Opinion mining has been a crucial research topic among recent studies, particularly concerning data from social media. However, a widely discussed communication concern called "the spiral of silence effect" has not been examined in opinion mining studies. In this paper, we propose an approach for detecting the spiral of silence effect in social media. We believe that the accuracy of opinion mining can be improved by considering the effect of the spiral of silence. The details and steps of the detection approach are discussed. We also collected data from two popular social networking websites, namely Facebook and Twitter, for performance measurement. Analysis findings show that the average accuracy of the proposed approach was higher than 0.85, indicating that the approach is highly effective.