[Koala]

Does the use of pipette in measuring the volume of solution reduce the random error in an experiment as compared to using measuring cylinder( which has larger uncertainty than pipette)?

[Borek]

Yes, see

http://www.titrations.info/pipette-burette

But remember that if you are using cylinder to add substances that just have to be in an excess, or in an approximate amount, it doesn't matter.

[Koala]

Does that mean that the tolerance of an instrument contributes to random error in an experiment? I read from some sources that tolerance of an instrument affects the precision of the experimental readings; and to reduce random errors, one needs to repeat the experiment several times to obtain average reading as the main approach rather than using instrument with smaller tolerance.

[Borek]

Does that mean that the tolerance of an instrument contributes to random error in an experiment?

No, it contributes to systematic, not random. If you have a pipette that is nominally 10.0 but 10.1 in reality, your average volume would be 10.1±random error, and 10.1 would be systematically +0.1.