Yes you can do. You need to arrange the data in a matched set up and look for p pvalue for the departures from symmetry. It is an extension of a Mc-Nemar Chi-square test, readily vailable in Stata.

Your Table should look like one as shown in the attachment. If there is considerable change in the frequencies from pre to post, there should be poor symmetry, showing very low p values (significant chi-square).

Good luck

Sreenivas

Thanks Sree i arranged the data and did the calculations with p-values

Is this statistical test also available in SPSS? Does the test have a official title to use when reporting it in a publication?

The test is a standard procedure to show whether your results are statistically significant or not?? The test will have to mentioned as a Chi-square test if you are planning to write a Article for Publication...The test is available in SPSS , Med Calc,R software . In these software you will have to do some minor tweaks to get your analysis done . Or if you don't want to go in these soft wares go for online calculators which easily gives you the desired results...

yes you can do such analysis. SAS and Stata are freindly packages. it can be done even with SPSS but it might be tedious

Hello - if the contingency table values are known (or you can count them if the data set is relatively small), you can do the analysis directly online at www.statpages.org

@Pallavi it seems you are one of those rare species ( Hard to find ) BWB , I will certainly be doing lots of data dredging in my thesis , so will need help frm every where , hoping fr the same .............

Sure...

The extension to the McNemar test (McNemar-Bowker test) is available easily in SPSS using the raw data from the various time points. Use the Crosstabs procedure and select the McNemar option under the Statistics button. For tables larger than the normal 2X2 for a McNemar test SPSS switches automatically to the Bowker test (see info in the Help menu).