Both methods must give the same results as they just two different approaches to measure the same exact quantity. That said, if there ir a discrepancy between both the results, then you should question the reliability of the data that you acquired.

Several reasons could be behind the discrepancies, but since you do not give many details it is difficult to say for sure. I will mention some possibilities:

-Samples with very low signals that result in a bad centering, either from the beginning of the measurements where at some point the centering was lost.

-Very heterogeneous samples, which more than one "magnetic moment center"-

-If the differences are in the absolute value but with the same trend, then you might be experiencing geometric or radial offset related effects which add inaccuracies to the final measured magnetic moment, even for high signal samples. This affects differently the magnetic moment measured using the DC and VSM methods. For this particular problem I recommend the following paper, where we provide a solution:

Amorim, C. O., Mohseni, F., Dumas, R. K., Amaral, V. S., & Amaral, J. S. (2021). A geometry-independent moment correction method for the MPMS3 SQUID-based magnetometer. Measurement Science and Technology, 32(10), 105602.

Thank you for your valueable information and suggestions. Carlos de Oliveira Amorim

If we have very small sample in large capsule then which mode is more reliabale? I think VSM mode is better as it can take average of measurments which can reduce noise?

Am I correct?

Actually it is not that straightforward. Even though the VSM is more sensitive, if you are using a capsule with a signal comparable to your sample, then you may loose the centering of the sample while doing the measurement. Particularly if you measure a different temperatures.

If that is the case there are several options:

1-the most reliable is to change the holder for some that has a signal much lower than the sample. For example quartz sample holder and using GE varnish to secure your sample. This is the most reliable approach, and in principle works for both measurements

2-Measure your sample holder and capsule with and without sample, deactivating any type of autocentering feature. In this case you can use VSM. Use the exact same centering conditions, making sure that you are centering the sample (when there is sample). Then subtract the sample holder contribution. This can work, but if you forget at any line of the sequence one line that changes the centering you will might end up with problems again.

3- Measure using the DC method, and reconstruct one by one the fit of the V vs z curves that give you each point of your measurement. This method can be quite tricky. I do not advise.