The efficacy of mobile crowd-sensing (MCS) fundamentally relies on sustained user participation. While game-theoretic incentive mechanisms have been widely adopted for user recruitment in MCS, most existing approaches fail to simultaneously consider real-world social effects as well as data quality. To bridge this gap, we propose a Social-aware and Quality-driven Incentive Mechanism (SQIM). First, we propose the social effect-enhanced two-stage Stackelberg game to model reciprocal interactions between users while investigating the impact of inter-provider cooperation on system performance. Then, we employ backward induction to derive optimal strategies that maximize utilities and prove the unique existence of the Stackelberg equilibrium. Besides, we introduce a multi-dimensional Data Quality Evaluation Scheme (DQES) to assess data reliability through feature-based clustering, provenance information and historical behavior while enforcing reputation-based constraints to obtain final optimal strategies. Experimental evaluations on real-world datasets demonstrate that SQIM excels in participant retention, system utility and data reliability compared to other mechanisms.