Positive LOH proxy for robustness

I found that Cv/Cb, Tc/Tg are positively correlated with G, and suggesting that they are positive proxies for robustness.  Because G~ARLS is negative, this also means that positive robustness proxies will correlative negatively with ARLS. In other words, trade-off will occur.

See _2013June18,R0-G-TcTg.regression.by.mean.R
> summary( lm( 1/tb$Tg.vs.Tc ~ tb$G*log(tb$R0)) ) #positive correlation
Call:
lm(formula = 1/tb$Tg.vs.Tc ~ tb$G * log(tb$R0))

Residuals:
     Min       1Q   Median       3Q      Max 
-0.16425 -0.04734 -0.01812  0.04014  0.22382 

Coefficients:
                Estimate Std. Error t value Pr(>|t|)
(Intercept)       1.8971     1.8101   1.048    0.329
tb$G              2.4978    12.6409   0.198    0.849
log(tb$R0)        0.2725     0.3067   0.888    0.404
tb$G:log(tb$R0)  -0.3646     2.0899  -0.174    0.866

Residual standard error: 0.1174 on 7 degrees of freedom
Multiple R-squared: 0.6008, Adjusted R-squared: 0.4297 
F-statistic: 3.512 on 3 and 7 DF,  p-value: 0.0776 


See _2013May31-H2O2LOH.R
>  ### Cv/Cb or Cb/Cv ~ robustness? I need a positive proxy
>  summary(lm( tb$Cv.vs.Cb ~ tb$G ) )  #positive, p=0.37,

Call:
lm(formula = tb$Cv.vs.Cb ~ tb$G)

Residuals:
    Min      1Q  Median      3Q     Max 
-1.2357 -0.7075 -0.3445  0.5111  1.9584 

Coefficients:
            Estimate Std. Error t value Pr(>|t|)
(Intercept)   0.5274     1.4296   0.369    0.720
tb$G         10.6344    11.3144   0.940    0.369

Residual standard error: 1.13 on 10 degrees of freedom
  (2 observations deleted due to missingness)
Multiple R-squared: 0.08117, Adjusted R-squared: -0.01071 
F-statistic: 0.8834 on 1 and 10 DF,  p-value: 0.3694