Jet installation causes jet noise to be amplified significantly at low frequencies and its physical mechanism must be understood to develop effective aircraft noise reduction strategies. A hybrid semi-empirical prediction model has recently been developed based on the instability-wave-scattering mechanism. However, its validity and accuracy remain to be tested. To do so, in this paper we carry out a systematic installed jet-noise experiment in the laboratory using a flat plate instead of an aircraft wing. We show that reducing $H$ (the separation distance between the flat plate and jet centreline) causes stronger low-frequency noise enhancement while resulting in little change to the noise shielding and enhancement at high frequencies. Decreasing $L$ (the axial distance between the jet exit plane and the trailing edge of the plate) results in reduced noise amplification at low frequencies and also weakens both the shielding and enhancement at high frequencies. Increasing the jet Mach number abates the installation effects. It is shown that the hybrid model developed in the earlier work agrees with experimental measurements and can capture the effects of varying $H$, $L$ and jet Mach number extremely well. It is concluded that the model captures the correct physics and can serve as an accurate and robust prediction tool. This new physical understanding provides insights into innovative strategies for suppressing installed jet noise.