The level of acceleration supported by the sensor output signal specification is usually expressed in ±g. This is the maximum acceleration that the device can measure and accurately represent through the output. For example, the output of a ±3g accelerometer is linear with accelerations up to ±3g. If you accelerate to 4g, the output may be invalid. Note that the limit value is specified by the absolute maximum acceleration, not by the measurement range. 4g acceleration will not invalidate the ±3g accelerometer.
The ratio of the acceleration (input) change to the change in the output signal. It defines the ideal linear relationship between acceleration and output (the gray line in Figure 1). Sensitivity is specified by a specific supply voltage, typically mV/g for analog output accelerometers and LSB/g or mg/LSB for digital accelerometers. It is usually expressed as a range (minimum, typical, maximum), or as a typical value plus a percentage deviation (%). For analog output sensors, the sensitivity is proportional to the supply voltage. For example, if the power supply is doubled, the sensitivity is doubled.
The change in sensitivity caused by temperature is generally expressed as a percentage (%) change per °C. The temperature effect is caused by the combination of mechanical stress and temperature coefficient of the circuit.
Ideally, there is a linear relationship between voltage and acceleration, which is described by the sensitivity of the device. The degree of non-linearity measures the deviation between the actual sensitivity and the ideal constant sensitivity, expressed as a percentage relative to the full-scale range (%FSR) or positive-negative full-scale (%FS). Usually, FSR = FS+FS. The nonlinearity of ADI accelerometers is very low and can be ignored in most cases.
If you want to get more details about accelerometer,pls visit https://www.ericcointernational.com/accelerometer/