No, there is no approximate correlation in statistical analysis; either a correlation occurs with a frequency of p>.05 or there is deemed to be no correlation of any significance, if any at all — ‘~’ is meaningless as a measure of anything.

Moreover, the threshold for statstical significance is p > .05, not p >.50 — is “with correlation~.5” a typo or does it actually reflect a correlation of 50%?