The acoustic behavior of ceiling absorbers can be predicted under different surface reaction assumptions: Local and extended reaction. This study aims to experimentally validate acoustic transfer functions near a ceiling absorber in an anechoic chamber based on the two surface reaction models. First, a ceiling absorber with two mounting conditions is modeled by equivalent fluid models, such as Delany-Bazley's, Miki's, and Komatsu's model, in various ways: (1) Local vs extended reaction and (2) plane-wave vs spherical-wave incidence. For a single absorber under anechoic conditions, the acoustic transfer functions for four source-receiver pairs are simulated using a pressure-based image source model, and then compared with measurements. For a rigid backing condition, both the local and extended reaction models agree well with the measurement. For an absorber backed by an air cavity, the extended reaction model agrees better at larger incidence angles at lower frequencies than the local reaction model.