How is the test error rate affected by k-fold cross validation




While I was learning about cv I got the understanding that cv better helps in getting an idea of the error rate in test sets due to repeated validation but:

i could not exactly understand the second part of the answer,
Can someone please help me with this??