If you perform a land magnetometer survey and you see random spikes in the data spread out over many cycles, these being real magnetic events, then the issue might not be the sensor but metal on your person.
A baseball cap often has a little steel button on the top, and for land magnetometers where the sensor is overhead, this steel is very close to the sensor. With movement during walking, this can create strong magnetic spikes in the data. To test this, you can mark the location of the magnetic sensor in relation to the steel button on a baseball cap and record the magnetic field standing still. Then lower the sensor 1 meter and you'll observe the anomaly increase by a factor of 16 or so.
For magnetometer surveys, it is important that the operator is free of electronics and metal. This means no cell phone, no belt, no steel toe boots, etc. Small quality checks can have a big impact on the data and save time in the field.