# Nonparametric Hypothesis Testing Paper

Only available on StudyMode
• Published : May 10, 2011

Text Preview
Nonparametric Hypothesis Testing Paper
ABC’s real estate agency has recently expanded its business and is in the process of conducting research on housing prices within 10 miles of its new office. Team B has been given the task by ABC real estate to conduct the needed research. The team will be able to answer at the end of the research if the prices in Santa Cruz, California, are less than the prices of house in Scott’s Valley, California. Throughout this research process the team will formulate the hypothesis statement, perform a five-step hypothesis test, discuss which nonparametric test was needed, and discuss the results from this data research compare to past research completed. Five-Step Hypothesis Test

The five-step hypothesis test will determine if consumers have a greater satisfaction when the consumer(s) pay more for the home. Step one is to state the null hypothesis and alternative hypothesis: Ho: µ1 = µ2 No difference in satisfaction

Ho: µ1≠ µ2 Satisfaction differs for the two groups

Step two evaluates if the assumptions have been met through the Mann-Whitney test selection that will compare the two populations and will assume equal variances with a 95% confidence level, the significance level is .05. The decision rule is to reject Ho if z>1.645. Step three the ranks are summed and calculated using the Wilcoxon – Mann-Whitney test. T1 = 264.5 and T2 = 244.5. The sum T1 + T2 must be n(n + 1)/2 where n = n1 + n2 = 18 +20 = 38. Since n(n + 1)/2 = (38)(39)/2 +741 and the sample sums are T1 + T2 = 264.5 + 244.5 = 509, our calculation check. α = 0.05. The results of the Mann-Whitney test:

Wilcoxon – Mann-Whitney Test

n sum of ranks
18384 Group 1
20357 Group 2
38741 total

351.00 expected value
34.11 standard deviation
0.95 z, corrected for ties
.3407 p-value (two-tailed)

Step four is to make a decision to reject Ho if z>1.645.
Step five is the conclusion based on the...