iSixSigma

Query

Viewing 4 posts - 1 through 4 (of 4 total)
  • Author
    Posts
  • #44798

    Alok
    Participant

    Hi
    Suppose i want to know whether a person clearing an exam is any way related to his age , qualification and experience . These are all non metric variables ,so how to go about in establishng any co relation or how to use multiple regression or any other statistical tool to use . also if i am doing a hypothesis testing and my null hupothesis is that a mean of a sample of people not clearing any exam is  realted to age then how to go about testing this hypothesis .
     
    Regards
    Alok

    0
    #144321

    SH
    Member

    Can u give some more details regarding this..
    1. Person clearing an exam, (do u have your responses as Yes/No, Pass Fail or Exam % like 93%,45%,and then a cut off %)
    2.Have u categorized age in buckets like 15-20 yrs, 20-25 yrs or you have taken the age in figures as 21, 22.5, 44 etc), same for the experience.
     

    0
    #144341

    Alok
    Participant

    For person clearing an exam it is categorised as Pass or fail
    age has been categorised as buckets like 10-12 yrs 12-14 yrs
    same holds true for the experience
    But for the qulaification how can i categorise i am thinking of catergorising them as below 10th , 12th pass , graduate , post graduate
    Pls help me as its impotant for me to establish a co realtion or devlop a model takinh these X”s into account and see what effect it has on Y which is the attrition.

    0
    #144355

    SH
    Member

    Alok,
    Your post suggest that  you want to know whether a person clearing an exam is any way related to his age , qualification and experience
    You have age and experience categorized in buckets and have pass/fail criterion.
    You can proceed with the qualification data as you have stated.
    Your Y here would be Pass/fail and your Xs are age, experience and qualification.
    You can use Chi Square test for each of the Xs and determine if there is any difference between Pass/fail and your various xs.
    You null hypothesis for the test would be:
    “Proportion of no of people scoring Pass or Fail is independent of the Age buckets classified into”
     Hope this helps.
     

    0
Viewing 4 posts - 1 through 4 (of 4 total)

The forum ‘General’ is closed to new topics and replies.