Consider the following scenario. An experimenter collects a random matched-pairs sample of binary data (1 or 0) and would like to conduct a sign test. Is a sign test a sensible choice? Explain why or why not.
NO.It is not sensible to use a sign test in this case because the basic assumption of a sign test is that the random variable X should be continuous and here it is a pair of binary numbers. Also sign test is a substitute for a one sample t test which we use to test the mean of population where we have a value of Mu and we accordingly give signs to the sample values as '+' if the value is greater than Mu or '-' if the value is negative. Here as we do not have value of Mu nor we have the continuous values of sample variables sign test cannot be used.
Get Answers For Free
Most questions answered within 1 hours.