. 6699.8 0.00. 9.79.0 0.00 0.67 0 0.0067 treatment *..7 7.8 0.087 0.07 ν9--- Parallel_példák 9 Tests of Homogeneity of Variances (PhEur.sta) : "Treatment" Hartley F-max Cochran C Bartlett Chi-qr. df p lny 7.767 0..6798 9 0.979 Levene's Test for Homogeneity of Variances (PhEur.sta : "Treatment" Degrees of freedom for all F's: 9, 0 M M F p lny 0.0088 0.006.68 0.097 nem megnyugtató Test of Lack of Fit (PhEur.sta) Dependent Variable Pure Err df Pure Err M Pure Err Lack of Fit df Lack of Fit M Lack of Fit F p lny 0.6707 0 0.006678 0.07 0.00686 0.967 0.0779 Parallel_példák 9
ANCOVA Univariate Tests of ignificance for lny (PhEur.sta Type I decomposition; td. of Estimate:.0809068 Degr. of M F p.. 70. 0.00 7.8 7.8 769.9 0.00.7.97 7.89 0.00 0.600 0.006 Parameter Estimates (PhEur.sta) (*Zeroed predictors failed tolerance check) Level of umn Comment lny lny (B/Z/P) Param. td.err 6.988 0.0908 0.90879 0.006 Biased -0.6608 0.09 T Biased 0.0886 0.09 U Biased -0.098 0.09 V 6 Zeroed* 0.000000 a T a 0.0886 + 0.6608 ln xt ln x 0.77 b 0.90879 xt exp x Parallel_példák 9 ( 0.77). 7 xt exp x ( 0.77). 7 P-Plot: lny: log(y) : "Treatment" (Plot of within-cell residuals) All Groups Expected Normal Value 0 - - 0.0 Predicted vs. Residual Values Dependent variable: lny (Analysis sample) 0. - -0.0-0. -0.0-0.0 0.00 0.0 0.0 0. 0.0 0.0 Observed Value Raw Residuals 0.0 0.00-0.0-0.0-0. -0.0 -. -.0 -. -.0 -. -.0-0. 0.0 0. Predicted Values Parallel_példák 96
PhEur... A completely randomised (0,,) design An assay of factor VIII 6 7 8 B 0.0 0.0 0.0 T 0.0 6 T 0.0 7 T 0.0 0.0 0. 0. 0.99 0.0 0.88 0. 0.0 0. 0. 0.99 0.9 0.88 0. 0.0 0. 0.6 0.99 0.8 0.90 0. 0.06 0.6 0.8 0.97 0.0 0.90 0.8 0.0 0.7 0.0 0.97 0.0 0.90 0.7 0.0 0.6 0.0 0.0 0. 0.9 0.7 0.0 0.8 0.9 0.99 0. 0.9 0. 0.0 0.7 0.8 0.0 0. 0.90 0. abszorbancia a jel slope ratio assay Parallel_példák 97 nem jó a lineáris függvény a blank-re Test of Lack of Fit (PhEur.sta) Dependent Variable Pure Err df Pure Err M Pure Err Lack of Fit df Lack of Fit M Lack of Fit F p y 0.0007 9 0.00000 0.0086 0.0008 9.906 0.00 Parallel_példák 98
Blank nélkül Test of Lack of Fit (PhEur.sta) Exclude condition: 0 Dependent Variable Pure Err df Pure Err M Pure Err Lack of Fit df Lack of Fit M Lack of Fit F p y 0.0006 0.00000 0.0000 0.000008.9897 0.0 dose doset Treatment Univariate Tests of ignificance for y (PhEur.sta) Type III decomposition; td. of Estimate:.00077 Exclude condition: 0 Degr. of M F p 0.097 0.097 68.00 0.00 0.898 0.898 609.6 0.00 0.89 0.89 08.6 0.00 0.0008 0.00000 Univariate Tests of ignificance for y (PhEur.sta) igma-restricted parameterization ive hypothesis decomposition; td. of Estimate:.00969 Exclude condition: 0 Degr. of M F p.97608.97608 0. 0.00 0.970 0.08 99.0 0.00 0.0006 0.00000 Parallel_példák 99 dose doset Parameter Estimates (PhEur.sta) Exclude condition: 0 y y Param. td.err 0.0979 0.00077 8.07 0.080 6.7696 0.080 6.77 R 8. 0.8 Parallel_példák 00
PhEur... A completely randomised (0,,,) design An in vitro assay of influenza vaccines conc (ug/ml) I II TI TII 6 UI 7 UII 8 8 8 7 6 0 9 0 0 9 7 0 6 7 8 7 7 haemagglutinin antigen content by radial immunodiffusion, zone of precipitation area mm Parallel_példák 0 Treatment Univariate Tests of ignificance for y (PhEur.sta) igma-restricted parameterization ive hypothesis decomposition; td. of Estimate:.000 Degr. of M F p 78.77 78.77 8. 0.000000 096.0 99.6 9. 0.000000.8.07 Test of Lack of Fit (PhEur.sta) Dependent Variable Pure Err df Pure Err M Pure Err Lack of Fit df Lack of Fit M Lack of Fit F p y.800.06797 8.989 8.067 0.9998 0.8888 Parallel_példák 0
PhEur... Four-parameter logistic curve analysis A serological assay of tetanus sera 6 7 8 9 0 dose Dilution P 6 P 0.0000 0.9.97.07.987 0.0000 0.79.6.80.808 0.000 0.0..0.0 0.00 80.6.68.98.96 0.006 60.07 0.97.6.99 0.00 0 0.8 0.666 0.86 0.8 0.006 60 0.6 0.6 0.97 0.96 0.00078 80 0.66 0. 0.0 0. 0.0009 60 0.8 0.97 0. 0.7 0.0000 0 0.76 0. 0.78 0. N0 A guinea-pig antiserum is assayed against a standard serum (0. IU/ml) using an enzyme-linked immunosorbent assay technique (ELIA). 0 two-fold dilutions of each serum were applied on a 96-well ELIA plate. Each dilution was applied twice. Parallel_példák 0 α δ y δ + + e α δ δ + dose + EC0 β ( x γ ) β x bx a e a > 0, b ln a y lépések: variancia konstans? normális eloszlás? jó-e a függvény illeszkedése? párhuzamosak-e a görbék? RP konf. intervallum Parallel_példák 0
. catterplot of multiple variables against PhEur.sta 6v*0c.0..0..0 0. 0.0-9 -8-7 -6 - - - - y yp Parallel_példák 0 6 7 8 9 0 6 7 8 9 0 6 7 8 9 0 6 7 8 9 0 Dilution e Pe 6 y 0 -.0 0.9 0 -.0 0.97 0 -.996 0.79 0 -.996 0.6 0 -.689 0. 0 -.689 0. 80 -.8 0.6 80 -.8 0.68 60 -.07 0.07 60 -.07 0 0.97 0 -.768 0 0.8 0 -.768 0 0.666 60-6.6 0 0.6 60-6.6 0 0.6 80-7. 0 0.66 80-7. 0 0. 60-7.88 0 0.8 60-7.88 0 0.97 0-8. 0 0.76 0-8. 0 0. 0 -.0 P 0.07 0 -.0 P 0.987 0 -.996 P 0.80 0 -.996 P 0.808 0 -.689 P 0.0 0 -.689 P 0. 80 -.8 P 0.98 80 -.8 P 0.96 60 -.07 P 0.6 60 -.07 P 0.99 0 -.768 P 0 0.86 0 -.768 P 0 0.8 60-6.6 P 0 0.97 60-6.6 P 0 0.96 80-7. P 0 0. 80-7. P 0 0. 60-7.88 P 0 0. 60-7.88 P 0 0.7 0-8. P 0 0.78 0-8. P 0 0. Parallel_példák 06
külön P 0.0 0.0 Predicted vs. Residual Values Dependent variable: y (Analysis sample) Include condition: 'P' s e 0.09 0.0006 s e Raw Residuals 0.0 0.0 0.00-0.0-0.0 treatment Univariate Tests of ignificance for y (PhEur kulon.sta) igma-restricted parameterization ive hypothesis decomposition; td. of Estimate:.09 Include condition: 'P'.0 Degr. of M F p. 6.8887 6.8887 68.6 0.000000.0.07 9.08 7.7 0.000000 0. 0.006 0 0.0006 Tests of Homogeneity of Variances : "treatment" Include condition: 'P' Hartley F-max Cochran C Bartlett Chi-qr. df p y.00 0..9 9 0.77 Expected Normal Value -0.0-0.0-0. 0.0 0..0..0..0. P-Plot: y All Groups : Predicted "treatment" Values (Plot of within-cell residuals) 0.0-0. -.0 -. Include condition: 'P' -.0-0.0-0.0-0.0-0.0 0.00 0.0 0.0 0.0 0.0 Observed Value Parallel_példák 07 Model: yd+(a-d)/(+exp(-bp*pe*(-gp))) (PhEur kulon.sta Dep. var: y Loss: (OB-PRED)** 0.0 0.0 Final loss:.00088 R.99976 Variance explained: 99.9% Include condition: 'P' 0.0 0.0 N0 d a bp gp 0.0 Estimate.068 0.8 -.090 -.68 0.00 td.err. 0.099 0.06788 0.00 0.06-0.0 t(6) 97.6 7.06 -.868-80. -0.0-9%CL.07 0.088669 -.0 -.77-0.0 +9%CL.906 0.987 -.060 -.67-0.0 p-value 0.00000 0.00000 0.0000 0.000-0.0 0.0006 s e s r F 0.0008 0 s.7 0 6. 0.7 0 r 0 se ( 6,0). F 0.0.69 Residual Values Predicted versus Residual Values Include condition: 'P' -0.06 0.0 0..0..0..0. Expected Normal Value.0..0 0. 0.0-0. -.0 -. Predicted Values Normal Probability Plot of Residuals Include condition: 'P' -.0-0.06-0.0-0.0-0.0-0.0-0.0 0.00 0.0 0.0 0.0 0.0 0.0 Residuals Parallel_példák 08
külön treatment Univariate Tests of ignificance for y (PhEur kulon.sta).0 igma-restricted parameterization. ive hypothesis decomposition; td. of Estimate:.069 Include condition: ''.0 Degr. of M F p 0. 9.008 9.008 09.9 0.00000 9.979 9. 00.6 0.00000-0. 0.00 0 0.00 Tests of Homogeneity of Variances : "treatment" Include condition: '' Hartley F-max Cochran C Bartlett Chi-qr. df p y 7.9600 0.986 7.06 9 0.669 Levene's Test for Homogeneity of Variances : "treatment" Degrees of freedom for all F's: 9, 0 Include condition: '' M M F p y 0.00067 0.000000.998E+8 0.00 -re nem megy! s e 0.069 0.00 s e Expected Normal Value Raw Residuals -.0 -. P-Plot: y : "treatment" (Plot of within-cell residuals) Include condition: '' -.0-0.06-0.0-0.0 0.00 0.0 0.0 0.06 0.08 0.06 0.0 0.0 0.00-0.0-0.0-0.06 Predicted Observed vs. Residual Value Values Dependent variable: y (Analysis sample) Include condition: '' Parallel_példák 09 All Groups -0.08-0. 0.0 0..0..0..0. Predicted Values Model: yd+(a-d)/(+exp(-b*e*(-g))) (PhEur kulon.sta) Dep. var: y Loss: (OB-PRED)** Final loss:.08787 R.9998 Variance explained: 99.86% Include condition: '' N0 d a b g Estimate td.err. t(6) -9%CL +9%CL p-value.0.67 0.696 -.8 -. 0.06 0.0979 0.07 0.087 9.67 7.07098-0.80-88.86.009 0. -.76 -.88.0 0.069 -.00 -. 0.00000 0.00000 0.0000 0.0000 Normal Probability Plot of Residuals Include condition: '' 0.08 Predicted versus Residual Values Include condition: '' 0.00 s e.0 0.087 s r.79 0 0 sr.79 0 F.8 s. 0 0 e ( 6,0). F 0 Expected Normal Value..0 0. 0.0-0. -.0 -. Residual Values 0.06 0.0 0.0 0.00-0.0-0.0-0.06 -.0-0.08-0.08-0.06-0.0-0.0 0.00 0.0 0.0 0.06 0.08 0.0 0..0..0..0. Residuals Predicted Values Parallel_példák 0
eddig: variancia konstans? normális eloszlás? jó-e a függvény illeszkedése? következik: párhuzamosak-e a görbék (a közös meredekség és aszimptoták feltétele teljesül)? RP konf. intervallum Parallel_példák yd+((a-d)/(+exp(-b*(pe*(-gp)+e*(-g))))) α δ step: 0. y δ + β ( x γ j ) + e Model: yd+((a-d)/(+exp(-b*(pe*(-gp)+e*(-g)... (PhEur.sta) Dep. var: y Loss: (OB-PRED)** Final loss:.09 R.999 Variance explained: 99.90% N0 d a b gp g Estimate td.err. t() -9%CL +9%CL p-value 0.6.999. -.68 -.07 0.08 0.06 0.097 0.08 0.08 0.897 96.99 7.7966-6.90 -.67 0.66.868.06 -.7 -.6 0.7.60.89 -.67 -. 0.00000 0.00000 0.00000 0.000 0.000 közös meredekség Correlation Matrix of Parameter Estimates (PhEur.sta) Parameter d a b gp g d.000-0.8 0.7-0.0-0.0 a -0.8.000-0.8 0.79 0.78 b 0.7-0.8.000-0.9-0.88 gp -0.0 0.79-0.9.000 0.697 g -0.0 0.78-0.88 0.697.000 Parallel_példák
Parallel_példák yd+((a-d)/(+exp(-(pe*bp*(-gp)+e*b*(-g))))) Model: yd+((a-d)/(+exp(-(pe*bp*(-gp)+e*b*(... (PhEur.sta) Dep. var: y Loss: (OB-PRED)** Final loss:.08796 R.999 Variance explained: 99.90% N0 d a bp gp b g Estimate td.err. t() -9%CL +9%CL p-value.967 0. -. -.68 -.7 -.07 0.07 0.08 0.00 0.09 0.06 0.08 9.776 0.066 -.9998-6.99 -.99 -.08.79 0.60 -.88 -.7 -.9 -.6.60 0.788 -.0 -.66 -.06 -.9 0.00000 0.00000 0.0000 0.000 0.0000 0.000 0.0-0.08766.8 0 - Correlation Matrix of Parameter Estimates (PhEur.sta) Parameter d a bp gp b g d.000-0.8 0.78 0.7 0.7 0.79 a -0.8.000-0.60-0.0-0.66-0.0 bp 0.78-0.60.000 0.78 0.69 0.6 gp 0.7-0.0 0.78.000 0.7 0.698 b 0.7-0.66 0.69 0.7.000 0.9 g 0.79-0.0 0.6 0.698 0.9.000 α δ y δ + β j ( x γ j ) + e külön meredekség F0.9 0 0.0876 6.8 0 Parallel_példák
Parallel_példák eddig: variancia konstans? normális eloszlás? jó-e a függvény illeszkedése? párhuzamosak-e a görbék? következik: RP konf. intervallum ln xt ln x lnγ ln γ T.07 +.68 0.77 x x T e 0.77.8 Parallel_példák 6
Model: y(pe*dp+e*d)+(((pe*ap+e*a)-(pe*dp+e*d))/(+e... (PhEur kieg.sta) Dep. var: y Loss: (OB-PRED)** Final loss:.096608 R.999 Variance explained: 99.907% N0 dp d ap a bp gp b g Estimate td.err. t() -9%CL +9%CL p-value.068.67 0.8 0.69 -.090 -.68 -.8 -. 0.009 0.0 0.09 0.08987 0.06 0.0 0.060 0.00 7.66 9.908.69 8.8-6.8 -.0 -.78-07.07.88.0909 0.077 0.9 -.7 -.7 -.8 -.97..7600 0.7099 0.006 -.008 -.60 -.06 -. 0.00000 0.00000 0.000006 0.000000 0.0000 0.000 0.0000 0.000.97 0 0.0-0.0966.097 0 - F0.08 0 0.0876 Parallel_példák 7 catterplot of multiple variables against PhEur.sta v*0c 0 - - - - - -6-7 -9-8 -7-6 - - - - ky kyp Parallel_példák 8