# Linear regression

#### gloew

##### New Member
Hi all

I was given a data sample of 35 countries which included the country name,the number of railways and the number of airports in the given country

My question is which is the independent variable and which is the dependent variable in this problem?

Pls help me,thanks!

#### bayesian

##### New Member
Hi all

I was given a data sample of 35 countries which included the country name,the number of railways and the number of airports in the given country

My question is which is the independent variable and which is the dependent variable in this problem?

Pls help me,thanks!
It all depends on what you are trying to model. You say this is a linear regression problem - are you wondering if the number of airports tends to be greater in countries that have a small number of railways? Or maybe you would like to determine how the number of railways is related to the number of airports in the country.

Which variable is dependent and which is independent is determined by the question you want to answer.

#### areteus

##### New Member
Lineer Regression

Dear friends,
I want to test if some of the the numerical factors such as age, serum hematocrit level have impact on numerical factor nerve conduction velocity.
Is Lineer regression OK with it?
Best regards,
Areteus

#### TheEcologist

##### Global Moderator
Hi all

I was given a data sample of 35 countries which included the country name,the number of railways and the number of airports in the given country

My question is which is the independent variable and which is the dependent variable in this problem?

Pls help me,thanks!
As both are random varaibles (you did not 'control' or select each variable) a linear regression is (strickly speaking) not relevant here. You should use a correlation technique.

#### areteus

##### New Member
I will use correlation
However will it let me know which factors does the nerve conduction velocity depend on or which factors have an impact on CVelocity such as height, drug dosage and ***?
If not which method do you suggest me to use?
Areteus

#### Philyuko

##### New Member
a correlation shows a relationship but not a cause and effect.

The number of ice-creams sold in any one day may correlate highly with the number of deaths by drowning ... but that doesn't mean that ice-cream sales CAUSE drownings.

In other words - be careful how you interpret the result.