Dear friend, according to the literature and theoretical background perceived organizational support has a positive impact on organizational citizenship behavior. However, you can justify your statistical analysis based on cultural differences. For instance, masculine society as one of the dimensions of cultural background according to Hofestede's cultural dimensions might have a negative impact on OCB. Therefore, in my opinion, it would be better to search for cultural dimensions for justifying your claims and research outcomes.

There are many factors that contribute to nonsignificant results. Dr. Nosrati mentioned cultural differences, but there is also sample size, the measures you used, the nature of the sample, and bad luck. There is also the issue of who completed the measures (self-report vs. other-report).

It is a broad question. Unfortunately, most of us have become technicians in management research. Hence, you are right to ask this question, but it will not lead you much further. Instead, think a little bit wider to envision, what behaviour (collective or individual) you hope to impart in a society or group? That will be a good starting point.

Saeid Nosrati, Paul E. Spector, Tariq H. Malik If I modify my question further, in my model Perceived Organizational Support has a significant impact on Job performance but results show that OCB is not a mediator, can I justify my results?

@Gitu Nijhawan As you already mentioned that perceived organizational support has negative SIGNIFICANT effect which means the it's minimal in terms of strength and might due to sample size, measurement scale used, data collection method, respondents' bias while responding (as stated by @Paul Spector). The data we collect doesn't always tell what we expect.

https://towardsdatascience.com/simpsons-paradox-how-to-prove-two-opposite-arguments-using-one-dataset-1c9c917f5ff9

I will help you in this regard just contact me

Yes, It is possible. The reason can find from previous researches like cultural impact, leadership, skills variety, task significance, autonomy & other Motivational factors.

I will simply refer this below article of prof Ronald. Which stated that dont go for p-values without having the real idea about the phenomena.

Wasserstein, R. L., Schirm, A. L., & Lazar, N. A. (2019). Moving to a world beyond “p< 0.05”.

If the result is insignificant, does the sign still matters?

Resepcted Kim Hoe Looi ofcourse sign matters. positive sign indicates that change 1 unit in IV leads to postive change in DV and vise virsa when results are significants.

Respected Waqas, sign matters when the result is significant, not when result is insignificant.

Respected Kim Hoe Looi Yes! you are right. I am also talking about significant results not about insgnificants. Signs does not matters in insignificant results.