Coefficient Alpha via RELIABILITY using COV & CORR MATRIX
Locked
 
|
All,
What's the code to calculate coefficient alpha via the RELIABILITY procedure for the following two sets of syntax?: MATRIX DATA VARIABLES = ROWTYPE_ x1 x2 x3 x4.
BEGIN DATA N 500 500 500 500 COV 1.022 COV .717 1.008 COV .713 .780 1.098 COV .712 .624 .747 .995 END DATA. and MATRIX DATA VARIABLES = ROWTYPE_ x1 x2 x3 x4. BEGIN DATA N 1000 1000 1000 1000 MEAN 11.0 12.0 10.0 10.1 STDEV 3.2 2.8 6.0 2.1 CORR 1.00 CORR .50 1.00 CORR .47 .38 1.00 CORR .44 .51 .77 1.00 END DATA. Or must I calculate coefficient alpha outside of RELIABILITY? TIA, Ryan |
Re: Coefficient Alpha via RELIABILITY using COV & CORR MATRIX
Administrator
|
Hi Ryan. I don't have SPSS on this machine, so cannot test. But, RELIABILITY has a /MATRIX subcommand, and can take a correlation matrix dataset as input. It cannot take a covariance matrix dataset as input, but you can use MCONVERT to convert your covariance matrix dataset to a correlation matrix dataset. Something like:
MATRIX DATA VARIABLES = ROWTYPE_ x1 x2 x3 x4. BEGIN DATA N 500 500 500 500 COV 1.022 COV .717 1.008 COV .713 .780 1.098 COV .712 .624 .747 .995 END DATA. MCONVERT matrix=out(*). Then add RELIABILITY with /MATRIX = in(*) as a sub-command. Something like: RELIABILITY /VARIABLES= x1 to x4 /MODEL=ALPHA /STATISTICS=DESCRIPTIVE SCALE CORRELATIONS /SUMMARY=TOTAL MEANS VARIANCE /MATRIX=IN(*). HTH.
--
Bruce Weaver bweaver@lakeheadu.ca http://sites.google.com/a/lakeheadu.ca/bweaver/ "When all else fails, RTFM." PLEASE NOTE THE FOLLOWING: 1. My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above. 2. The SPSSX Discussion forum on Nabble is no longer linked to the SPSSX-L listserv administered by UGA (https://listserv.uga.edu/). |
|
|
Bruce is mostly correct below. The reliability procedure expects
N, MEAN, STDDEV to be present if you are using MATRIX=IN(*). So, for your correlation matrix, the following works: RELIABILITY VARIABLES=ALL /MATRIX=IN(*). However, the Mconvert procedure will produce a correlation matrix that will not have a row of means which will cause reliability to fail. If you don't have the raw data, you'll have to calculate alpha outside of reliability. -Mike Palij New York University [hidden email] ----- Original Message ----- From: "Bruce Weaver" <[hidden email]> To: <[hidden email]> Sent: Sunday, January 26, 2014 8:43 PM Subject: Re: Coefficient Alpha via RELIABILITY using COV & CORR MATRIX > Hi Ryan. I don't have SPSS on this machine, so cannot test. But, > RELIABILITY has a /MATRIX subcommand, and can take a correlation > matrix > dataset as input. It cannot take a covariance matrix dataset as > input, but > you can use MCONVERT to convert your covariance matrix dataset to a > correlation matrix dataset. Something like: > > MATRIX DATA VARIABLES = ROWTYPE_ x1 x2 x3 x4. > BEGIN DATA > N 500 500 500 500 > COV 1.022 > COV .717 1.008 > COV .713 .780 1.098 > COV .712 .624 .747 .995 > END DATA. > > MCONVERT matrix=out(*). > > Then add RELIABILITY with /MATRIX = in(*) as a sub-command. Something > like: > > RELIABILITY > /VARIABLES= x1 to x4 > /MODEL=ALPHA > /STATISTICS=DESCRIPTIVE SCALE CORRELATIONS > /SUMMARY=TOTAL MEANS VARIANCE > /MATRIX=IN(*). > > HTH. > > > > > Ryan Black wrote >> All, >> >> What's the code to calculate coefficient alpha via the RELIABILITY >> procedure for the following two sets of syntax?: >> >> MATRIX DATA VARIABLES = ROWTYPE_ x1 x2 x3 x4. >> BEGIN DATA >> N 500 500 500 500 >> COV 1.022 >> COV .717 1.008 >> COV .713 .780 1.098 >> COV .712 .624 .747 .995 >> END DATA. >> >> and >> >> MATRIX DATA VARIABLES = ROWTYPE_ x1 x2 x3 x4. >> >> BEGIN DATA >> >> N 1000 1000 1000 1000 >> >> MEAN 11.0 12.0 10.0 10.1 >> >> STDEV 3.2 2.8 6.0 2.1 >> >> CORR 1.00 >> >> CORR .50 1.00 >> >> CORR .47 .38 1.00 >> >> CORR .44 .51 .77 1.00 >> >> END DATA. >> >> >> Or must I calculate coefficient alpha outside of RELIABILITY? >> >> >> TIA, >> >> >> Ryan > > > > > > ----- > -- > Bruce Weaver > [hidden email] > http://sites.google.com/a/lakeheadu.ca/bweaver/ > > "When all else fails, RTFM." > > NOTE: My Hotmail account is not monitored regularly. > To send me an e-mail, please use the address shown above. > > -- > View this message in context: > http://spssx-discussion.1045642.n5.nabble.com/Coefficient-Alpha-via-RELIABILITY-using-COV-CORR-MATRIX-tp5724158p5724159.html > Sent from the SPSSX Discussion mailing list archive at Nabble.com. > > ===================== > To manage your subscription to SPSSX-L, send a message to > [hidden email] (not to SPSSX-L), with no body text except > the > command. To leave the list, send the command > SIGNOFF SPSSX-L > For a list of commands to manage subscriptions, send the command > INFO REFCARD ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
|
|
Thanks, Mike and Bruce. I added the row of means. Best, Ryan On Sun, Jan 26, 2014 at 9:23 PM, Mike Palij <[hidden email]> wrote: Bruce is mostly correct below. The reliability procedure expects |
Re: Coefficient Alpha via RELIABILITY using COV & CORR MATRIX
|
Administrator
|
In reply to this post by Mike
Right! Good catch, Mike.
--
Bruce Weaver bweaver@lakeheadu.ca http://sites.google.com/a/lakeheadu.ca/bweaver/ "When all else fails, RTFM." PLEASE NOTE THE FOLLOWING: 1. My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above. 2. The SPSSX Discussion forum on Nabble is no longer linked to the SPSSX-L listserv administered by UGA (https://listserv.uga.edu/). |
Re: Coefficient Alpha via RELIABILITY using COV & CORR MATRIX
Administrator
|
In reply to this post by Mike
This strikes me as silly/stupid on the part of SPSS/IBM as
One does NOT require means to calculate alpha! WHY should MCONVERT take such liberties and clobber the mean vector? Furthermore, why doesn't RELIABILITY just simply read in the COVARIANCE matrix and work with it as is?
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?" |
Re: Coefficient Alpha via RELIABILITY using COV & CORR MATRIX
|
Administrator
|
MCONVERT does not "clobber the mean vector"--sorry if I suggested it does. In Ryan's example, there was no mean vector to start with. If the means are there to start with, they are retained in the MCONVERTed matrix. I don't know what Ryan's actual means are, but plugging in some values to illustrate:
MATRIX DATA VARIABLES = ROWTYPE_ x1 x2 x3 x4. BEGIN DATA N 500 500 500 500 MEAN 40 50 60 70 COV 1.022 COV .717 1.008 COV .713 .780 1.098 COV .712 .624 .747 .995 END DATA. MCONVERT matrix=out(*). FORMATS x1 to x4(f8.3). LIST. Output: ROWTYPE_ VARNAME_ x1 x2 x3 x4 N 500.000 500.000 500.000 500.000 MEAN 40.000 50.000 60.000 70.000 STDDEV 1.011 1.004 1.048 .997 CORR x1 1.000 .706 .673 .706 CORR x2 .706 1.000 .741 .623 CORR x3 .673 .741 1.000 .715 CORR x4 .706 .623 .715 1.000 Number of cases read: 7 Number of cases listed: 7 ;-)
--
Bruce Weaver bweaver@lakeheadu.ca http://sites.google.com/a/lakeheadu.ca/bweaver/ "When all else fails, RTFM." PLEASE NOTE THE FOLLOWING: 1. My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above. 2. The SPSSX Discussion forum on Nabble is no longer linked to the SPSSX-L listserv administered by UGA (https://listserv.uga.edu/). |
«
Return to SPSSX Discussion
|
1 view|%1 views
| Free forum by Nabble | Edit this page |