Claim A: In Karn/Partridge algorithm to compute the timeout, if the variable of sampleRTT is small, it is an indication that the timeout value should NOT be tightly coupled to the EstimateRTT. Is the Claim A is true or false? Please briefly explain.