Which statistics should be used

[Mallika Shekhar]

Dear Sir and all Group Member,

For comparing performance of two groups (like Gender: M/F; Background Urban/Rural) which statistics should be used. So that, it can be concluded that which group perform better than other.

Thanks

With Regards:

Mallika Shekhar

[Vikas Deep]

I think independent sample t test would give desired results. Neeraj Sir, what's your take please.

Vikas Deep

9876076000

--
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistrai...@googlegroups.com.
To post to this group, send email to dataanalys...@googlegroups.com.
Visit this group at http://groups.google.com/group/dataanalysistraining.
For more options, visit https://groups.google.com/groups/opt_out.

[ishani chopra]

If data is metric and normally distributed then we use t test if not then, when it is ordinal we use u test and when continuous then t test or KS test. Is it right Neeraj Sir ?

[ishani chopra]

For gender and residential comparisons , will not Chi square test be the best ?

[Neeraj Kaushik]

The choice of statistical tool/technique depends on the type of measurement of Dep var.

Plz explain how performance is measured?

If it is metric & normally distributed-> Use 2 indep sample t-test

If it is metric & not normally distributed-> Use Mann Whiteney U-test

If it is Ordinal -> Use Mann Whiteney U-test
If it is Nominal -> Use Chi sq test

Let me put it again regarding the doubts on Likert scale.

Working on Individual statement is Ordinal measurement

However Summing the statement as new var (This was what Likert suggested when he called his scale as Summated Rating Scale), then it is Interval measurement i.e. Metric (which may or may not be normally distributed, 1 sample KS test checks it)

Best wishes
Neeraj

[Mallika Shekhar]

sir
i hv taken sum of statements to measure performance. but hvnt checked
its normality.After checking this i will use statistics as per your
suggestion.thank-you sir.

and thanks to ishani mam n vikash sir