0

I would like to know what is the specific term given to this case, or how is this justified in technical words:

x1 is not correlated to y but x2 which is correlated to x1 is correlated to y.

Joehat
  • 111
  • 3
  • Could you clarify what you mean by "related to"? – whuber Sep 09 '21 at 13:48
  • "correlated", sorry. When plotting a scatter plot between x1 and x2, correlation is noticed. When plotting x2 and y a correlation is also noticed. However, a weak correlation (or no obvious correlation) is seen when plotting x1 and y. – Joehat Sep 09 '21 at 13:54
  • 1
    Have you looked at our threads on mutual correlations? Start with https://stats.stackexchange.com/questions/256440 and https://stats.stackexchange.com/questions/5747. https://stats.stackexchange.com/questions/32169 is helpful for constructing examples of any possible sets of correlations. For the terminology, search our site for "regression" along with "mediat*" or "confound*", or include both "significan*" and "not". – whuber Sep 09 '21 at 13:57

0 Answers0