Just like it suggests: that the measures CWB and DJ are negatively correlated. The higher the CWB will be, the lower the DJ will be, and vice versa (however only with ca. -30% correlation only).
The p-value suggests that the hypothesis "the correlation is zero" can be significantly rejected. In a non-technical "sloppy" explanation: the correlation is not zero. Note that the hypothesis is NOT "the correlation is positive/negative", which needs a one-tailed test instead of a two-tailed one.
Technically, the correlation is weak as stated by Robert A Canales ; however when interpreting the results you are also advised to take account for context and the standards of your discipline. In social science or economics research a correlation of 31% is not that weak. However, here we do not know what the variables are.
Since the p-value is approved by rejecting the independence with magnitude less than (0.05), then the correlation exists indeed between two variables X & Y.
But, low absolute value of this correlation (0.312) may be due to small sample size.
Whereas, the negative (sign) of correlation value refers to the converse relationship between X & Y.
negative correlation means it has an indirect relationship, while one of the variables grows, the other decreases, but this only occurs in approximately 31% of cases.
the significance, it turns out to be less than alpha, so it says it exists, but in the analysis, this is considered very weak.
Cindy Burgos Actually, the negative correlation coefficient indicates that there is an inverse linear relationship between two variables. Your result indicates the weak association.