A service condition, a cause of a malfunction, or another aspect of a device in a large group of devices to be managed can be analyzed in an accurate and efficient manner. A complete test involving the entire number of devices in a large group of managed devices (T) is periodically performed to determine whether the devices are operating normally or have a malfunction; a test result (Ic) is recorded for each cycle of the complete test, and a device that has been found to be malfunctioning is repaired or replaced; and analysis data G, E are created showing a malfunctioning frequency (N) of each of the managed devices (T) on the basis of the test result (Ic) of the complete test that spans a plurality of cycles.