normality test for extracted factor.

134 views
Skip to first unread message

Jugal Kishor

unread,
Jan 18, 2014, 12:51:20 AM1/18/14
to dataanalys...@googlegroups.com
Respected sir,
                       Good morning sir. Pls tell me Do we need to test the normality of the extracted factor from factor analysis for further analysis of these extracted factors.

--
Thanks & Regards
Jugal Kishor
Research Scholar
FMS, Gurukula Kangri University
Haridwar
Mob- 7417949125

Neeraj Kaushik

unread,
Jan 18, 2014, 1:28:33 AM1/18/14
to dataanalysistraining
Yes u shd check normality of the factors using 1 sample KS test


--
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.

Amarbir Bhalla

unread,
Jan 20, 2014, 12:49:47 AM1/20/14
to dataanalys...@googlegroups.com
Respected Neeraj Sir,
Good Day!!

Sir, we need to check the normality of the factors extracted even if
the data has been measured on Likert Scale?
Regards,
A.S.Bhalla
Research Scholar,
Department of Commerce
and Business Management,
Guru Nanak Dev University,
Amritsar.

Contact No. +919463169631

Neeraj Kaushik

unread,
Jan 20, 2014, 1:31:41 AM1/20/14
to dataanalysistraining
Dear ASB
Yes the extracted factors (from EFA) need to be checked for the normality coz Likert scale does not guarantee normality.
I've seen may cases where the factor scores were not normally distributed.
Best wishes
Neeraj

Amarbir Bhalla

unread,
Jan 20, 2014, 3:03:53 AM1/20/14
to dataanalys...@googlegroups.com
Ok Sir, Thank you very much for your Invaluable guidance.
Regards.
A.S.Bhalla
Research Scholar,
Department of Commerce
and Business Management,
Guru Nanak Dev University,
Amritsar.

Contact No. +919463169631


jugal77

unread,
Jan 20, 2014, 11:57:05 PM1/20/14
to dataanalys...@googlegroups.com
thank you so much sir fro your immediate help.

Jugal Kishor

unread,
Jan 21, 2014, 3:07:48 AM1/21/14
to dataanalys...@googlegroups.com
                                               
Tests of Normalitya
Age Kolmogorov-Smirnovb Shapiro-Wilk
Statistic df Sig. Statistic df Sig.
FACTOR1 18-19 yrs .170 6 .200* .961 6 .831
20-21 yrs .190 111 .000 .900 111 .000
22-23 yrs .185 628 .000 .810 628 .000
24 yrs or older .186 3254 .000 .803 3254 .000
*. This is a lower bound of the true significance.
a. FACTOR1 is constant when Age = 17 yrs or younger. It has been omitted.
b. Lilliefors Significance Correction

Gd afternoon sir,
                         I have a confusion whether the above spss output is normally distributed or not because Shapiro-wilk sig. is greater than .05 for 18-19 age group but for others this is less tha .05. This is the normality test after factor analysis on extracted factor1. according to the table data is normally distributed for 18-19 yeras but not for others group of age. Should I consider this as normally distributed or Not normally distributed. how can I use Anova or H-test on this as You said earlier.




--
You received this message because you are subscribed to a topic in the Google Groups "DataAnalysis" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/dataanalysistraining/wh1kPGRm4js/unsubscribe.
To unsubscribe from this group and all its topics, 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.

Neeraj Kaushik

unread,
Jan 21, 2014, 7:41:44 AM1/21/14
to dataanalysistraining
Dear JK
You're checking the normality for the each age group.
Plz check the normality of the dep var alone (not in group)
Plz use Analyze->Non parametric test->Legacy Dialog->1 sample KS test
Just send dep var to RHS pane and click ok
If Sig value>0.05, factor is normally distributed else not. {Plz note output will give u many rows, just see the last row of the table depicting sig value. Secondly just below the table it will be written: Test distribution is Normal. It is not the result/conclusion but the Null hypothesis for this table}
Best wishes
Neeraj


--
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.

Jugal Kishor

unread,
Jan 22, 2014, 12:56:44 AM1/22/14
to dataanalys...@googlegroups.com
thank you very much sir, I got the solution. d way u guide its awesome. :) pls keep showering your blessings.  
Reply all
Reply to author
Forward
0 new messages