Can we use negation transformation if the correlation is positive?

> cor(my . data[, 5], my . data [, 13]) [1] 0. 5085044 > cor (my . data [, 6], my . data[, 13]) [1] 0. 637115 > cor(my . data [, 7], my . data[, 13]) [1] 0. 5821095 > cor(my . data[, 8], my . data [, 13]) [1] 0. 4839867 > cor (my . data [, 9], my . data [, 13]) [1] 0. 8119323 > cor(my . data [, 10], my . data[, 13]) [1] -0.2751204 > cor(my . data[, 11], my . data[, 13]) [1] -0. 07430394 > cor(my . data[, 12], my . data[, 13]) [1] 0. 134105> skewness (my . data [ , 5]) [1] -6. 89452 > skewness (my . data [ , 6]) [1] 0. 6054125 > skewness (my . data [ , 7]) [1] -1.01471 > skewness (my . data [, 8]) [1] 3.435634 > skewness (my . data[, 9]) [1] -0. 1824606 > skewness (my . data [ , 10]) [1] 0. 8492184 > skewness (my . data [, 11]) [1] 0.5037336 > skewness (my . data[ , 12]) [1] 16.92895