Scribd
Upload
1K views20 pages

Median Test for Comparing Two Medians

This document discusses how to use the median test to make inferences about the difference between the medians of two populations. The median test involves taking two independent random samples, finding the grand median, classifying observations as above or below the median, and constructing a contingency table. A decision is made by comparing the test statistic to critical values. If the test statistic is in the critical region, the null hypothesis that the population medians are equal can be rejected, supporting the alternative that the medians are different. An example applies these steps to compare stroke index values between two patient groups.

Uploaded by

Rohaila Rohani
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
1K views20 pages

Median Test for Comparing Two Medians

This document discusses how to use the median test to make inferences about the difference between the medians of two populations. The median test involves taking two independent random samples, finding the grand median, classifying observations as above or below the median, and constructing a contingency table. A decision is made by comparing the test statistic to critical values. If the test statistic is in the critical region, the null hypothesis that the population medians are equal can be rejected, supporting the alternative that the medians are different. An example applies these steps to compare stroke index values between two patient groups.

Uploaded by

Rohaila Rohani
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
  • Introduction
  • Learning Outcome
  • Assumptions for Median Test
  • Median Test
  • Hypotheses for Median Test
  • Steps to Conduct Median Test
  • Contingency Table
  • Example Table
  • Sample Classification and Assumptions
  • Contingency Table Example
  • Making Decision
  • Conclusion
  • End Note

MAKING INFERENCES ABOUT THE DIFFERENCE BETWEEN TWO LOCATION PARAMETERS BY MEDIAN TEST

Learning Outcome

Student should be able to make inferences about the difference between two location by using Median Test.

MEDIAN TEST
The simplest and most widely procedures for testing the null hypothesis of two populations that have the same median.
Only two-sided alternatives will be discussed.

Assumptions for Median Test


1) The data consist of two independent random

samples : X1, X2, X3, ..., X n, and Y1, Y2, Y3, ..., Y n 2) The first sample is from a population with unknown median, MX and the second sample is from a population with unknown median, MY. 3) The measurement scale employed is at least ordinal. 4) The variable of interest is continuous. 5) The two populations have the same shape. 6) If the two populations have the same median, then for each population the probability p is the same that an observed value will exceed the grand median.

Hypotheses for Median Test H0 : H1 : MX = M Y MX M Y

Step 1 : State the hypothesis: H0: MX=MY H1: MXMY Step 2: Find the grand median from the sample values.
Step 3 : Classify each sample observation according to two criteria: whether it belongs to sample X or sample Y whether it is above or below the computed sample median Step 4 : Draw the contingency table.

CONTINGENCY TABLE
SAMPLE RELATION ON THE SAMPLE ABOVE X A Y B TOTAL A+B

BELOW

C+D

n1 = A+C

n2 = B+D

N=n1+n2

Where, A : The number of observations from sample X falling above the median B : The number of observations from sample Y falling above the median C : The number of observations from sample X falling below the median D : The number of observations from sample Y falling below the median

TABLE 3.2 Stroke-index values, milliliters, for patients admitted to the myocardial-infarction research unit of a university hospital
Diagnosis

Anterior transmusal infarction and anterior necrosis(X)


25 25 17 26 18 30 24 13 30 20 23 26 12 20 9 17 37 20 11 32 16 46 20 25 17 36 54 8

Interior transmural infarction and interior necrosis (Y)


31 21 38 19 38 41 68 43 42 30 20 29 13 32

21

37

31

26

28

30

Step 1: Classify the sample observation as sample X and Y.


Anterior transmural infarction and anterior necrosis as sample X Interior transmural infarction and interior necrosis as sample Y H0:MX = MY H1:MX MY (claim)

Assumption:
This data consists of two independent random samples X and Y.

Step 4: Draw the contingency table.


Anterior transmural infarction and anterior necrosis
12 20 32

Relationship to 25.5

Inferior transmural infarction and interior necrosis


12 4 16

Total

Above Below Total

24 24 48

Step 6: Making decision.


We are using standard normal distribution and the critical value are 1.96. Since -2.45 are in the critical region, so we reject the null hypotheses. By using P-value method we get: 2(0.5 - 0.4929) = 0.0142

Step 7: Conclusion.
Since we reject the null hypothesis, we have enough evidence to support the claim that the two population medians are not equal.

THANKS FOR WATCHING!!


[Link] pju_nbB0i8&feature=[Link]

MAKING INFERENCES ABOUT THE DIFFERENCE BETWEEN TWO LOCATION PARAMETERS BY MEDIAN TEST
Learning Outcome • Student should be able to make inferences about the difference between two location by usi
MEDIAN TEST • The simplest and most widely procedures for testing the null hypothesis of two populations that have the s
Assumptions for Median Test 1) The data consist of two independent random samples : X1, X2, X3, ..., X n, and Y1, Y2, Y3,
Hypotheses for Median Test H0 : MX = MY H1 : MX ≠ MY
Step 1 : ►State the hypothesis: H0: MX=MY H1: MX≠MY Step 2: Find the grand median from the sample values. Step
SAMPLE RELATION ON THE SAMPLE X Y TOTAL ABOVE A B A+B BELOW C D C+D n1 = A+C n2 = B+D N=n1+n2 Where, A : Th

You might also like