1

I am quite new with R and while i am able to perform the basics i am not yet able to understand the output results. For example:

summary(lmodel) generates the following:

Call:
lm(formula = temperature ~ altitude + sea.distance, data = meteodata)

Residuals:
     Min       1Q   Median       3Q      Max 
-1.40156 -0.46083 -0.04078  0.50961  1.49844 

Coefficients:
              Estimate Std. Error t value Pr(>|t|)    
(Intercept)  19.736000   0.257222  76.728  < 2e-16 ***
altitude     -0.016509   0.004006  -4.121 0.000714 ***
sea.distance  0.531111   0.414166   1.282 0.216927    
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.7425 on 17 degrees of freedom
Multiple R-squared:  0.864, Adjusted R-squared:  0.848 
F-statistic:    54 on 2 and 17 DF,  p-value: 4.317e-08

I know about the ?summary, help(summary) and i am familiar with the basic concepts, but i would kindly ask for some thorough explanation of the summary in regards to my model (residuals, coefficients, F-statistic, Residual standard error, etc.).

It seems that there are several similar answered questions in this forum, however none of them macthes exactly my case.

Any help is welcome. Thanks in advance

lynx
  • 567
  • The terminology is all standard from any regression text, with the exception of Residual standard error which is the conditional variance. Never did figure out why. – conjugateprior Mar 24 '14 at 11:03
  • This question is an obvious duplicate, which you seem to recognize. If previous threads did not provide the information you needed, you should state what you learned from them & what specifically you still need to know. – gung - Reinstate Monica Mar 24 '14 at 16:10

1 Answers1

1

The residuals is a five number summary of the residuals from your model; somewhat helpful but plot(lmodel) will show more. The 5 number summary looks fine - the min is opposite the max, the 1st and 3rd quartiles likewise

The coefficients table shows the parameter estimates, their standard errors, associated t and p values

R squared and adjusted r squared are standard measures of how much of the variation in the DV the model is accounting for.

The F statistic is a test of the significance of the entire model

Peter Flom
  • 119,535
  • 36
  • 175
  • 383
  • i was hoping for something more thorough tbh. please take some time expanding your answer so i can accept it – lynx Mar 24 '14 at 10:19
  • 2
    What do you need explained? I think that, if you don't understand the output with the explanation I gave, you may need to learn more about regression. – Peter Flom Mar 24 '14 at 10:49
  • 1
    As i explained i am familiar with the basic concepts of statistics, in general. I will accept your answer even though i doesn't provide much extra – lynx Mar 24 '14 at 11:44