Download INFERENTIAL STATISTICS II: TESTING HYPOTHESIS ON TWO

TESTING HYPOTHESIS ON TWO SAMPLE MEANS BY USING MICROSOFT EXCEL
©2016
TESTING HYPOTHESIS ON TWO SAMPLE MEANS BY USING MICROSOFT EXCEL
Learning Objectives:



How to state null and alternative hypotheses for comparisons between means & How to interpret the
results based on p-value (Exercise 1 & 2)
t-Test for Two Independent Samples (Exercise 3)
Paired-Samples t-Test (Exercise 4)
EXERCISE 1
1.
2.
Create a new Word Document named Test-by-examples and save it in Lab09 folder.
A study was conducted to “investigate the effect of continuous training programme on blood pressure and serum
creatinine concentration in black African subjects with hypertension” (Sikiru L, Okoye GC. Therapeutic effect of
continuous exercise training program on serum creatinine concentration in men with hypertension: a randomized
controlled trial. Ghana Med J. 2014;48(3):135-42.). Copy the following image extracted from this article:
For this request, the answers are already provided as example on how the
answer is supposes to look like.
a.
b.
For SBD (mmHg) variable (where SBD = systolic blood pressure):
i. State the null and alternative hypotheses:
 H0 (null hypothesis): The pre-test mean of SBP in continuous group is not significantly
different by the pre-test mean of SBP in control group.
 H1 (alternative hypothesis – two-sided test): The pre-test mean of SBP in continuous
group is significantly different by the pre-test mean of SBP in control group.
ii. Write the name of the test used for comparison:
 Student t-test for independent samples.
iii. Provide the values of means and standard deviation for each group and interpret from statistical
point of view the presented result
 Continuous group: 170.45±15.57 mmHg
 Control group: 160.87±13.23 mmHg
 p = 0.597  Since the obtained p-value is higher than 0.05 we fail to reject the null
hypothesis and conclude that there is no significant differences in regards of pre-test
SBP between continuous group compared to control group.
For DBP (mmHg) variable (where DBP = diastolic blood pressure):
i. State the null and alternative hypotheses
ii. Write the name of the test used for comparison
iii. Provide the values of means and standard deviation for each group and interpret from statistical
point of view the presented result
-1-
TESTING HYPOTHESIS ON TWO SAMPLE MEANS BY USING MICROSOFT EXCEL
©2016
EXERCISE 2
1.
2.
Insert a new page on Test-by-examples file using Page Break option.
A study was conducted to investigate the hemodynamic changes in two variables, named pulmonary artery
pressure and pulmonary vascular resistance index by comparison of baseline measurements with measurements
after application of inhalated nitric oxide (iNO) and respectively inhaled treprostinil (Vorhies EE, Ivy DD. Drug
Treatment of Pulmonary Hypertension in Children. Paediatr Drugs 2014;16(1):43-65.). Copy the following image
extracted from this article in the inserted new page:
a.
For Pulmonary artery pressure variable: baseline vs. iNo
i. State the null and alternative hypotheses:
 Null hypothesis (H0): The mean of pulmonary artery pressure at baseline is not
significantly different by the mean of pulmonary artery pressure after iNO.
 Alternative hypothesis (Ha): The mean of pulmonary artery pressure at baseline is
significantly different by the mean of pulmonary artery pressure after iNO.
ii. Write the name of the test used for comparison under the assumption that the samples are
paired: Paired t-test
iii. Write the obtained p-value and give the interpretation of this value:
 p <0.01  Since the obtained p-value is less than 0.05 we reject the null hypothesis
and conclude that there is a significant decrease in the mean of pulmonary artery
pressure after iNO treatment
b.
For Pulmonary artery pressure variable: iNo vs. inhaled treprostinil (NS = not significant)
i. State the null and alternative hypotheses:
ii. Write the name of the test used for comparison under the assumption that the samples are
paired:
iii. Write the obtained p-value and give the interpretation of this value:
c.
For Pulmonary vascular resistance index variable: baseline vs. inhaled treprostinil (NS = not significant)
i. State the null and alternative hypotheses:
ii. Write the name of the test used for comparison under the assumption that the samples are
paired:
iii. Write the obtained p-value and give the interpretation of this value:
-2-
TESTING HYPOTHESIS ON TWO SAMPLE MEANS BY USING MICROSOFT EXCEL
©2016
PROBLEM – TO BE USED FOR EXERCISE 3 & 4
A study was conducted to assess two different therapeutic schemas for treatment of ferriprive anemia.
Newborn from rural and urban environments were included in the study. Two different schemas were
studied: a daily schema (mother administrated the treatment once a day), and a bi-weekly schema (mother
administrated the treatment on Monday and Thursday). The collected data are in the Anemia.xlsx file.
EXERCISE 3: IS THE MEAN OF HAEMOGLOBIN AT 6 MONTHS DIFFERENT FOR PATIENTS
WITH DAILY TREATMENT COMPARED TO PATIENTS WITH BI-WEEKLY TREATMENT?
1. Download the Anemia.xlsx file and save it in Lab09 folder. The analysis is conducted under the
assumption of normal distribution of quantitative variables.
2. Create a new sheet named Exercise 3. Copy in this sheet the value of "Haemoglobin (mg/dl) 6 months",
and “Treatment schema” columns. Sort data ascending by Treatment schema variable.
3. Apply the F-test [Data – Analysis – Data Analysis – F-test Two-Sample for Variances] to test if the
variances of Haemoglobin at 6 months for patients with daily treatment and bi-weekly treatment are
equal or not.
a. Put the results at the right of the columns with the data.
b. Under the table with the results, state the null and alternative hypotheses.
c. Interpret the obtained results (p-value). Which test on means is appropriate to be applying
according with the interpretation of p-value? Write the name of the proper test using the
possible tests implemented in Excel (see the image bellow):
4. Apply the correct t-test for independent samples – two tail (according with the results obtained on Fvariance tests) - Haemoglobin (mg/dl) 6 months. Display the results at the right of the results on F-test.
5. Under the resulted table:
a. State the null and alternative hypotheses
b. Write the statistic of the test and associated p-value (two-tail test)
c. Interpret the obtained result from statistical and clinical point of views.
EXERCISE 4: IS THE MEAN OF HAEMOGLOBIN AT 6 MONTHS DIFFERENT BY THE MEAN
OF HAEMOGLOBIN AT 12 MONTHS FOR PATIENTS WITH DAILY TREATMENT?
1. Create a new sheet named Exercise 4. Copy in this sheet the value of "Haemoglobin (mg/dl) 6 months",
“Haemoglobin (mg/dl) 12 months”, and “Treatment schema”.
2. Sort the data according with Treatment schema.
3. Apply the Paired T-test in order to find the answer for the subjects who received Daily treatment. Put
the results near the column with the data.
4. Under the table with the results:
a. State the null and alternative hypotheses.
b. Write the value of statistic and its associated p-value.
c. Interpret the obtained result from statistical and clinical point of views.
-3-