Exactly as described, for the latency and PDV tests, this sets the interval, in units of 100 milliseconds, to wait between each of the configured samples. Combined with the max-samples setting, this value determines the total wall-clock time the latency and PDV test portion actually takes and how spread out in time the samples are, which matters for catching intermittent, time-varying latency issues like periodic congestion versus taking all samples in a tight burst that might miss them. For example, a sampling interval of 50, or 5 seconds, combined with 20 max samples would spread the latency measurement across a full 100 seconds rather than measuring everything in a two-second burst.