I have collected information through the 5 and 7 point likert scale questionnaire, through the Google drive, but I am little bit confused which statistical tool to use for analysis.
Before worrying about what statistical test to use, you should really have a good understanding of your research question and any hypotheses you wish to test. Once you know these, then deciding on how to analyse the data becomes much clearer. The fact that you have Likert scales is only a marginal consideration compared to the above.
Some statistical tools to be used to test the 5 or 7 point likert scale questionnaire in SPSS other than factor analysis is also avaiable. They are regression analysis and one sample t-test, reliability and validity
Before worrying about what statistical test to use, you should really have a good understanding of your research question and any hypotheses you wish to test. Once you know these, then deciding on how to analyse the data becomes much clearer. The fact that you have Likert scales is only a marginal consideration compared to the above.
I concur with Adrian. Please provide some information about your question, as this will inform which tests might be most appropriate.
Also, whether the data mean is something worthwhile depends on your interpretation of the Likert scales and your specific subfield. In psychology, for example, it's common to work with the means from Likert scale data. In other fields, the median is more accepted. If the median would make more sense for your data, then the non-parametric statistics will work if you are making comparisons across groups. Some specific tests would include the Chi Square, Mann-Whitney U (for comparing two different groups), Wilcoxon-t test, and Kruskal-Wallis.
Hi Nuzhath (and it's fine to call me "Jay" here as we're all colleagues on research gate).
It seems to me that you're working with a descriptive study using a survey for the data. In this case, using correlations would make sense. Specifically, you should use Spearman's Correlation Coefficient instead of the standard Pearson Correlation Coefficient for Likert scale data.
You could perform more interesting analyses if your data also has some demographic information. For example you could examine differences in MBA graduates' perspectives based on age, gender, income, or other similar variables. If you group your participants based on any of these, then many non-parametric statistics will work well. For example, you could compare female versus male MBA graduates' perspectives using a Mann Whitney U test, which is easy to work with in SPSS.
Please let me know whether you have any demographic information and what data you do have. Then, I can better suggest some statistics.
I completely agree to everything stated by Jay and Adrian. Particularly, much more information can be obtained from investigations of interactions with other factors. If you do this, the data type of the other variable you would like to investigate to be correlated matters. If you investigate an interaction ordinal - ordinal (e.g. Likert scale - Likert scale or Likert scale - years of education), gamma is the best for measurement of strength of ordinal association (https://learn.bu.edu/bbcswebdav/pid-826908-dt-content-rid-2073693_1/courses/13sprgmetcj702_ol/week05/metcj702_W05S03T04_gamma.html), I think. Parametric tests (e.g. Pearson) are a no-go and non-parametric tests (Mann-Whitney-U or Kruskal-Wallis) work only if you have a low percentage of ties. If you wish just to describe your data, MODE might be appropriate.
I guess you can use PCA (Principal Component Analysis).
With PCA you can transform the information you have collected (several-item scale, I suppose) in a much smaller number of variables.
Latent class analysis is another option. However it is less frequently used and it is not implemented in SPSS. But I suggest you give it a try using an open source software like R.