The Zero-One version of the model

Very quickly, lets look at the model when we just look at a failure or not.

402 > Bn <- cbind(sign(Fail), 1 - sign(Fail))
402 > Bn
   [,1] [,2] 
 1    1    0
 2    1    0
 3    1    0
 4    1    0
 5    0    1
 6    0    1
 7    0    1
 8    0    1
 9    0    1
10    0    1
11    0    1
12    0    1
13    1    0
14    1    0
15    0    1
16    0    1
17    0    1
18    1    0
19    0    1
20    0    1
21    0    1
22    0    1
23    0    1

402 > mymod <- glm(Bn ~ Temp, family=binomial)
402 > summary(mymod)

Call: glm(formula = Bn ~ Temp, family = binomial)
Deviance Residuals:
       Min       1Q     Median        3Q      Max 
 -1.061112 -0.76128 -0.3782822 0.4523828 2.217506

Coefficients:
                  Value Std. Error   t value 
(Intercept)  15.0422911  7.3366528  2.050293
       Temp  -0.2321537  0.1076141 -2.157279

(Dispersion Parameter for Binomial family taken to be 1 )

    Null Deviance: 28.26715 on 22 degrees of freedom

Residual Deviance: 20.31519 on 21 degrees of freedom

Number of Fisher Scoring Iterations: 4 

Correlation of Coefficients:
     (Intercept) 
Temp -0.99716   
402 > anova(mymod)
Analysis of Deviance Table

Binomial model

Response: Bn

Terms added sequentially (first to last)
     Df Deviance Resid. Df Resid. Dev 
NULL                    22   28.26715
Temp  1  7.95196        21   20.31519

So far things look very similar, including the residual plot (believe me).