third variable problem?

Jaapaap

New Member
Dears,

I have a question with regards to "third variable problem".

I have 3 variables

A: specific bacteria
B: Symptom expression in a certain crop
X: susceptibility of the crop, which is actually a pre-condition to develop symptoms. (why the crop is susceptible is not known but quite likely can it be explained by a genetic, a physiological or a physical factor)

The interactions are as follows.

The bacteria is needed to develop specific symptoms. The specific symptoms are always caused by the specific bacteria. The presence of the bacteria does not mean that symptoms will be developed in the crop. When the crop is not susceptible the crop will not show symptoms even when the bacteria is present and able to infect the crop.

In other words

A is necessary to cause B
B need A otherwise B does not exist.
X without A will not cause B
A and X must be present to cause B

Is this an example of a third variable problem or is there another definition?

DevinUMN

New Member
I think about sentential logic with problems like these. Basically a bunch of "if...then" and statements like that. What you're looking for is a contradiction or gap in the logic, especially when it concerns A and X together. You can skip past the numbered steps if you don't really care how I got to an answer!

1.) If NOT A, then NOT B (100% of the time)
-By "NOT," I mean that the variable (A in this case) does not happen/is not present
2.) If NOT X, then NOT B (100% of the time)
3.) If A, then B (sometimes)
4.) If X, then B (sometimes)
5.) If NOT A or NOT X, then NOT B (100% of the time)

Though we've established that B is affected by BOTH A and X, there has been no evidence put forward that A is affected or affects X. Within the statements you gave, there is nothing that says that either A or X has an effect on each other. Because the Third-Variable problem relies on the Third Variable influencing BOTH independent (A) and dependent (B) variables, and because there is no evidence that T influences A, this is NOT a third-variable problem. Now if you see correlation between A and X in statistical analyses, then it WOULD be a Third-Variable Problem.