How to explain the impacts as shown by the coefficients of beta+ and beta- in NARDL models?

[Tetsuya]

Dear everyone,

I am not sure about the interpretation of those coefficients of beta+ and beta- of x+ and x- in NARDL models. I found there are different interpretations in youtube videos. A -0.2 of beta- for log-log NARDl model implies that, if x- decreases 1%, y increases 0.2%. Am I right?
Could anyone tell me? Thanks so much.

Best regards,
Tetsuya