You can still use the t-test. Only for very small sample sizes you have to ensure normal distribution. Another option would be the Mann-Whitney-U-Test, which is more appropriate on non-normal distributed data.
For salary - as stated before, Independent t test (considering that te conditions are met) , and for designation you have to know the type of data. If it's categorical (A, B, C etc) then probably a Chi-square test is what u need (in SPSS you go to Analyis>Crosstab> Options). This will test if a certain gender is associated with a certain designation.
I was just asking, because you started with a incorrect statement. The t test relies on some assumptions to give valid results, but a sample size below a fixed number is none of them.
Gary is correct, Dhruba. But if you really want another test, Z test of arithmetic mean difference between two large independent samples will do, You might be surprised to find out that the results of the t test (between two independent samples) and that of Z test will be the same.
This method may not be available in SPSS, though. Ed
Hi, Im looking for any instrument to study employee salary effecting retention, if you could share your instrument will be wonderful ([email protected])