Answering "Why ln(length) is Better for ANOVA & F-Test Homework

  • Thread starter Thread starter MaxManus
  • Start date Start date
  • Tags Tags
    anova
Click For Summary

Homework Help Overview

The discussion revolves around the use of logarithmic transformation, specifically ln(length), in the context of ANOVA to assess the dependency of length on age. Participants are exploring the implications of using ln(length) versus length in their analysis.

Discussion Character

  • Exploratory, Conceptual clarification, Assumption checking

Approaches and Questions Raised

  • Participants are examining the differences in F-values and p-values between the two models. There is a focus on understanding why ln(length) might be preferred despite a lower F-value. Some participants question the clarity of the original poster's inquiry regarding the comparison of ln(length) and length.

Discussion Status

The discussion is ongoing, with participants attempting to clarify the original question and explore the reasoning behind the choice of transformation. Some guidance has been offered regarding the comparison of models, but there is no explicit consensus on the best approach or interpretation.

Contextual Notes

Participants are working within the constraints of their homework assignment, which may limit the depth of exploration into the statistical implications of their findings. There is also a mention of comparing R^2 values between the models, indicating a focus on model fit.

MaxManus
Messages
268
Reaction score
1

Homework Statement

I am using Anova to check if length is dependent of age. The question is, "why is it better to use ln(length)?"

How should I answer this?
I think both seems linear on the two plots(attachment)
But ln(length) gives lower F-value and isn't that a bad thing so why is it better to use? Both has p-value = 2.16*10^-16 in R.

F-value ln(length) = 483
F-value length = 522
 

Attachments

  • agevslength.jpeg
    agevslength.jpeg
    11.6 KB · Views: 539
  • agevsloglength.jpeg
    agevsloglength.jpeg
    12.8 KB · Views: 531
Last edited:
Physics news on Phys.org
MaxManus said:
The question is, "why is it better to use ln(age)?"
You looked at ln(length) v age.

If that is the question, why didn't you look at answering the question?
 
D H said:
You looked at ln(length) v age.

If that is the question, why didn't you look at answering the question?

Sorry, the question is why is ln(length) vs age better than length vs age
 
Have tried to use a linear regression on the two models to see which is best by comparing R^2. On linear-linear I got 0.5678. To compute R^2 for log linear I first ran the regression ln(length) = b0 + b1*age
and then computed ^length^ = exp(^b0)exp(^b1^x) and estimated the R^2 between ^length^ and x. R^2 = 0.49, which is worse

where ^text^ means estimates for text
 

Similar threads

  • · Replies 7 ·
Replies
7
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K
Replies
5
Views
2K
  • · Replies 2 ·
Replies
2
Views
3K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 11 ·
Replies
11
Views
3K
  • · Replies 2 ·
Replies
2
Views
12K
  • · Replies 2 ·
Replies
2
Views
4K
  • · Replies 6 ·
Replies
6
Views
7K
Replies
1
Views
1K