Skip to main content
SAGE
Search form
  • 00:04

    HERSCHEL KNAPP: Welcome to Practical Statisticsfor Nursing Using SPSS.This video shows how to process the MANOVA test.You can watch the entire video, or use the time sliderto navigate to any time point.[MANOVA Video]The MANOVA test is quite similar to the ANOVA test.

  • 00:25

    HERSCHEL KNAPP [continued]: Before proceeding, it's recommended that you first viewthe video Ch06-ANOVA.mp4.[MANOVA Overview]In terms of setup and results, the MANOVA an ANOVA testsare quite similar.We've already seen that the ANOVA test can compare

  • 00:45

    HERSCHEL KNAPP [continued]: the effects of three different drugs on systolic bloodpressure.But what if we also wanted to seehow these drugs affected another dependent variable,such as pulse rate?Whereas ANOVA can process only one dependent variable,such as systolic blood pressure, MANOVAcan process additional dependent variables, such as pulse rate.

  • 01:08

    HERSCHEL KNAPP [continued]: In this example, we'll see that the results of the MANOVA testwill look like we ran two separate ANOVA tests, onefor systolic blood pressure, and another for resting pulse rate.[MANOVA Pretest Checklist]The MANOVA test has five pretest criteria, n quota, normality,

  • 01:31

    HERSCHEL KNAPP [continued]: moderate correlation, homogeneityof variance-covariance Box's M Test,and homogeneity of variance Levene's test.We'll check for the first three now.We'll see the results of the remaining twowhen we run the MANOVA test.This example uses the data set Ch08-Example01-MANOVA.sav.

  • 01:58

    HERSCHEL KNAPP [continued]: This data contains three variables.Group is a categorical variable thatcontains three values, Drug A, Drug B, and Drug C. Systolic BPis a continuous variable that contains the systolic bloodpressure of each participant at the conclusion of the study,

  • 02:18

    HERSCHEL KNAPP [continued]: and pulse is a continuous variablethat contains the resting pulse rate of each participantat the conclusion of the study.We'll begin by checking for the n quota.Click on Analyze, Descriptive Statistics, Frequencies.Move group to variables and click OK.

  • 02:42

    HERSCHEL KNAPP [continued]: The frequency table indicates 30 participants per group,hence this criterion is satisfied.Next we'll check for normality, starting with Drug A, Group 1.Click select cases, click if condition is satisfied,click if.

  • 03:02

    HERSCHEL KNAPP [continued]: In the dialog box, enter group equals 1.Click Continue.Click OK.Now order the histogram with a normal curve for systolic BPand pulse.Click Analyze, Descriptive Statistics, Frequencies.

  • 03:28

    HERSCHEL KNAPP [continued]: Move systolic BP and pulse into variables.Click charts and select histogram with normal curve.Click Continue, click OK, and it'll process.The histograms for systolic BP and pulse

  • 03:48

    HERSCHEL KNAPP [continued]: are normal for Group 1.Next repeat the process for Drug B, Group 2.Click select cases, click if.In the dialog box, enter group equals 2.Click Continue.

  • 04:09

    HERSCHEL KNAPP [continued]: Click OK.Now order the histogram with a normal curve for systolic BPand pulse.Click Analyze, Descriptive Statistics, Frequencies.Click OK, and it'll process.The histograms for systolic BP and pulse

  • 04:31

    HERSCHEL KNAPP [continued]: are normal for Group 2.Now repeat the process once again for Drug C, Group 3.Click select cases, click if.In the dialog box, enter group equals 3.Click Continue.

  • 04:53

    HERSCHEL KNAPP [continued]: Click OK.Now order the histogram with normal curve for systolic BPand pulse.Click Analyze, Descriptive Statistics, Frequencies.Click OK, and it'll process.The histograms for systolic BP and pulse

  • 05:13

    HERSCHEL KNAPP [continued]: are normal for Group 3.The normality criterion is satisfied.Before proceeding, reactivate all the records.Click on filter underscore dollar sign,and press the Delete key.Now we'll check for moderate correlationbetween the two continuous variables, systolic BP

  • 05:36

    HERSCHEL KNAPP [continued]: and pulse.Click Analyze, correlate bivariate.Move systolic BP and pulse to variables, and click OK.The correlation between systolic BP and pulse is 0.339.

  • 05:59

    HERSCHEL KNAPP [continued]: Since this is between 0.3 and 0.9,this satisfies the criteria of moderate correlation.[MANOVA Test Run]To run the MANOVA test, click on Analyze, general linear model,multivariate.Move systolic BP and pulse to dependent variables,

  • 06:24

    HERSCHEL KNAPP [continued]: and move group to fixed factors.Click on Post hoc, move group to post hoc tests.On the equivalent variance assumption option,check Bonferroni.Click Continue.Click options.

  • 06:46

    HERSCHEL KNAPP [continued]: Move group to Display Means, and check homogeneity tests.Click Continue.Click OK, and it'll process.[MANOVA Results]To finalize the pretest checklist,we look to boxes test of equality

  • 07:07

    HERSCHEL KNAPP [continued]: for covariance matrices, which rendered a p value of 0.703.Since this is greater than 0.001,this indicates that there is no significant differencebetween the variance covariances in this model,hence this pretest criteria is satisfied.

  • 07:29

    HERSCHEL KNAPP [continued]: Levene's test of equality of error variancesshows p value of 0.656 for systolic BP,and 0.765 for pulse.Since these p values are greater than 0.05,this indicates that there is no statistically significantdifference among the variances for the two

  • 07:50

    HERSCHEL KNAPP [continued]: dependent variables, hence this criterion is satisfied.Proceeding with the results of the MANOVA test,we look to the multivariate test table.We see that Pillai's trace rendered a p value of 0.003.Since this is less than or equal to 0.05,this tells us that a statistically significant

  • 08:12

    HERSCHEL KNAPP [continued]: difference has been detected among the pairs of groupsfor the outcome variables systolic BP and pulse.Finally, we look to the multiple comparisons tableto determine which group performed statisticallysignificantly different from which.The estimates table provides us with the means for each group

  • 08:34

    HERSCHEL KNAPP [continued]: for systolic BP and pulse.These figures will be useful when documenting the results.Notice that the multiple comparisons table essentiallyresembles to ANOVA multiple comparisons tables, whereinthe top half of the table shows the three pairwise comparisons

  • 08:55

    HERSCHEL KNAPP [continued]: for systolic BP, and the bottom halfshows a separate set of comparisons for pulse.To identify which groups outperformwhich for each of the variables, begin by identifyingthe pairs of groups that have a p value that's less thanor equal to 0.05 for systolic BP.Then do the same for pulse.

  • 09:19

    HERSCHEL KNAPP [continued]: This concludes this video.

Video Info

Series Name: Practical Statistics for Nursing Using SPSS

Episode: 9

Publisher: SAGE Publications, Inc.

Publication Year: 2016

Video Type:Tutorial

Methods: Multiple analysis of variance, SPSS

Keywords: blood pressure; mathematical computing; mathematical concepts; nursing; pulse; Software ... Show More

Segment Info

Segment Num.: 1

Persons Discussed:

Events Discussed:

Keywords:

Abstract

Chapter 9 of the nursing statistics with SPSS series outlines the principles of MANOVA. Professor Herschel Knapp demonstrates analysis of multiple dependent variables and provides a guide to generating effective results with the software.

Looks like you do not have access to this content.

MANOVA

Chapter 9 of the nursing statistics with SPSS series outlines the principles of MANOVA. Professor Herschel Knapp demonstrates analysis of multiple dependent variables and provides a guide to generating effective results with the software.