We study the SDSS-IV MaNGA database consisting of spatially resolved spectra of ~10,000 galaxies to understand the emission properties of the associated ionising sources. With the help of the BPT diagrams’ classification scheme, we train Spender (Melchoir et al 2021) to distinguish the dominant ionising mechanism with information from the spectra outside of the strong emission lines usually used for classification. We find that a large fraction of LINER-like emission spectra can be distinguished by stellar absorption features associated with an old, metal-rich stellar component. This is consistent with previous work finding that LINER emission could be driven by a stellar population of old low-mass evolved stars.