Normal Distribution and Min 1q Median

> data=read.table("d:/111113/1.txt",header=T)
> model1=lm(S~u_direction+mx+my+mz,data)
> summary(model1)

Call: lm(formula = S ~ u_direction + mx + my + mz, data = data)

Residuals: Min 1Q Median 3Q Max
-11.8430 -0.3962 0.3252 0.7887 18.3963

Coefficients: Estimate Std. Error t value Pr(>|t|)
(Intercept) -0.50372 0.12738 -3.955 7.93e-05 *** u_direction -0.40368 0.07996 -5.048 4.85e-07 *** mx -0.40573 0.01292 -31.404 < 2e-16 *** my -0.25862 0.01292 -20.018 < 2e-16 *** mz -0.36557 0.01292 -28.296 < 2e-16 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 1.809 on 2043 degrees of freedom
Multiple R-squared: 0.8874, Adjusted R-squared: 0.8872
F-statistic: 4026 on 4 and 2043 DF, p-value: < 2.2e-16

> data=read.table("d:/111113/2.txt",header=T)
Call:
lm(formula = S ~ u_direction + mx + my + mz, data = data)

Residuals: Min 1Q Median 3Q Max
-11.0966 -0.6784 0.3573 1.0221 15.9857

Coefficients: Estimate Std. Error t value Pr(>|t|)
(Intercept) -1.16784 0.31215 -3.741 0.000188 *** u_direction -0.18336 0.08816 -2.080 0.037656 * mx -0.43009 0.01424 -30.195 < 2e-16 *** my -0.25563 0.01424 -17.947 < 2e-16 *** mz -0.29883 0.01424 -20.979 < 2e-16 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 1.995 on 2043 degrees of freedom
Multiple R-squared: 0.8557, Adjusted R-squared: 0.8554
F-statistic: 3028 on 4 and 2043 DF, p-value: < 2.2e-16

> data=read.table("d:/111113/2.txt",header=T)
> model1=lm(S~u_direction+mx+my+mz,data)
> summary(model1)

Call: lm(formula = S ~ u_direction + mx + my + mz, data = data)

Residuals: Min 1Q Median 3Q Max
-11.4155 -0.5540