Question

Problem 3: [5 pts.]In order to investigate if the systolic blood pressure measurements vary in standing and lying positions the systolic blood pressure levels of sample of 12 persons were measured in both positions (first in standing position and then in lying position). The data of this experiment are recorded in SSPS file Assn1Q3.sav.

1. A researcher believes the mean systolic blood pressure level in standing position is more than that in lying position. Test the appropriate hypothesis to investigate the researcher’s belief at alpha=0.05.

2. Find the 95% confidence interval for the difference in mean systolic blood pressure levels in two positions. Interpret the result.

Please mentioned the steps in SPSS

The table is for the data.

Lying	Standing
132.00 146.00 135.00 141.00 139.00 162.00 128.00 137.00 145.00 151.00 131.00 143.00	136.00 145.00 140.00 147.00 142.00 160.00 137.00 136.00 149.00 158.00 120.00 150.00

Answer 1

First you have to note down all the values in a single column( both lying and standing) and then you have to define a grouping variable where there will be two distinct values- "Lying" and "Standing". So, in the variable "Group", which is a grouping variable, you should have 12 values as "Lying" and 12 values as "Standing" corresponding to the original values of those groups. Your data should look like this.

132.00 Lying

146.00 Lying

135.00 Lying

141.00 Lying

139.00 Lying

162.00 Lying

128.00 Lying

137.00 Lying

145.00 Lying

151.00 Lying

131.00 Lying

143.00 Lying

136.00 Standing

145.00 Standing

140.00 Standing

147.00 Standing

142.00 Standing

160.00 Standing

137.00 Standing

136.00 Standing

149.00 Standing

158.00 Standing

120.00 Standing

150.00 Standing

Now, you have to go to SPSS>Analyze>Compare Means>Independent Samples t-test. Populate the dependent variables as measurement variable and Independent variable as "Group". You will get an option to define the grouping variables. You have to assign any one of the "Lying" or "Standing" to group 1 and other to group 2. Now press OK and run. You should have a table like the following,

Independent Samples Test 95% CI

F sig t df sig(2-tail) Mean diff std err diff Lower Upper

Lying Equal variances assumed .152 .700 -.601 22 .554 -2.50000 4.15817 -11.12352 6.12352

Equal variances not assumed -.601 21.629 .554 -2.50000 4.15817 -11.13211 6.13211

I have assumed Lying as group 1 and Standing as group 2

Now note that you won't get the answer automatically from SPSS as you do not have any option to do one-tailed independent sample t test in SPSS. ( You need one tailed t-test for Q1 as your alternative hypothesis is BP at standing is higher than that at Lower). For this you will calculate the t-statistic, which will be mean diff/std err diff=-2.5/4.15817=-0.60122602, df=22

Now, you will go to Transform>Compute Variable. In target variable write some name, it does not matter. Now, among the functions select CDF and Non-central CDF and then CDF.T Double-click on it. Write -0.60122602 as the first qty and 22 as the second qty. You will find sig level as 0.28 which indicates null hypothesis is true at 5% level. Do not use the sig level 0.554 in the table as it is two tailed.

Also, you need not use the row Equal variance not assumed from the table, since sig for the assumption of equality in variability of the two groups is 0.7, indicating the assumption to be true.

95% confidence interval indicates that, we cant say that BP at standing position is higher than that of the lower.