Are there larger stellar velocity dispersions in low redshift Type-1 AGN than in Type-2 AGN?

The main objective is to check the Unified Model (UM) expected similar stellar velocity dispersions between Type-1 AGN and Type-2 AGN, then to provide further clues on BH mass properties. Not similar as previous comparisons of BH masses estimated by \msig relations to Type-2 AGN but Virial BH masses in Type-1 AGN, reliable stellar velocity dispersions $\sigma$ measured through absorption features around 4000\AA~ are directly compared between so-far the largest samples of 6260 low redshift ($z~<~0.3$) Type-1 AGN and almost all the Type-2 AGN in SDSS DR12. Although half of Type-1 AGN do not have measured $\sigma$ due to unapparent absorption features overwhelmed by AGN activities, both properties of mean spectra of Type-1 AGN with and without measured $\sigma$ and positive dependence of $\sigma$ on [O~{\sc iii}] luminosity can lead to statistically larger $\sigma$ of all the Type-1 AGN than the 6260 Type-1 AGN with measured stellar velocity dispersions. Then, direct $\sigma$ comparisons can lead to statistically larger $\sigma$ in Type-1 AGN, with confidence level higher than 10sigma, after considering necessary effects of different redshift and different central AGN activities. Although Type-1 AGN have $\sigma$ only about $(9\pm3)$\% larger than Type-2 AGN, the difference cannot be well explained at current stage. Unless there was strong evidence to support different \msig relations or to support quite different evolution histories between Type-1 AGN and Type-2 AGN, the statistically larger $\sigma$ in Type-1 AGN provides a strong challenge to the Unified model of AGN.