ࡱ> 5@ %1bjbj22 %XX)zzzzzzzvvvvvv8vlw,<Ryhy~y~y~yzzzHJJJC  $RDzZzzZZzz~y~y4ZBz~yz~yHZH :zz8~yFy DAgvv| 0<֙bνVν48zzzzνz8Dzx *t2RE$1REMultiple Regression - II Extra Sum of Squares An extra sum of squares measures the marginal reduction in the error sum of squares when one or more predictor variables are added to the regression model, given that other predictor variables are already in the model. Equivalently, one can view an extra sum of squares as measuring the marginal increase in the regression sum of squares when one or several predictor variables are added to the regression model. Example: Body fat (Y) to be explained by possibly three predictors and their combinations: Triceps skinfold thickness (X1), thigh circumference (X2) and midarm circumference (X3). Body fat is hard to measure, but the predictor variables are easy to obtain. Model (X1) Fit and ANOVA Model (X2) Fit and ANOVA Model (X1, X2) Fit and ANOVA Model (X1, X2, X3) Fit and ANOVA Extra Sum of Squares Decomposition of SSR into Extra Sum of Squares What are other possible decompositions? Note that the order of the X variables is arbitrary. ANOVA Table Containing Decomposition of SSR Source of VariationSSdfMSRegression3111ErrorTotalANOVA Table with Decomposition of SSR - Body Fat Example with Three Predictor Variables. Source of VariationSSdfMSRegression396.983132.27352.171352.2733.17133.1711.54111.54Error98.416.15Total495.39Computer Packages SAS use the term Type I to refer to the extra sum of squares.  EMBED Equation.3 Example in SAS Using Body Fat Data The GLM Procedure Dependent Variable: y Sum of Source DF Squares Mean Square F Value Pr > F Model 3 396.9846118 132.3282039 21.52 <.0001 Error 16 98.4048882 6.1503055 Corrected Total 19 495.3895000 R-Square Coeff Var Root MSE y Mean 0.801359 12.28017 2.479981 20.19500 Source DF Type I SS Mean Square F Value Pr > F x1 1 352.2697968 352.2697968 57.28 <.0001 x2 1 33.1689128 33.1689128 5.39 0.0337 x3 1 11.5459022 11.5459022 1.88 0.1896 Source DF Type III SS Mean Square F Value Pr > F x1 1 12.70489278 12.70489278 2.07 0.1699 x2 1 7.52927788 7.52927788 1.22 0.2849 x3 1 11.54590217 11.54590217 1.88 0.1896 Standard Parameter Estimate Error t Value Pr > |t| Intercept 117.0846948 99.78240295 1.17 0.2578 x1 4.3340920 3.01551136 1.44 0.1699 x2 -2.8568479 2.58201527 -1.11 0.2849 x3 -2.1860603 1.59549900 -1.37 0.189 For more details on the decomposition of the SSR into extra sums of squares, please see the Schematic Representation in Figure 7.1 on page 261 Mean Squares  Note that each extra sum of squares involving a single extra X variable has associated with it one degree of freedom.Extra Sum of Squares from Several Variables  Extra sums of squares involving two extra X variables, such as SSR(X2, X3| X1), have two degrees of freedom associated with them. This follows because we can express such an extra sum of squares as a sum of two extra sums of squares, each associated with one degree of freedom.Uses of Extra Sums of Squares in Tests for Regression Coefficients Test Whether a Single Beta Coefficient is Zero (two tests are available). t-test (6.51b) discussed in chapter 6 General Linear Test Approach Full Model Hypotheses Reduced Model when H0 holds General Form of Test StatisticForm of Test Statistic for Testing a Single Beta Coefficient Equal Zero We dont need to fit both the full model and the reduced model. Only fitting a full model in SAS will provide the MSR(X3| X1 , X2) and MSE(X1,X2,X3). See the SAS outputNote: (1) here the t-test and F-test are equivalent test. (2) the F test to test whether or not (3=0 is called a partial F test (3) the F test to test whether or not all (k=0 is called the overall F test. Test Whether Several Beta Coefficients Are Zero (only one test available). General Linear Test Approach Full Model Hypotheses Reduced Model when H0 holds General Form of Test StatisticForm of Test Statistic for Testing Several Beta Coefficients Equal Zero Example: Body FatOther Tests When Extra Sum of Squares Cannot be Used, therefore both full model and reduced model have to be fitted. Example Full ModelHypothesesReduced Model when H0 holdsGeneral Test StatisticCoefficients of Partial Determination Descriptive Measures of relation ships, uses extra sum of squares. Useful in describing causal relationships. Two Predictor Variables The coefficient of multiple determination measures the proportionate reduction in the variation of Y achieved by the introduction of the entire set of X variables. Coefficient of Partial Determination uses Y and X1 both adjusted for X2 and measure the proportionate reduction in the variation of the adjusted Y by including the adjusted X1. (comments 2 on page 270) General CaseExample When X2 is added to model containing X1, SSE is reduced by 23.2% When X3 is added to model containing X1 and X2, SSE is reduced by 10.5% When X1 is added to model containing X2, SSE is reduced by only 3.1% Multicollinearity and Its Effects Some questions frequently asked are: What is the relative importance of the effects of the different predictor variables? What is the magnitude of the effect of a given predictor variable on the response variable? Can any predictor variable be dropped from the model because it has little or no effect on the response variable? Should any predictor variable not yet included in the model be considered for possible inclusion? If the predictor variables included in the model are uncorrelated among themselves and uncorrelated with any other predictor variables that are related to the response variable but are omitted from the model then relative simple answers can be given. Unfortunately, in many nonexperimental situations in business economics, and social and biological sciences, the predictor variables are correlated. For example: Family food expenditures (Y). Correlated predictors in model: Family income (X1), Family savings (X2), Age of head of household (X3). Correlated with predictors outside model: Family size (X4). When the predictor variables are correlated among themselves, intercorrelation or multicollinearity among them is said to exist. Example of Perfectly Uncorrelated Predictor Variables (Table 7.6) Models:X1 and X2 are uncorrelated. (the regression coefficient for X1 is the same for both model (1) and (2). The same holds for regression coefficient for X2. (conduct controlled experiments since the levels of the predictor variables can be chosen to ensure they are uncorrelated (SSR(X1|X2)=SSR(X1) SSR(X2|X1)=SSR(X2)(1) Source of Variation SS df MS Regression 402.250 2 201.125 Error 17.625 5 3.525 Total 419.875 7 (2) Source of Variation SS df MS Regression 231.125 1 231.125 Error 188.75 6 31.458 Total 419.875 7 (3) Source of Variation SS df MS Regression 171.125 1 171.125 Error 248.75 6 41.458 Total 419.875 7  Example of Perfectly Correlated Predictor Variables Case i Xi1 Xi2 Y Pred-Y (Model 1) Pred-Y (Model 2)12623232328983838336863636341010103103103 Models:(1)Perfect Relation between predictors: X2=5+0.5 X1(2)Two Key Implications The perfect relation between X1 and X2 do not inhibit our ability to obtain a good fit to the data. Since many different response functions provide the same good fit, we cannot interpret any one set of regression coefficients as reflecting the effect of different predictor variables. Effects of Multicollinearity We seldom find variables that are perfectly correlated. However, the implication just noted in our idealized example still have relevance. The fact that some or all predictor variables are correlated among themselves does not, in general, inhibit our ability to obtain a good fit. The counterpart in real life to many different regression functions providing equally good fits to the data in our idealized example is that the estimated regression coefficients tend to have large sampling variability when the predictor variables are highly correlated. The common interpretation of a regression coefficient as measuring the change in the expected value of the response variable when the given predictor variable is increased by one unit while all the other predictors are held constant is not fully applicable when multicollinearity exits. Example: Body Fat  Effects on Regression Coefficients Estimates of coefficients change a lot as each variable is entered in the model. In Model (3) although the F-test is significant, none of the t-tests for individuals coefficients is significant. In Model (3) the variances of the coefficients are inflated.The standard error of estimate is not substantially improved as more variables are entered in the model. Thus fitted values and predictions are neither more nor less precise.Theoretical reason for inflated variance: As the correlation between the predictors increases to one, the variance increases to infinity. The primed variables Y, X1, X2 are called the correlation transformation. The XX matrix of the primed variables is the correlation matrix rXX. As (r12)2 approaches 1 the variances march off to infinity. For more details, please read page 272-278 of ALSM PAGE 17 1E C D E ] ^ _ { | }         / 0 1 2 < = > @ A B D E F V W X j%hn EHUj'hnUjhn EHUjhnUj1hn EHUjM h\Ujd hn EHUh`jhnU h~vH* h~v6]h%8h& h~v>*h~v;. 9kd $$IfTl% t4 laT$If#$d!%d!&d!'d!N!O!P!Q! 1$1   $Ifskd$$IfTl0,"UC t04 laT  / 1 3 $IfskdX$$IfTl0,"UC t04 laT3 4 V X Z $Ifskdl$$$IfTl0,"UC t04 laTX Y q r t u   G P \ ] ` a c d e f i j l m n o r s u ukaWjJhn EHUjuHhn EHUjFhn EHUjChn EHUjAhn EHUj?@hn EHUjn=hn EHUj@;hn EHU h~v6]h!hB*phhB*ph hu6n56hu6nhu6n56jL7hn EHUj2hn EHUjt.hu6nhn EHUh~vj(hnUZ [ q s $Ifskd-$$IfTl0,"UC t04 laTs t v w 75`kd6$$IfTl" t04 laT$If`kd^2$$IfTl" t04 laT  3 G J M P  $$Ifa$$Ifgd`kd :$$IfTl#$ t04 laT $$Ifa$ P Q \ ^ ` b } $$Ifa$$Ifrkd:$$IfTl\%KL  t   t4 laTb c e g i k } $$Ifa$$Ifrkd?$$IfTl\%KLt t4 laTk l n p r t } $$Ifa$$IfrkdF$$IfTl\%KLt t4 laTt u w y { } } $$Ifa$$IfrkdL$$IfTl\%KLt t4 laTu v w x { |   + , > ? O P l m |rg_gWjh-Uh B*phh!h~vB*phjkjhn EHUjghn EHUjbehn EHUjbhn EHUj`hn EHU h~v6]j\hn EHUj [hn EHUj;Xhn EHUjaVhn EHUj0Thn EHUjkQhn EHUj;Ohn EHUh~vjDMhn EHU } ~ } $$Ifa$$IfrkdS$$IfTl\%KLt t4 laT }} $$Ifa$$IfrkdhZ$$IfTl\%KLt t4 laT    {{{ $$Ifa$$Ifrkd^$$IfTl\%KL  t   t4 laT   ! # * }t} $$Ifa$ $$Ifa$$IfrkdV_$$IfTl\% uQ  t   t4 laT* + - 4 6 = }t} $$Ifa$ $$Ifa$$Ifrkd_$$IfTl\% uQt t4 laT= > @ F H N }t} $$Ifa$ $$Ifa$$IfrkdYb$$IfTl\% uQt t4 laTN O Q W Y _ }t} $$Ifa$ $$Ifa$$Ifrkdd$$IfTl\% uQt t4 laT_ ` f l n s }t} $$Ifa$ $$Ifa$$IfrkdYg$$IfTl\% uQt t4 laTs t z }tt $$Ifa$ $$Ifa$$Ifrkdi$$IfTl\% uQt t4 laT MNdeOwwwwwwwww 5$9Dgd-gd-rkdCl$$IfTl\% uQ  t   t4 laT  $%RSɹ|of]TM=jqh8hn 5EHU\ hk2hk2hk25CJaJh~v5EH\hk25EH\jenhn 5EHU\ hh hh~v5B*CJaJph hh5B*CJaJph' *hJ h-CJOJPJQJ^JaJh-CJOJPJQJ^JaJ h-h-h~vh8jh-Ujlh-h-EHUj{$G h-CJUVaJh-OPbcds)*89:;<= 5$9Dgd-=wk  % $Ifgdk2  $Ifgdk2$Ifgdgd- 5$9Dgd- %&RTVl}}$If$If$Ifokdp$$Ifl0$z t04 laSTUVklm !>BJKXYuqjqeq[qSh5>*\j zhn EHU h>* hhh hK)hW>=CJOJQJ^JaJhK)hW>=CJ OJQJaJ hK)hW>=CJ aJ hK)hCJ aJ hK)hK)CJ aJ hK)h~vCJ aJ h~vhk2hk2hk25CJaJhk25CJH*aJhk25CJaJ h~vEHjuhn EHU h~v5\lm!>IJLypp $Ifgd8 $Ifgd8 & F gdgdK)okdfy$$Ifl0$z t04 laLMXY[xo $Ifgd8 $Ifgd8 $Ifgd8tkd2}$$Ifl0c $ - t04 laYZ\cpqxyzH_`cdhituwxz{|ݶynhcyhX&B*phhcyhX&B*H*ph jbhcyhX&B*phhX& h hi h H*h hihK) hEHjhhhn EHUjhhn EHUjhhn EHU h5\ hH* h>*hj}hhn EHU+[\xy{xo $Ifgd8 $Ifgd8 $Ifgd8tkdH$$Ifl0c $ - t04 la{|x $Ifgd8 $Ifgd8tkd$$Ifl0c $ - t04 laxx $Ifgd8 $Ifgd8tkd‡$$Ifl0c $ - t04 ladxxvvpjj$If$Ifh^hgd h^hgdtkd%$$Ifl0c $ - t04 la  =ABCET[cd )*+󾺶ytmcYj>hn EHUjȔhn EHU h~v5\ h~vH*jhn EHUh~v5>*\jˍhn EHU h~v>*hW>=h~v5CJaJhW>=hW>=5CJOJQJaJhW>=h~v hX&hK)hcyhX&B*H*ph jbhcyhX&B*phhcyhX&B*phh hK)hX&hW>=hX&>*"}$If$If$Ifokd$$Ifl0c $ - t04 la }$If$If$Ifokd"$$Ifl0c $ - t04 la  *,$If$Ifokd$$Ifl0c $ - t04 la,-uwx$If$Ifokd$$Ifl0c $ - t04 la+tuvwx  ()>?GH,ghjyqyfa h~v>*h@h@B*phh@B*phhSpB*phh@jhn EHU h~vH*&jhn CJEHUaJmHnHujhn EHU&j|hn CJEHUaJmHnHuhhh~v>*B*phjhn EHxUhi h~vEHjhn EHU h~v5\h~v$xy$If$Ifokd$$Ifl0c $ - t04 la $If$Ifokd֥$$Ifl0c $ - t04 la'$If$Ifokd$$Ifl0!F  t04 la'(*F$If$Ifokd$$Ifl0!F  t04 laFGI`$If$Ifokd$$Ifl0!F  t04 la`a||v$If $Ifgd@$Ifokd$$Ifl0!F  t04 laj ./0NOPnoru""/#0#1#D#E#F#c#d#e#######$'$\$h$s$t$u${$|$~$#hn hn CJH*OJQJ^JaJ hn hn CJOJQJ^JaJhsh8h*xjݼhn EHU h~vH* h~v6]j~hn EHfUjhn EHU h@h@h~vh?*:$If$Ifokdض$$Ifl0N $  t04 la)np||| & F$Ifgd~v$Ifokd7$$Ifl0N $  t04 lapqrstu m B!x!!"{{ & Fgd~v & Fgd~vokd$$Ifl0N $  t04 la""""h##&$i$j$r$s$$ %%%% $Ifgd8gd8gd8#$d!%d!&d!'d!N!O!P!Q! & Fgd~v~$$$$$$$$ % % % %%%%%%%%%%%%%%%%%%®wiZiZiZiZiZiZiI hshsCJOJQJ^JaJhsCJH*OJQJ^JaJhsCJOJQJ^JaJ& jhshsCJOJQJ^JaJ#hshn CJH*OJQJ^JaJ hshn CJOJQJ^JaJ& jhshn CJOJQJ^JaJh8CJOJQJ^JaJ hn hn CJOJQJ^JaJ hn hn CJOJQJ^JaJhn CJOJQJ^JaJ%%%%%riiU$Ifgd8l,K$ $Ifgd8kd1$$IfTl4F}8`]  t0    4 laT%%%%o&v&w&x&&&&&&','.'/'5'6'7'''''1(2(3(5(6(7(8(:(;(<(=(?(F(G(X(Y(((((((thXhVCJH*aJhXhVCJaJjChVEHUhV h~vH*h~v6H*] h~v6]h~v hohohojh8hn EHU hsEH h8EHjh8hn EHUhsh8hsEHjh8hn EHUh8-%%&&!ekd$IfK$L$l t044 la$Ifgd8l,\K$ekd$IfK$L$l t044 la&A&B&n&r$Ifgd8l,K$ekd$$IfK$L$l t044 la$Ifgd8l,K$n&o&p&q& $Ifgd8ekdh$IfK$L$l t044 laq&r&v&x&&riiU$Ifgd8l,K$ $Ifgd8kd$$IfTl4F}8 ]  t0    4 laT&&&&!ekd:$IfK$L$l t044 la$Ifgd8l,TK$ekd$IfK$L$l t044 la&''-'r$Ifgd8l,\K$ekd~$IfK$L$l t044 la$Ifgd8l,K$-'.'/'0' $Ifgd8ekd$IfK$L$l t044 la0'1'5'7'`'riiU$Ifgd8l,K$ $Ifgd8kd$$IfTl4F}8 ]  t0    4 laT`'a'''!ekd$IfK$L$l t044 la$Ifgd8l,K$ekdH$IfK$L$l t044 la''''r$Ifgd8l,K$ekd$IfK$L$l t044 la$Ifgd8l,K$'''' $Ifgd8ekd$IfK$L$l t044 la''''''''',(3(8(rmhhhhhhf``$IfgdogdokdX$$IfTl4F}8 ]  t0    4 laT 8(=(@(R(d(e(g(i(k(n(q(t(`kd.$$IfTlֈA^ RF? n n     t4 laT$If t(u(w(y({(~(((f``````$Ifkd$$IfTlֈA^ RF?nn t4 laT((((((((f``````$Ifkd$$IfTlֈA^ RF?nn t4 laT((((((((f``````$Ifkd9$$IfTlֈA^ RF?nn t4 laT((((((fd^^^$Ifkd$$IfTlֈA^ RF? n n     t4 laT((((((smmmd $IfgdV$Ifkd$$IfTlFS8   t0    4 laT(((((((( ) )))*)+)0)1)2)------..}/~/%0&0*0+0^0a0000000000011׻ó׫ףלל׏ׅwohRZB*phhEB*ph h~vEHjhn EHU h~vH*h~v5H*\ h~v5\jhnUjXhnUjhUjPhUh h~vH* h~v6]hXh~vjLhVEHUhVhXhVCJH*aJhXhVCJaJ+(((((rlll$Ifkd$$IfTl4FS8 `  t0    4 laT(( )p))*F**`+o,-riaa_]UUU & Fgd~v & Fgd~v HgdXkd$$IfTl4FS8    t0    4 laT --------....  & F$Ifgd~v$Ifakdw$$Ifl@ t04 la$If ..}//}w$If  & F$Ifgd~vtkd2$$Ifl02 $ ^ t04 la// 0Z00001wwwqq$If  & F$Ifgd~v$Iftkd $$Ifl02 $ ^ t04 la111111111 1!1#1$1%1h%I1hh%I10J+mHnHu h0J+jh0J+Uhdh~vhEhEB*ph 1111!1"1#1$1%1|v*h]h*&`gddtkd$$Ifl02 $ ^ t04 la#P/ =!"#$%*2 Dd     0  # A2 H@õ/S#i D`!a H@õ/S#~9hGA/ xڥMLYl&OCa)XxIsXDpJ/6 spFÜzI^*a.mF-JX@m<=Nj7kg3YfkRU9JSW\*YڍQJoU% Z박JUtx[ FozS7GVQgtٗ{RoG)RV*ERx ݘ3 nz=>~kJIT0>6pK=+h__tdЭO?bJ#ero2kWuf2ޥezְƯ06ֿ!pexx|,k fZ^?$e?ٶL+.sHpǾf|n/up'9O٦F$}S6~O}N~}o4r% }SkAO '}C\ʭ}ܧ>dsL.s$i&{}ܧ#ȭwý t${; c>L } pq6n#spw{/ܯX?iWMpbsM}Pw{/ܱM3m;{ w=nBvr{;e{l;{н;`cBvr{\gi/=#t' k][н3p92zr֝盹Ɔz➁{;2^'tHupv [:Er{ǵq2^jwƆz^jwg^*-t"4܏ u9n=B*rO]PW{,w.sdJ {WL}IOs2p?1τ\/M{݇l|#^PXOO!~n>ܟ:n=B]OpjVo~w[2p.ܷ UM w}[&=o}WCp2x~oK .qBMr߆{az[Ip0,1}[Ippz cCa=q߆;| b=$r}=$&su+p!Gp_5+GB#ȳߗ}`|p_L.˽}mOpqW}/al('pGiG>+tGpuy[Y=rɵ=B]~g>z cCa=q`}܋pu-z^#^MeE~_!{Q>BEtWVp{?;ݣ_Gp49; ~{ýӹƆzNwU|5~=7y}m?? _Pz'rְ!(0.!p.?ڳiѮb~^0P̂*{^iܖ'5X Բ(?>3SPy)N)*O6^#=@Y$^㽊a%P~j<:rvTEyxMW2-,oQy \^K T$,oP9Byx9ޫ9ZWy'jQX^"3khe*Ϣ8>ה9dk徻 py3۸7gds 1Q^ VcwF*|/]7dž>3<1c~b|쾱g{`ΗC7uM34[ț5]Ϛӱd g-L?[X:B-yۿk8V\[50NYBJ˳[wqsmfӽ'WE?_R?ܥ5}Q,?Z+#omSo OX~q,*=W>MNnA~ 4==.GE-cu!Grt*umu~琢xO񩂎Q$$If!vh5%#v%:V l t5%4TDd 0  # A2e8%UrA `!98%Ur  x4 xڥMKAǟdqKt BHx$HR(hdKZ Sz~̳6˰yeq%0|PB>EbS"4h#x 0A{| qԘPWvkP_`&"w6nm9}NjG9<>O ! Ds۽ب?jv}{2!~ܽbtuhef}ײ 1] PSˢ(؈" H 2jz$,OEYc*2x>V-kQB|DiL{} ^/E⟐BL*RӏQN]>VeJMIuQ ߃𾜙u<%m ]Cpgv*-v.Nm'e3 fPݶ$G\ERtI  z"WVфx_5]Q:Dd @ 4:    0  # A2%FV*Cfٷ `!%FV*CfٷHjXxڵXoE~3Q nU R YIԦTCIBQi\ŭH%_8zB'vfv)70by{?yւƈb.81rSҿסߕiJ| H$Q>L,gnPLVt%R25^ۢ-'OR#$ T!UnLyL3yhs!UhD }k }X%9ŵ@# D<=[#<~1tG48~Ac 估C"Ra!F܃k1ߋ$d.,?3.ڡkA?qȺ@g pkBs+n*_#ci*x'pxS,XΈnRq&#q ';_:q{еc0Y"XpuNnȝ䙹9K|GFnȝdgySͭTHN]nSꓲ|C} xV\ժvc> <=xP_[m*vUl?[o5 /i gBO"fm\dj}T< i9t/c~O):_Ƿڽ&'OMW4kHi#]1+{zyv]1cg<,4|\Gr{S22+Isz8&t +AG &3TV)ݟkQE{dѷXTWjfoJ|={7͗%rY+AL5Lh-L<^F&.LԐ p<+ժuTu{?{!303dfWf&8YLsw,>4wd/}+z/?.~τ91Ԏ)n}]wTxzk$$If!vh5U5C#vU#vC:V l t05U5C/ 4TDd p0  # A2h9 /nDu`!<9 /n" 4 xڥ=K#Aǟ8٘pQNwF#$1A _¤F5)u)C@Ep+y<ήb ;?@!9 3]m&2>IsEh:嬿t&1%Gi0S-a|(= /ԫVR=)> P޸)RWT;xީm:A+,&{5pxqZJVsk*l.حa(y[OJ'x& X zj.XYs#;8L&gX2[JSi2ޢ697! $)2EGu{Q~,U,# b=,nJ#;|Mt6ɸ[~ES]Ob#uD r9tj1{ Z j oĿ;Dd @ 00    0  # A2-|L3`t9a`!-|L3`t9HjYxڵXoE~3v\ .*TEE*q(I(Gq &:D8=sI4pCO=ĹP8 Yh.%xc)]㛷o-h([(Eo$:)9)݄ ֤J~r"fi mD + 4qa q_mm=zt>i.KA`I-RqҙgU}%+M/_+Umz,oM%,TZ':9';_,;OGxlhI qz8`y9aDP`!F܃k6߈d.,_gv$=ZbЏD.|>5VTЃ%U#ci*x'pxg&;[9p&s#q iw=.fgetL$;,Cjr4yvܼ%#Z#rg7N i3o ܩ{$[ʏ.ϋVIZ>x~!}MqZ}~HR6c[f _6oj t&jmi8jպ[oU!303d Lp屳4'Y|`d}Kz/5/?.~τ91)QwT#z $$If!vh5U5C#vU#vC:V l t05U5C/ 4T%Dd 0 F<0  # A2yc>!#yG}F`!uyc>!#yG:4 CxڥMkAǟtVPėjM{zR!Zt`=~'( @[czq`]hDm!BnlL (zԹt=ͨis8{|?bBzo'?^l]?O0̭|03(`yB|zxzF m\W7sK"-4&f 4 &IFש>kr d2x";;gPYCQGUaPy+wT-T+i=.Xgk9W'C{&Q<aaYw#e[]DmY5M2xw0tvX4OB/{EDd @ 00    0  # A2Y-Y4BQYk`!Y-Y4BQYHjcxڵXoE~3v\ *Jj%XHTCIBQI^ŭH%_8^=?'$" eoE%ko[ &Hҟ81!\GJJ}7"5)֡FN cC%4R:̊M\wQRJHwT_iԿERƱkI9Kd,I%DHZgIsﳅTW}!N󰰏'{j;e-t8IAJTUs.!9ë"*bV8m_ z nV&Q3Q[$ tf5mOcJC몛LczR2}ˋ+ÙbvbW!ov.z\oKՀ' f^ VedLY.,t~Y =f 2⩧vRaCܫR vz2{.Ȁ {e~̿27ʢcm~Om?<|s8[JDf%aOϑ kC59|+Y (?lZ0(i_GG\AGBzzI؋}AA/?Up&oUna]]oO!؅ҒstәY(aV˘0 I&{flb#~\%Ql=qJVBZ-Rm޹UΝH r6dmc*q *9ϛ?6dZ[3*9FsYm2eוh.䒌fwnvq>H$--A^QR38.)IsYTRѣtz{3LD#9 e9 dwXJeK;yrcs\U>Ř@ƀ*/x>7eCEu>͎ZN_s) }H5IX4QWӸƢ:*:t`NT^C(KR)P<JHDd @ 00    0   # A 2d2f cS*h(`!d2f cS*hHjfxڵXO[W>bcS%i"jJMp6*Rh*,F8K Tv t!S CiJE߻ׄD KOq=;?{y(!z#ӧ"Tq ӨִReDLf"cDQ1ACfŷMw(5;cٵM$z [[2\֊F?}H_Q!OFw}Z(O[P$ܳg-hz OS\̨c0z yȓJD~_OB șlj*;#@'*  mg3\fZN-!PE kJvGn-j˭` :Kˢ ?m&pp q~&u-yQ2k&DzM1Cgp T&f]I@) {0~ a9|-KMU?!wvCme@1[ vlbz,uރbZ)NYG=OAxЄ{𠺺R+dUlR|?lo*2 /qauΠaa|R, Ņ#e;@ ^:_wfͷjkݳ.'\WtkH)-®'Qm^+C+&TT4 5 ,Oʹ؜@S'Ȯ3/w(x5M!Z"7ݸ"@ #m2TV->p-.W۫ ,ZOͧLݝ>~:-`'~lmm3yo;#9N?-m= >p(7s֝8R*Fb%:>]:no}=LnY]d g2렞5I~.]dIzSfiEUЊ|Ψ(Ki'^?Hj(L~LᩧW+(Ly ڹ8X;WWBX?7s"5Jّu*1T)3Y?1/ q~^St3ΨΨ}6ޙ%YLiNxλ T\S t3@L7`=51K#nrmAk,q AتUg, b iZZSN$$If!vh5#v:V l t05/ 4TDd 0   # A  2B lC/63`! lC/N  8i1(xڥOhPy^V焷aV6DeAAAa&́xO;ѳ@Axx]ҤocL^F pÇ1 ($ i6 *9IJDi#EɊ7G}4h h=^4QR[54W(%k{d'DV]jLN[ pɏCo C?ͧ/*gਹ_qwc'Sײ{oS;t*//?Rץ/ F}xXmfBqO%ȫL6FjccNsǽLzLeR:d<$IljX?7Y# .Hb%%&c&c)Ljfb,^Na)km]u>oBzRϏA1;qYL{5&.>x]L4ۚp],,׬c;-oBB3x {$$If!vh5#v:V l t05/ 4TDd  0   # A  29(Qw5'w4 zL7`! (Qw5'w4 zL%\9hxcdd`` @b1##X`=F !#T57LS A?d* @P5< %! `vL@(\q.PHq]d\v$8AJ5_2(eHRf,e@oM-VK-WMcp ؤ @Vpp |J`8fYH"XED`mn>t.BU&۲P]4vrzE*D aXz,@R`bi2d(e!ps1~t]bJqG7'V]e9邁t `;8fj"E^(t ` oZPRjp."'Fy/ ႖\2q"` {.#RpeqIj.Cz0}($$If!vh5$#v$:V l t05$/ 4T$$If!vh555t5#v#v#vt#v:V l t555t5/  4T.Dd T0   # A  2e#]>n21? $;`!~e#]>n21? $ hLxڍJ@g&/Q`! zpgeKO(H`!,,l83je~]`I OBT(M2:`RA Gw<_]o9-8NQhn?ɵ@<^/2GLGQsP׏>%uS;Q^?qK?#x# _*%GׇǻUG:.Z&OuC2n }tri".?&W / !;L}}zߝ!!N~ǎxvSK p/ּh24.Dd 0  # A 2Ѯ _)=`!~Ѯ _)  hLxڍJ`Ͻߴ8QPE\vh:8Z!`B Rp`^܋{o-ա q͹>Bp|++/Zd18ҪSpQYj5j¢GR=Ɋ\@TSE3hpf ?~X;˽UhGhIuNl#{~L]n!+U2 ^3Z e N$SD8R5(»U%<"Yr_HË;ǿ9LxSS ,*O=!scy5me䭘7cA<٩eSvry A\Οf$$If!vh555t5#v#v#vt#v:V l t555t5/ 4TDd @@0  # A29Ny󿳶~o'А@`! Ny󿳶~o'Аh  xu @ ƿ?`[q*n #0>K[7r_~B+F{*5YVBlJZZ0j> ɺ:0 iH 7x<=DDQ# Dc8Ntvxuғoq5vgMFAz~;-u8KG7Pn;1wf1b gPRhf7Dd 0  # A2}˴-nn.L%Y@B`!Q˴-nn.L%hxڍQ1JQ}3,DA>B케))RI",(I .GvB+x >wf%X>{v"؆>ukWb"Cӎ!m.97_.ВR'N߬8 ·^1ɀ1bItiE~9 ]$~pOKVS}_􊪾wgSn]wG\&N$թ3Y"]ru_D$p?մqҶ1A\m/ -cOy/-,h5ydo1NDd 0  # A2 =<\AD`!T =<@8 h"xڍQJA}30XBR"N#(xI .?c!/v!3`->ޛyP؀~U5֪" !vr@JvOIb NEhʽAo!p$ziӟ\Q s'srXV3ɳw/芜3J|*u^>J]e8>Io0 TЙt.TMS5|nO3bx\v{QQQ #Dez%[ ֍ گCG 8}A" 7w hX\ =f1rLIREꍓT1utDyf[$rR_IM(_ \u70QPMH Dd 0  # A2.3M#L|kNUvH`!n.3M#L|kNUp4 h<xcdd```d``beV dX,XĐ  A?d @P5< %! `vL@(\q.PHq]d\v$ `fjM-VK-WM \1~I @`8fYH"Xa"[3 \V CR`*di2b(LcE@`.e [\ sj=4Ĥ\Y\ːff: ^Dd O0  # A2-,zqu؊tJ`!l-,zqu؊p0, h:xڍRJP=6-ZDA([&:4ܴB@$O A18IjMxpιU:!dUTKH4M9MEL9ƅ1Ǖ5vʫBjDh"Ջ, shuG7hW]<,oÏ]@P1kЙe( H_3eC#;a0t΃{=MdN$SBjmxMx1ɺ:)~SOᕒL})LXx"铺?'Ye{lU1opgm^^$$If!vh555t5#v#v#vt#v:V l t555t54TDd @0  # A2sz-p@D9OM`!Gz-p@D9 xڅJpƿB 퐢CjE)8tpr9|I߻KG/B`ATI1b眑Sj0&s-/mby^I fi|8')pi ~x X:H"}LIN䵤G7'/\%o>+}asוtTƽr?ٿ =Omc gyq!b>Ogiqut5P0Dd h0  # A2P<*~! -0O`!P<*~! -0@ hNxcdd``fd``beV dX,XĐ ɁKRcgb vx\ @P5< %! `vL@(\q.PHq]d\v$u `fjM-VK-WMcpb# 0:00YA  (Tq "DzXEND PπVwbbS6 H1C\ $b$r$bA7'ؾ07:18@ ĔX5padbR ,.IeC3Xof00Dd 0  # A24VLEL FͣQ`!4VLEL Fͣ (hNxڍJPsbk, DQPurVV*Ђ:|_A+sOZC.?' `<YhI)!Xm(2gyesJ[\&b|GKF{pJ~'s[u>qt(Ԗ~Щe${ sk?H 50{gF9H*7"%QoN_U>Usd'vF^>Ztތl-R#7!5!pߙG=oыoA{83h fYvcX>,|6O$*$D%HI(%%M4JF V { 4Y;͜Q\ģGaopam2rN=Sk3|@*y;>~#Sˣ^IDM~iYH+G6bwž3z]F2w;0M-h6$Z8_ҼO-f/Dd 0  # A2V> EFEp2V`!*> EFEp`:Rxcdd`` @bD"L1JE `x,56~) M @ +030$5rC0&dT20(]! A6&ZZPr.Pz#.F<`ې4#27)?+]  v`t.c+ f1͊ 5\Prbą"` {Ĥ\Y\ːBf:(@>-Dd 0  # A2y>|/c-X`!}y>|/c-  hKxڍJPƿsI+VQ *Nn:TT(Ђ} A1{8$&\ɏ;K#}l֨ !ؐ0jR΂>2לQ|~o߆= ^dVA~Kwc溜#RK[7r_~B+F{*5YVBlJZZ0j> ɺ:0 iH 7x<=DDQ# Dc8Ntvxuғoq5vgMFAz~;-u8KG7Pn;1wf1b gPRhf7$$If!vh555t5#v#v#vt#v:V l t555t54TDd @0  # A2[ p2s 'j^72c`!/ p2s 'j^@ xڅ=A_ո  f z/ನ# "xS#/b010|aꚟ@_zus®EUDcTiSMS.q1WH,c9*N#bi,v>گCG 8}A" 7w hX\ =f1rLIREꍓT1utDyf[$rR_IM(_ \u70QPMH $$If!vh555t5#v#v#vt#v:V l t555t54TDd @0  # A2sz-p@D9Oe`!Gz-p@D9 xڅJpƿB 퐢CjE)8tpr9|I߻KG/B`ATI1b眑Sj0&s-/mby^I fi|8')pi ~x X:H"}LIN䵤G7'/\%o>+}asוtTƽr?ٿ =Omc gyq!b>Ogiqut5P$$If!vh555t5#v#v#vt#v:V l t555t54TDd 0   # A2V> EFEp22h`!*> EFEp`:Rxcdd`` @bD"L1JE `x,56~) M @ +030$5rC0&dT20(]! A6&ZZPr.Pz#.F<`ې4#27)?+]  v`t.c+ f1͊ 5\Prbą"` {Ĥ\Y\ːBf:(@>$$If!vh555t5#v#v#vt#v:V l t555t5/ 4TDd 0 ! # A 2T%)1o70j`!(%)1o7 Rxcdd`` @bD"L1JE `x,56~) M @ &5rC0&dT20(]! A6&ZZPr.Pz#.F<`ې4#27)?+]  ^`t.c+ f1r!5\Prbą"` {Ĥ\Y\ːBf:@$$If!vh555t5#v#v#vt#v:V l t555t5/  4TDd @b & c $A!? ?3"`?!2ȜB1o!*m`!ȜB1o! Ƚ kxcdd  !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklnopqrstuvwxyz{|}~YRoot Entry  FEAgData mcWordDocument %ObjectPool 0AgEAg_1207379067F0Ag0AgOle CompObjfObjInfo  FMicrosoft Equation 3.0 DS Equation Equation.39qT Oh+'0 ( D P \hpxEquation Native #1TableSummaryInformation( DocumentSummaryInformation8 D``d!0 ĜL0##0KQ* W2AA?H1Z @øjx|K2B* R\``0I Dd  0 " # A"2 {CF滞8n`!{CF滞8``PxڥSO@ JKVM\q7D0G7w'fb7gc4Z޵CcH,9] 2 W/E #8leB2\v=@\\ V"0ͫ,@6aC>(uU[b~`:FSoɎR {V3mEPg1uc\g~Rܐ|ĕTxתy07Mxaڲv'VQ'][\Pn1/7:!RgֆϞImb";׎qjRz%(DYh]G8Lr<`Ih ^?Krwb@8$[{B4:Ea'6;,vZj ӽ+2F$$If!vh55z#v#vz:V l t055z/ 4Dd b # c $A? ?3"`?#2fDMTp@Bq`!:DMTp@ @-4 xڝKQϝ̺tj20@W‡2Z'MAE|I"| +Z2Bz${޻di9=^@bW-!BJRaV'[58qGr[X$BA'zXy ইba`tpn2ХڿûEpUOO粗&D|V^?ؙ]'gαr z0ڞW')߳!-.ﺗ8 ̋&i7Ҽ!7Ҽ4o_w/yhH7w*& Viި|?h.۪|?h.[X]R[Ҽ[˻ޏTT̝6##JR@0&.T0 g3SsDv Jͳ-z5ߝ3]5ꈣow.;NUkLtz(cFiYD֏=?T5~O0TS y6gѿ Sǟ3.f8 j[0H`f@1(*SfN3~1fF3a q& $ L)Oivi]jm=ѴׂTs=-]Ժ||QVK|{.7KBk YG_@" jwxvHr7}4Iw?ho G~ctld@:75Brnra'=z@ Dd H0 $ # A$2,Xq]2%oX'u`!Xq]2%oX' @ X-(xڥOhAL6n&M6AdiE/4=(bZrMj/z(xI/Ao==xP xKf ;{73/Y2@lGLȆ@ P˲#Aʔ~ Ж<9ڣ#, Ex(=z?ٖfϭܘ^]j(WoLrj_J}ym3QJrl9q ׉'~L*([,ohΥ?ѢD|fctq4؜[d@dI٣z} w1Rl1yveSXV{+kbp6d *L0WfpPW ,;[ gK읂NyprtMAX.q3)(a<3g]O`w=D\dMʓya/ }ϳoH/>ؙ6$]c1+AMlb<ȞUAHdՄۘ83=O{jHhWUk+ {wL-:).O%8YIcZ C29 +*&$&&&&{`M)& & Q?k7\z:>+~'5.5׳aps(]])X8\ X DxڕkAǿ3;Yw7Y/Fl*FPȶM+-H.ҋ="MrA ԋ%?ME[JzED fI Q^2g@,ݛo6>ڱ7?SΎO{Q-MYٔN޸:y0'}1:(?8n]ؠg)%i.gHvI'O'b/NP$)'ys!SNN6i7yb=XNU:3Ұt򃪜wT 0'#bž47:7nRO˓U0 ?%7K qס9J}&Cj+@GAX $$If!vh5 5-#v #v-:V l t05 5-/ 4pDd b ' c $A"? ?3"`?&2B2pPM~`!B2pPM@ 8 d\xcdd``ed``baV d,FYzP1n:&6! KA?H110`ꁪaM,,He`x7ӂ`'0LY ZZǰc@`^c:]j mĕ jP 0 po[3P~(PFT0*'Ͻ ǘ#AFW>pO2ȯeTb 7& Ma`庳p~f, #vgIp;mPy.hrc8`` #RpeqIj.c;:@ĞB a ye$$If!vh5 5-#v #v-:V l t05 5-/ 4*Dd t b ( c $A#? ?3"`?'2t%ӳc _ ĕP2`!H%ӳc _ ĕN @q8 xcdd``^ @c112BYL%bpuU>UF\В \: `pyĤ\Y\ 7P"CXY=M0F2$$If!vh5 5-#v #v-:V l t05 5-/ 4Dd h b ) c $A$? ?3"`?(2N^6 +c=*`!"^6 +c=0@0J@2xڝTM/QNK;mYX 4"jDW:VV4M?‚K!_$D+S_43sO{w  dh~yG!!vD4MEH]]%OCB!u4B+< ﳨ;G̤\/[>ySd5ә'M݌އ`w;`w8<,eػ_M9//6~KEԫޛNcйNTxTz7Yz9~Fk%%u5k;RB5hUs-m(ް»IGe6xXٴ pgG (eŜcńu29}N$9g)Ne+:3b8 ܎/VBΉؽTjrGUṘRTg*Z".i:UÚ(RWBoϤ$$If!vh5 5-#v #v-:V l t05 5-/ 4Dd Dnnb * c $A%? ?3"`?)2Y؂q.CI* `!Y؂q.CI* $  8'xڝW[HQ>猳3RYi=E&f8XiK׼ƲT BЋPo` =BAEj"%9s%l3g@ 8\ X DxڕkAǿ3;Yw7Y/Fl*FPȶM+-H.ҋ="MrA ԋ%?ME[JzED fI Q^2g@,ݛo6>ڱ7?SΎO{Q-MYٔN޸:y0'}1:(?8n]ؠg)%i.gHvI'O'b/NP$)'ys!SNN6i7yb=XNU:3Ұt򃪜wT 0'#bž47:7nRO˓U0 ?%7K qס9J}&Cj+@GAX $$If!vh5 5-#v #v-:V l t05 5-/ 4Dd 0 , # A&+2naՓLlm,Yn@E3bXQ`++YT=ZAc.i(]82ꅌ1QƠVj1O녹Z!NJ r8%Ȧ;<|-t&'y^n!$:'\Tr%#̳+ëO.ĴGRpɒ 򄷉9{qOkdfհ-R5h¦)nN:(KEhѬg3$~ˉaw4y\)wx6w9]T]) $$If!vh5 5-#v #v-:V l t05 5-/ 4Dd l`0 - # A',2Lec@UƢ( `! ec@UƢ XhX xڕ?H@ƿdMVpO X,EA BbDlq`+w6 bv^)\wqefq5a70 `d䰸rpƈ0;!δ:#_+G]]zE(*grc!=Zӊʦ2|};Z7_Kኊfwخ紆~|/6t@^6;M6?1sa~5h?{[io0/-"UX3:hQdƊ!g}J RĔ }o+iq8TWX*WQ̥J.2I2I5pxn)'6 ňm$oGIKoX([6-a.6b*[۞rr_^Q$$If!vh5 5-#v #v-:V l t05 5-/ 4Dd  0 . # A(-24՞fl#5 {jld`!՞fl#5 {jld`+ Pxcdd``Nc ĜL  312Ec21BUs30%41@  h 7T obIFHeA*L]2d W&@\ F\@J4? _RIp <ٿb?0~1AB 6BUI õvB世 0ĵ07-ᇹ@Dlf? 3C2sSRs 0\ 6}H`a d27Pswc=Ie'vy`frE3kJX'B f~&B7Y ,FCgg.r b#B$L&n1F-bPśe T.\d;s&f>$/@"V\021)W2+O$$If!vh5 5-#v #v-:V l t05 5-/ 4VDd nn0 / # A).2;35S1`!;35S1Z @)txڥV_HAͬެM2L4z28}4RzD4R I1z^^%(KArBPX/E\ewެ;ݲs (*(؟]a{TDOCf 9b i(60pT}j1CANz.F& @[@H@\?, Gs|𠓨:6?jզt]xMۭ葫O 0ߔyWr\^Qտ(t_SV*5(^5,]E]Ů~}gLW˷r5ƮcYY ;4QRX11Ӫ3v)f)>Ōϛ.ן"/ ?E=Y0sZ0Z(󼜁ʹ=V] ;('ִB6f:;[ It޲y0[[3lC"b-ɢN q*ѭ*1> _*j*WMT1IB)C{B[%I!YRI!1f|eih8׌+Ռ> Ĵysk,rAr]D]7*[v[Ӧ^vp7Zs!]BAHBs5q ˜Rь#>6D("Ggw'N,sêԻYȤ!#5ru#gL$5zR'*'¢h`Coϸ;kBeSh ,AI4VeN{@tNa>+eOqK9O!DtNQ]F^E&'߷G,7V6F 1I{uQE8ǟۗ݉}>¸s|to,+$$If!vh5 5-#v #v-:V l t05 5-/ 4>Dd  jj0 0 # A*/2njv"BMܠ`!njv"BMD !@Y3d\xڥoheǟw5]vk :t,mn еF X&F`" f EA&f|1DbX2@'_ʖi~s{P!A."nMd_Y(zNfɣ$+Kh]y^5޾V'}{.ֹ$g)ج_{UE/zNcF'G,V?CF>$fn!ߛ?+. hʼnAJSxBhzz J{𷻈}7$TNZQgj;ڗI׏*'8!/c}yUbnEIXآۊ򳻞ۊsgj|?*`%F<`HnΈ0]/"[b}G%QLk~:j+58YQ_r„$Br̃ʗErcj1:rMCF\;fƞ ?!hI}[%҄gO(;6IxgIާ];UuU՞ZW ţjt{XqىZpQSDM53-[̳6uӪ[Ȫ< kz]Y; XlO@3E}DŽg87D!Eͣh0{>l$o"(DyZ^Q>1lIUeU P=З2"YNH<$ezF!F{S HN$E]5N{Z2?_O7Iu7]_m ݁.Z#!qc0&`1+*ZQ?\HuflY~ZN,ςl j}3{)Xލ[pW`&w0~neO5| Y˳xd6bwl ϟ.  LoiC#Tu,M/Pά$$If!vh5 5-#v #v-:V l t05 5-/ 4lDd H @0 1 # A+02EۧD؜`!EۧD؜@H xڅ=KP{aV*(šTlEp(bL[ht'(AA0{҄x+4pɓsr޼\ =1# yEҸbqƈ08e6C0;νΦD&1G϶MlQ$SWm#BQiXf'kr-TRz*+@' IǸJ. sθγp;3 lJ I>JUr񈡸ArߪNW o'aIdMSJS1F%'UN3R|sQɁͤJҒ'&$V ՘JhvJ>4>w΄+u5CN|tM MI$$If!vh5F5 #vF#v :V l t05F5 / 4TDd 0 2 # A,12'5CϺwҩ`!'5CϺw@`0= PrxڕJAϝ$G\N1iA&&$E `e#X 3w6kD |{̙;4  (@? B,A0Y&EA:+rfmTo%9Ezl5~P{=JK&jr({~^98 ^Aucg[ȝ$JRIh.d/quieW[.3SGq/LMh(z"㵮dCjR!3IMJuL#H&,7Qðzg"{5G}6xN'廈Hcjº{)XZ5y/o+N쨺dLyVw_˕~zh3K}ĝ$$If!vh5F5 #vF#v :V l t05F5 / 4Dd  0 3 # A-22xcr/ue@9̬`!xcr/ue@98\ hxڕKAg'aQL,N-L ә/ 4vVbRI!; ?"}4\86cٽ77; !B#b#0 "K,E$l c 8_QC@]*& [aYH9#er!~UrMT=63Zl}xd>FOEtv>~A?_ֵu[k;UXSfe`FbzAv&ɐL E T 6&bR2(úMt>/kad"չA&gnA% ^D&q(V>ڒ#G&)\N A{Oϓb%LPɈqu% QAVHp|t̔$H @7!'uM\ܔ$$If!vh5F5 #vF#v :V l t05F5 / 4RDd  0 4 # A.32ڊk]x*`!ڊk]x*8\ @ pxڥAhAf7ilnA T=Vҋ M4EIJȥ^D x=ySgO^KQTΛd'&.L>7ow Gx)Bh#G$TPKSnNYY8|_-ɸ'ԤbBA `/$*y (?b|wuEV4)SUrBiG.$J@yDpߏ|c"ʈ]uG:WW3e=ZV!jzii.5.oAGZfi=?Z^Lz%Ԡ=W`H6KߖP,^6?wB&,~K_*biy o֥6Dew-_9=1=oA!m£IUukvʩQEdA}g }e,Hԓ#C88bMCX,PI))EķÇ!DhNϙݻ/&;n;|g AջFehIuCi,eTӿiŇLo2~QU&}L]LQ%Itɤex5^OL.?Uvu&a&b^y sJ1gɾPÑpC[S{;&g /KY6Űr&s<\f,Y fKkgk(놪MkM$ -_-F# Uov C2lyktu;V&OQnuwn v0xM%o)#4[ddU^m߃V{,Lv" [֜+\?ʿ 8'; O}u3 Dh5YPudǖF/."L\RĮ9M0RkHTcdQ:j^8v?~S?n9jl57 r4%1CdRvtƎ C˾uAFeՔ;M˖Z3 2͉ED+v5`6"[jaUkյ,Mh- iUUky|R8wOk6x_h-/w·#JQ UB Vϐ*M#UʛpM$D$6wM1Q-_it>E`rD>Ô"f^d5}yB(ڞ:Sxb-*Ǥ$$If!vh5 5#v #v:V l t05 5/ 4Dd  0 7 # A162*< ֻ!`!< ֻ0 @$xڥMhGήww䙑І\I|h+ShH]Lc NKډIPIi.An AB>(9zj{QHɥQcgֶ$歴`wW0+_q$qPݖ쑤TݏtA"_" 5_+0j{z5W(L ߅™ *8 g{\Z#?gB*lW;i???NׁQtu}-գK CeCX:{DEqA9/{._j'a[m|5Q7֧bw.4cm76m[-1^bfRÍM3៬LjJټ'=!dI4:;eN1j)07jx)Y mҠb6f=.sa1#HR$Lz?ud%;'R?ex1S-ۉn4r;P3j߷WqDLM&V b3O ͨyqQڸIlhtAj`N| ?KCcF| ) hGo)hU=:U窣cѕ&XZU٬j}-tUUeUr#G;E7r>jJ91LJN$)>-V~;dʙ|;`SQ$Y9䦕s Ѻ(#Ik&{&R۫Jm*2ʥO9D@U9(w3Ёr}inJݿp;$$If!vh5 5#v #v:V l t05 5/ 4$$If!vh5855] #v8#v#v] :V l4 t0+5855] / /  4TDd  hb A c $A;? ?3"`?723[_ y;XʃK`!3[_ y;Xʃ@R|xcdd``.`d``baV d,FYzP1n:&B@?b ~ ㆪaM,,He`I? @201d++&1X +|-vNHq`#jT  +!/G ܞAFDTrByA,STH^ =S%o ` 39;csS `al 9LRft̨:2p5 36XB +ssb f:|aFnp{'S94-p 8R)ANaLLJ% wA2u(2t5B 1 5B$IfK$L$q!vh5#v:V,l t5B$IfK$L$q!vh5#v:V,l t5B$IfK$L$q!vh5#v:V,l t5B$IfK$L$q!vh5#v:V,l t5$$If!vh5855] #v8#v#v] :V l4 t0+5855] / /  4TtDd hb ? c $A9? ?3"`?82mںRHi.`!mںRHi.~` @.|`xcdd``gd``baV d,FYzP1n:&! KA?H1 < ?qC0&dT20ͤ KXB2sSRsv,\~ bF\?qVB[<@ 7qr0Y榩0 z0Lv#_ N#cߚ$ 16XI9 g4N.(3 ~cmmqnk).plO `p321)W2ԡ"|b@Wf~tB$IfK$L$q!vh5#v:V,l t5B$IfK$L$q!vh5#v:V,l t5B$IfK$L$q!vh5#v:V,l t5B$IfK$L$q!vh5#v:V,l t5$$If!vh5855] #v8#v#v] :V l4 t0+5855] / /  4TlDd hb @ c $A:? ?3"`?92]| S3= `!]| S3=~ @_|Xxcdd``gd``baV d,FYzP1n:&! KA?H1 < ?qC0&dT20ͤ KXB2sSRsv,\~ NT T~0i5y12ps0nQ /`L Ra*<9@*La )LG4 fT@id[3#|;0ߘP 27)?׾pop-b w8m͹1MpBIF&&\ {:@ĞB @ufB$IfK$L$q!vh5#v:V,l t5B$IfK$L$q!vh5#v:V,l t5B$IfK$L$q!vh5#v:V,l t5B$IfK$L$q!vh5#v:V,l t5$$If!vh5855] #v8#v#v] :V l4 t0+5855] / / 4T$$If!vh5?5n5n555#v?#vn#v#v:V l t5?5n55/  / 4T$$If!vh5?5n5n555#v?#vn#v#v:V l t5?5n554T$$If!vh5?5n5n555#v?#vn#v#v:V l t5?5n554T$$If!vh5?5n5n555#v?#vn#v#v:V l t5?5n554T$$If!vh5?5n5n555#v?#vn#v#v:V l t5?5n55/  4T$$If!vh585 5 #v8#v :V l t0585 / 4TADd |0 = # A7:2N\%Y`!N\%Y ` &0_xڅJPƿ{jM#V`Th[ElA] [nuT}oC&W''_]0s\Jpɏ;\48P!~ޘ6RJ 9!ERQt< 3[JMa'!}6G(F9u[.iM-Bp HJٸ k(Yeιm;[ P4|UZx6%deb@4L;QWC$ \ iE ^XyՇoصXfvp:G8'2dfRĊ6og9+'EP;$ F&j$$If!vh585 5 #v8#v :V l4 t0+585 / /  4T<Dd |0 > # A8;2JkFm1Qb!`!JkFm1Qb!@ `h0ZxڕJAƿѻŀX"6!iB $y*,ϝr(Aca|3 @GY}H)!8TH%<: oFGK)#fogiM5ƴ4#r7{M5'-b ZmA vlM %|Z*K}ӸzIDATxv:`۽Ճۼ*w\~E:Hgx9=L0Bt@ ? Ly|^!r!r!r!.׷@^fAz-U/.?3#c/ZMNz ፴7lޓ7aH|XʫR~!9y<ʗ>yC yy!179лݾoG-\Qo~67u~9=:n{R *vbk;7nn7pSeIJC\S87Uޗ6\,+<S+3~)o|0 ?f9\67;d2o CjmU*ƔǢR^Y@ ıF~ªyUr_*9]r145FynsEv _^xPN)R(w)_h[2C/Ofѹvs"G{Euf%׉R>eTWG"}(@־Z9u.T#׌<T'{$)zrUE=rΌ]T U ҕ 쎷gVn{-xD(v-4KS'/+sr;_{vωk[s-vǸs 7+{ h-G5״|5uroZٵE>{x+{ɏ7 '״X,M6fіe}/lO> 'P儹6;~;/lMǕo7/A^6˹3okWoCjߢNέTG!{}M윈!YK?>P^FN\¦[?W{] ȣ?;_Ѯ\6/Q 9:%#:hpNBo0rϊjsD9͛xP q|Frv5rNoY%Z8Νǵfja+'Ayyק@ ԇ}]Mc.TIENDB`$$If!vh5#v:V l t05/ 4QDd 0 9 # A3=2%֚EZYK`!%֚EZYD<oxڵOOGgCl)}jn*DžUQ+U X B+` X]亮JU \TJ`~p 8P-鳳;ÚF3U-Y,;ϳywvKIbɫaҍ}섺GY9-sɠ\D{!?~+.qGA&N1BN)v sbKey DbQbZhu"̹y΅=bosKg]\2ݖao*5LO cxdd䂬se^ԣzW_c ԇөQ=&2mC=mcCW{ژ4f0;/.zHPw@}Ffkc g|C%Eu51PRWԳ֞1PV, λLh[0ςy`àѦ'ԇ@ PaMeḟ CDm-<hJG%i԰Ω(3>buګbj=&̈SE< P~+v[̌[,ALw bfU~dX>&2m>`{bmpjk}efC]d=ef-{ȳ AjH] mef)UWn"6T< eCE 7(.(eWn"6TˠdԢ|i IOĹmF N*|ϭ7щ핥H0N~n!W*ixu+p-8p0:XC"X,/ί/s6+7s8=E.\qsr0cXiUTx *5#nЌoDU@Y3׋a9W|9I8,:PPmz1s[|xBMS>5}+V e>'b14Gt+ØIHxCM50ֳȕյRnaHfqz8.*H2t_yߟ31o;>wwxh8m_jFFkw^o_y;xnbkb ᷴG-!\׆b@Ew p>IK8zJ6yHH+_$)O٪3=;pŶGjFȸ_]~:rSp|\]J)Z(^'W= =OF*#o[guկ.UHg'=$)jYrˤDDo2VhA%`Վ2`!*hA%`Վ&hK.@xڵogG%Q\; `85!e\@6poHǤl) X<CDrta !Kle h ~wvg,Yb7@\ί>fv;X;# _3=s]R 2X]sU㏂sbO26Nsu&yoge*JހCQgDm(:vNo#<. zP=&{g&rtUڗU\MdiBJkg^EUTA;ti hO:vu=nڹtv]6U@+3xԣbIeC[a틠-O΁~E[!D6h'or&AܘSQ3]YyKr +ʃ,7O)z- k d&({Be;cX d=d2I.81A$g(#Qu2A3ȍvʮQCHS~$r[|l d\Uw1sz\p6OO'?7§rYs^1{#<}5V$]dkk.ԿRkXQ*6h3MatJy\Zj45?)Y.?ln[TɛQ]{h+]h)U {~}us'B9tnc8ܿ+Ÿq%D./"'8aP$~OF Fׯcoy6\)Gpue2+]v2 5W#|N.$+᲍0B|v:~~ًvLYwoe8v_[{F-‚9 p(3SEˁV$$pO&I'q 7I9>gCux^8~;(;k%)}ZZ*=~|??\c7cZޑzM󶘏gWiu ~(S\/ӵv+݊|iwOӼ G[t sռu$$If!vh5 5^#v #v^:V l t05 5^/ 42Dd 0 ; # A5?2$E[7x}`!$E[7x}HpG0;PxڭYoTE-"Mn--tP$6i֏jd齵mwawK/UBb/&VBd&oUO@F>$J[Z2;`2n9sowf BJ: $]B |2fߪ gPcȬ.$<(!uRFH)o4ςH|1€uMu>=$+C"QV=,CX[l,=!#V=g#T~Mm'O^:-ʳ@+%9侕qzcY I'՛}NO2eU›Xf}dYT9RR-w-i Y,EpcrP؆jhVPGGE+(\!D) W(c-zI`o?[(nmjÞ@v2޹k "5kOJqҷ z[kgTg_6wHjz*LUu(UšMUCUPCU@ŢzI`SURUvkښ@WU;4=`PCU PՎf- TUF}u:V"3 URŪZ,UT%j1*TfYD:RUrLg?fI`SUH* UqMztb CUn٢=` )UYĶoA?v[{WU"2v-_lԅMJ:@?6i])845CnPa ϲf[!c;{Ǽ.96⶧rh&]Dğpo?= U}'Ǖ#vV:5o?JEͣM$<\t<+ド?h'AofsWD +cҽA+EΣ9+bg2LW GK~7D=s T?.{'bLz+V}[ḵmc:5Z6esm\يoU$ӿ"oƏƒ/SY_բKN[r_2=Sh_K.?F+JeD"uSZvYVxϋڠKo?`Ql_L?^T{nuܗUxGL̻LZ}'62J:nj}3s}ik)v4==0pd{62j& ⫬zr rX&ͭ{ҩl.7Š_m_6kk`*d6c2Jwl_Ě1Śγ~3c6d93my*s28'ǘmNd.'D_ 孢FS\tIV 'r`Sܞ{`62j] nxnܮ# /̇y̋m4ň\ƇC`0-ȸ1iʽ3-;c e-le`!]z(>c e-l( D#>+xڥ}hgǟ\.{ib%R@Rlos*4nUVfS2*Aġ_": &Ne6aT؜5{.Ք#O~yC~̻Mx#{l:Q/6z?(M;=X#LSw.)v!*-Mib-lmE륿TK}g<"Pʳas4®'v)mWOYmݻk Wp?n7 [5#گq>%ۊrE^G:g:t!Td}PK,Ҋ Jf%fxGfj]E:JjեգJI+5,ʬg#s6M2q"$t֨,#KbA1Yh?%M!%Hvi\R$t:Mk۔mT6F'# va<= 2$?8-!_Z6 ;mXV=Ͽc1]tEhϒ\0z$Yƒ`Im`]bu8JH]lNyRjW|+~K`[Q?m!4+g·aNR |KO?1ܣ=͘H hr}p4Y5lpuL|Hf5{J H6 ¹V?)R 7 RqJ\ @e.ckh#,kA6h{S'ypl`JlKק:h(cɼeM ٕ*3JX8Akg'ĵ3֫8Qw*3Zv㟐f$ͼ%N^ hW4x;Kҩ8R3,W_#&ӎ{[Ă-:$$If!vh5 5^#v #v^:V l t05 5^/ 4Multiple Regression - II.ultDouglas JonesssougougNormal  xinping cui7npMicrosoft Word 10.0@xA@wg@8g!"՜.+,0 hp  Unknown OrganizationsJ)A Multiple Regression - II Title  FMicrosoft Word Document MSWordDocWord.Document.89q      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWX1`@` Normal,Pdd5$7$8$9DH$CJ_HaJmH nHsH tHT@T Heading 1,H1$<@&5CJ0KH$\aJ0P@P Heading 2,H2$<@&5CJ$\aJ$P@P Heading 3,H3$<@&5CJ\aJH@H Heading 4,H4$<@&5\N@N Heading 5,H5 <@&5CJ\aJN@N Heading 6,H6 <@&5CJ\aJDA@D Default Paragraph FontRi@R  Table Normal4 l4a (k@(No List 8O8 Address <<6]HOH Blockquotehh<<]h^h&O& CITE6]4O!4 CODECJOJQJ^JaJnO2n Definition Compact,DL COMPACTh<<^h`XOBX Definition List,DL0<<^`0POQP Definition Term,DTCJOJQJ^JaJBOaB Definition,DFN6CJ]aJJOJ Directory,DIR @ h^h4X@4 Emphasis,EM6]O Horizontal Rule,HR]d$d %d &d 'd -DM N O P Q :O: Hypertext,A >*B*phHOH Keyboard,KBD>*CJOJQJ^JaJ0@ List Bullet,UL~ & F>ThTf^`1@ List Number,OL~ & F>Th.Tf^`O Menu~ & F>ThTf^`O PRE WIDEC /  #&),N& /CJOJQJ^JaJO Preformatted,PRE= ) ` @`! %(+-& /CJOJQJ^JaJ`O` RestartList!d5$7$8$9DH$_HmHnHsH tHu:O!: Sample,SAMP OJQJ^JDO1D Strikethrough,STRIKE78W@A8 Strong,STRONG5\FOQF Typewriter,TTCJOJQJ^JaJJOaJ Variable,VAR6CJOJQJ]^JaJ\]@\z-Bottom of Form'$$dNa$CJaJLOL z-HTML Tag >*B*CJOJQJ^JaJphV\@V z-Top of Form)$&dPa$CJaJ4 @4 Footer * !.)@. Page Number@ s Table Grid7:V,0,dd5$7$8$9DH$@ j Table Grid 3:V-0  jjj0  -dd5$7$8$9DH$5\5\u 1ETable Classic 4:V.0  jj0 j0 jj .dd5$7$8$9DH$X5\B* `J ph56B*\]`JphB* `J ph5\s@ 1ETable Classic 2#:V/0  j% j#j0 jjj%  /dd5$7$8$9DH$:5\B*`JphB* `J ph5\4@4 @Header 0 !%). /134VXZ[qstvw3GJMPQ\^`bcegiklnprtuwy{}~ !#*+-46=>@FHNOQWY_`flnstzMNdeOPbcd s ) * 8 9 : ; < = w  k %&RTVlm!>IJLMXY[\xy{|d  *,-uwxy '(*FGI`a)npqrstumBxh&ijrs ABnopqrvx-./0157`a, 3 8 = @ R d e g i k n q t u w y { ~ !p!)"F""`#o$%%%%%%%%&&&&&}''' (Z(((())))!)")#)&)00(0(00 0 000 00 00 0 0 00 0 0 00 0 0 (00[ 0[ 0[ 0[ (00w 0w (080(00 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 800 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 (08080000000000000000000000000000000000000000000000800 0  0 0  0  0  0 0 0  0  00 0 000 0 0 00 0 0 00 0 0 0 0 0 0 000 000(0(000 0 0 00 0 0 00 0 0 0 0 0 0 000 0 0 0 0(00 0 0 0 0 0 0 0 0 0 0 0 0(0a0p00p0p 0p 0t 00p 0t 0p 0p 0p 0p 0p 0t 000000u 0u 0u 0u 0u0u 0u 0u0u0u 0u 0u 0u(0u(0u0&0& 0& 0&0&0&0&0&0& 0& 0&0&0& 0&0& 0& 0& 0& 0& 0& 0& 0& 0& 0&0& 0& 0&0& 0&0& 0&0& 0& 0& 0& 0& 0&0& 0& 0&0& 0&0& 0& 0& 0& 0& 0& 80&000000(0u0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00 0 0 000 00 0 0 0 0 0 80 0  0 (0u0)" 0)" 0)" 0)"80)"80)"80)"0% 0% 0%0% 0% 0% 0% 0% 0%  0% 0% 0% 0% 0% 0% 0% 0%0% 0% 0%0(@*0@*0@00L. /134VXZ[qstvw3GJMPQ\^`bcegiklnprtuwy{}~ !#*+-46=>@FHNOQWY_`flnstzMNdeOPbcd s ) * 8 9 : ; < = w  k %&RTVlm!>IJLMXY[\xy{|d  *,-uxy '(*FGI`a)npqrmBxh&ijrs ABnopqrvx-./0157`a, 3 8 = @ R d e g i k n q t u w y { ~ !p!)"F""`#o$%%%%%%%%&&&&&}''' (Z((()))&)00(0(00000000000000000000(00[0[0[0[(00w0w(080(0000000000000000000000000000000000008000000000000000000000000000000000000(08080000000000000000000000000000000000000000000000800 0 0 0 0 0 0 0 0 0 *0m 0 000000000000000000000 0(0n(0n0000000000000000000000(00000000000000*0m00000000 0 0 000000 0 0 0 00 0 000 0 0 0(0(000 0 000000 0 00 0 0 0 0 0 0 0 0 0 0 0 00 0 00 00 00 0 0 0 0 000 00 00 00 0 0 0 800T0T0T0T0T0T(00[ 0[ 0[ 0[ 0[ 0[ 0[ 0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[0[ 0[0[80[ 0Y 0Y(00  0  0  0 80 (80 80 0#p 0#t 0#0# 0# 0# 0#0#0# 0#p 0#p 0#t 0#p 0#p 0#p 0p 0p 0t 0wX u SY+j~$%(1%1'269>CJOQdj  3 Z s P b k t }  * = N _ s O=%lL[{ ,x'F`p"%%&n&q&&&-'0'`''''8(t(((((((-./1%1 !"#$%&()*+,-./0134578:;<=?@ABDEFGHIKLMNPRSTUVWXYZ[\]^_`abcefghik$1%): !@  @H 0(  0(  B S  ?nNumbered Styles 1Numbered Styles 2 OLE Objects 1 OLE Fields 83 OLE Objects 2 OLE Fields 82 OLE Objects 3 OLE Fields 81 OLE Objects 4 OLE Fields 80 OLE Objects 5 OLE Fields 79 OLE Objects 6 OLE Fields 78 OLE Objects 7 OLE Fields 77 OLE Objects 8 OLE Fields 76 OLE Objects 9 OLE Fields 75OLE Objects 10 OLE Fields 74OLE Objects 11 OLE Fields 73OLE Objects 12 OLE Fields 72OLE Objects 13 OLE Fields 71OLE Objects 14 OLE Fields 70OLE Objects 15 OLE Fields 69OLE Objects 16 OLE Fields 68OLE Objects 17 OLE Fields 67OLE Objects 18 OLE Fields 66OLE Objects 19 OLE Fields 65OLE Objects 20 OLE Fields 64OLE Objects 22 OLE Fields 62OLE Objects 23 OLE Fields 61OLE Objects 24 OLE Fields 60OLE Objects 25 OLE Fields 59OLE Objects 26 OLE Fields 58OLE Objects 27 OLE Fields 57OLE Objects 28 OLE Fields 56OLE Objects 29 OLE Fields 55OLE Objects 30 OLE Fields 54OLE Objects 31 OLE Fields 53OLE Objects 32 OLE Fields 52OLE Objects 33 OLE Fields 51OLE Objects 34 OLE Fields 50OLE Objects 35 OLE Fields 49OLE Objects 36 OLE Fields 48OLE Objects 37 OLE Fields 47OLE Objects 38 OLE Fields 46OLE Objects 39 OLE Fields 45OLE Objects 40 OLE Fields 44OLE Objects 41 OLE Fields 43OLE Objects 42 OLE Fields 42OLE Objects 43 OLE Fields 41OLE Objects 44 OLE Fields 40OLE Objects 45 OLE Fields 39OLE Objects 46 OLE Fields 38OLE Objects 47 OLE Fields 37OLE Objects 48 OLE Fields 36OLE Objects 49 OLE Fields 35OLE Objects 50 OLE Fields 34OLE Objects 51 OLE Fields 33OLE Objects 52 OLE Fields 32OLE Objects 53 OLE Fields 31OLE Objects 54 OLE Fields 30OLE Objects 55 OLE Fields 29UU  NN##::PPee||))>>$$HHxxCCXXLLnnuukkTTGG`"`"{"{"""$$$$$$&)  !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmUU  NN##::PPee||))>>$$HHxxCCXXLLnnuukkTTGG`"`"{"{"""$$$$$$&)/7ekJL  S ] c n  " uVeVX1 2 A E S W 4"E"u%%(()#)&) ! y { 0 2 ) + z | Z^lt""d&o&)#)&)333333333333333333333 [r3]^agjpsy|~,-?@PQmn0 8 = !KMZ\z y )Iuxjsr1, !%%%%)#)&))#)&) xinping cui HnR8cAk),JM8cAs,T,]8cA ve8cA@w8cAhh^h`. hh^h`OJQJo(*@hh^h`.^`o(() ^`hH. pLp^p`LhH. @ @ ^@ `hH. ^`hH. L^`LhH. ^`hH. ^`hH. PLP^P`LhH.@hh^h`.^`o(() ^`hH. pLp^p`LhH. @ @ ^@ `hH. ^`hH. L^`LhH. ^`hH. ^`hH. PLP^P`LhH.@hh^h`.@hh^h`.@hh^h`. |sHnR,] veM@ws,Tk)] @hh^h`OJQJ^Jo( h        I        KJ."n  &HIr*xJ naf$'~(K)?*%I1k2%8<B<`]<W>=?9GBmGJ=HfkL{Yo%iKju6nSpsshyk-z3E1E!VJm-q8djXl:`zFi@]8Q'cy^@e\CuRZs8o~v dX&_ru134XZ[qstvw3GJMPQ\^`bcegiklnprtuwy{}~ !#*+-46=>@FHNOQWY_`flnstz %&Rlm!>JLMY[\y{|  *,-uxy '(*FGI`anpqu&jrsABnopqrvx-./0157`a, 3 8 = @ R d e g i k n q t u w y { ~ %%%%&&&}'''()))&)a0*a0a0a0a0a0a0@Brother MFC-7820N PrinterNe06:winspoolBrother MFC-7820N PrinterBrother MFC-7820N Printer<  ndXXL PRIV ''''du  pX7 '''' ,@XCUIENGLISHBRMS104B.EXEBRLMF04B.DLLBRB7504B.DLLBO7820N.INIdArialdXinping CuiBRLMF04B.DLLEOSCBrother MFC-7820N Printer<  ndXXL PRIV ''''du  pX7 '''' ,@XCUIENGLISHBRMS104B.EXEBRLMF04B.DLLBRB7504B.DLLBO7820N.INIdArialdXinping CuiBRLMF04B.DLLEOSC 0(0(vrQQ0(0(@{%)@UnknownGz Times New Roman5Symbol3& z ArialC5  SAS Monospace?SimSun;Wingdings?5 z Courier New"AhĤ&&Ĥ&!"J!"J!xx24d)) 3Q$?~vMultiple Regression - II Douglas Jones xinping cui4         CompObjj