query on reliability test

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

query on reliability test

pooja chawda



Dear Sir,
 
I am having query on covariance matrix which is showing the warning like, "The determinant of the

covariance matrix is zero or approximately zero. Statistics based on its inverse matrix cannot be
computed and they are displayed as system missing values." i m having 36 variables and when
cronbach alpha calculated it was, Reliability Statistics
Cronbach's Alpha (.817) Cronbach's Alpha Based on Standardized Items (.806) after deleting 5 items
which were having very low (one is in minus value) the cronbach alpha increases to, (.848) and
áCronbach's Alpha Based on Standardized Items (.836) now what should i do with that warning? i am

looking for factor analysis as i am having altogather 80 variables in the different category of HRM
practices (Independent Variables), Firms Operational Performance, Firm performance (financial and
non financial) as well as business operation modes. How is this warning affect the results? Please
help me.á


Pooja Chawda
Researcher and Asst. Professor,
Indu Management Institute,
Baroda, Gujarat, India.








On Sun, Jun 17, 2012 at 1:13 PM, L-Soft list server at UGA (1.8d) <[hidden email]>
wrote:
     Sun, 17 Jun 2012 03:43:21

     Your ásubscription to áthe SPSSX-L álist (SPSSX(r) áDiscussion) has ábeen
     accepted.

     Please save this message for future áreference, especially if this is the

     first time you are subscribing to an electronic mailing list. If you ever
     need to leave áthe list, you will find the ánecessary instructions below.
     Perhaps ámore importantly, ásaving a ácopy of áthis message á(and of áall
     future subscription notices áfrom other mailing lists) in áa special mail

     folder will give you instant access to the list of mailing lists that you
     are subscribed áto. This may áprove very useful the ánext time you ágo on
     vacation and áneed to leave áthe lists temporarily so áas not to áfill up
     your ámailbox while áyou áare away! áYou should áalso ásave the á"welcome
     messages" from the álist owners that you will áoccasionally receive after

     subscribing to a new list.

     To send áa message to áall the people ácurrently subscribed to áthe list,
     just send mail to [hidden email]. áThis is called "sending mail
     to the álist," because áyou send ámail to a ásingle address áand LISTSERV
     makes ácopies áfor all áthe ápeople áwho áhave subscribed. áThis áaddress
     ([hidden email]) is áalso called áthe "list address." áYou must
     never send a command áto that address, as it would ábe distributed to all
     the ápeople áwho have ásubscribed. áAll ácommands ámust ábe sent áto áthe

     "LISTSERV *server* address," [hidden email].

     You may leave the list at any time by sending a "SIGNOFF SPSSX-L" command
     to [hidden email].

     For áinformation áabout ásubscription áoptions, send áa á"QUERY áSPSSX-L"

     command to [hidden email].

     Contributions sent to this list are automatically archived. You can get a
     list of the available archive files by sending an "INDEX SPSSX-L" command
     to [hidden email]. You can then áorder these files with a "GET

     SPSSX-L LOGxxxx" command, or using LISTSERV's database search facilities.
     Send an "INFO DATABASE" command for more information on the latter.

     This álist is áavailable áin digest áform. áIf you áwish áto receive áthe
     digested áversion áof the ápostings, ájust áissue áa SET áSPSSX-L áDIGEST
     command.

     More áinformation on áLISTSERV ácommands ácan be áfound áin the áLISTSERV
     reference ácard, which áyou can áretrieve áby sending áan "INFO áREFCARD"
     command to [hidden email].




Reply | Threaded
Open this post in threaded view
|

Re: query on reliability test

David Marso
Administrator
Rather than repeat what has been stated in this group many many times before I suggest you to search this group and the internet for the following:
"covariance matrix is zero or approximately zero"
"singular matrix"
"cannot be inverted"
"matrix is not of full rank"
---
Briefly:  
a. One or more of your variables are a linear combination of others?
b. You have more cases than variables (after listwise deletion)?
c. One or more variables is a constant?
d. some other data specific weirdness...
---
pooja chawda wrote
Dear Sir,
>


> I am having query on covariance matrix which is showing the warning like,
> "The determinant of the
>
> covariance matrix is zero or approximately zero. Statistics based on its
> inverse matrix cannot be
> computed and they are displayed as system missing values." i m having 36
> variables and when
> cronbach alpha calculated it was, Reliability Statistics
> Cronbach's Alpha (.817) Cronbach's Alpha Based on Standardized Items
> (.806) after deleting 5 items
> which were having very low (one is in minus value) the cronbach alpha
> increases to, (.848) and
> áCronbach's Alpha Based on Standardized Items (.836) now what should i do
> with that warning? i am
>
> looking for factor analysis as i am having altogather 80 variables in the
> different category of HRM
> practices (Independent Variables), Firms Operational Performance, Firm
> performance (financial and
> non financial) as well as business operation modes. How is this warning
> affect the results? Please
> help me.á
>
>
> Pooja Chawda
> Researcher and Asst. Professor,
> Indu Management Institute,
> Baroda, Gujarat, India.
>
>
>
>
>
>
>
>
> On Sun, Jun 17, 2012 at 1:13 PM, L-Soft list server at UGA (1.8d) <
> [hidden email]>
> wrote:
>      Sun, 17 Jun 2012 03:43:21
>
>      Your ásubscription to áthe SPSSX-L álist (SPSSX(r) áDiscussion) has
> ábeen
>      accepted.
>
>      Please save this message for future áreference, especially if this is
> the
>
>      first time you are subscribing to an electronic mailing list. If you
> ever
>      need to leave áthe list, you will find the ánecessary instructions
> below.
>      Perhaps ámore importantly, ásaving a ácopy of áthis message á(and of
> áall
>      future subscription notices áfrom other mailing lists) in áa special
> mail
>
>      folder will give you instant access to the list of mailing lists that
> you
>      are subscribed áto. This may áprove very useful the ánext time you
> ágo on
>      vacation and áneed to leave áthe lists temporarily so áas not to
> áfill up
>      your ámailbox while áyou áare away! áYou should áalso ásave the
> á"welcome
>      messages" from the álist owners that you will áoccasionally receive
> after
>
>      subscribing to a new list.
>
>      To send áa message to áall the people ácurrently subscribed to áthe
> list,
>      just send mail to [hidden email]. áThis is called "sending
> mail
>      to the álist," because áyou send ámail to a ásingle address áand
> LISTSERV
>      makes ácopies áfor all áthe ápeople áwho áhave subscribed. áThis
> áaddress
>      ([hidden email]) is áalso called áthe "list address." áYou
> must
>      never send a command áto that address, as it would ábe distributed to
> all
>      the ápeople áwho have ásubscribed. áAll ácommands ámust ábe sent áto
> áthe
>
>      "LISTSERV *server* address," [hidden email].
>
>      You may leave the list at any time by sending a "SIGNOFF SPSSX-L"
> command
>      to [hidden email].
>
>      For áinformation áabout ásubscription áoptions, send áa á"QUERY
> áSPSSX-L"
>
>      command to [hidden email].
>
>      Contributions sent to this list are automatically archived. You can
> get a
>      list of the available archive files by sending an "INDEX SPSSX-L"
> command
>      to [hidden email]. You can then áorder these files with a
> "GET
>
>      SPSSX-L LOGxxxx" command, or using LISTSERV's database search
> facilities.
>      Send an "INFO DATABASE" command for more information on the latter.
>
>      This álist is áavailable áin digest áform. áIf you áwish áto receive
> áthe
>      digested áversion áof the ápostings, ájust áissue áa SET áSPSSX-L
> áDIGEST
>      command.
>
>      More áinformation on áLISTSERV ácommands ácan be áfound áin the
> áLISTSERV
>      reference ácard, which áyou can áretrieve áby sending áan "INFO
> áREFCARD"
>      command to [hidden email].
>
>
>
>
Please reply to the list and not to my personal email.
Those desiring my consulting or training services please feel free to email me.
---
"Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis."
Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?"