Similar words for similar concepts turn up in many widely scattered languages. Some linguists say this is chance while others claim that many if not all of the world's languages descend from a single prehistoric language. Yet neither position has been analyzed or supported with statistics. Computerized statistical techniques can be used to help determine whether or not words in different languages have ancestral connections. These techniques are explained and broken down to provide the necessary principles for those linguists with no background in statistics. This methodology measures the probabilistic significance of sound correspondences between short word lists. Many rules of thumb used to obviate chance resemblance are shown to decrease the power of quantitive testing. The procedures presented here are straightforward, but the author also presents the extensive linguistic work needed to produce word lists that will not yield nonsensical results. Examples analyze 200 words in eight languages.