I am wondering if anyone can point me to a resource that explains (in detail) the different ways that data is treated and how you decide which postcommand to use when you run the following commands after going through all other steps of EFA (factortest, factor, etc. etc.) to get your final items that load on the factors.
alpha (varlist), item
alpha (varlist), item asis
alpha (varlist), std
I had been using alpha (varlist), item asis because I had taken steps to reverse code my items before analysis and thus wanted the variable to be taken as it was coded but for one of my scales I get very odd results when I run alpha with item asis. I get values that are much more expected for both alpha (varlist), item and alpha (varlist), item std.
I have included below the results I get when I run with each option and am looking to understand the significant variance to better understand how to treat my data.
alpha a1 a2 a3 a4 a5 a6 a8 a9 a10 a11 a12 a13 a15, item asis
Test scale = mean(unstandardized items)
average
item-test item-rest interitem
Item | Obs Sign correlation correlation covariance alpha
-------------+-----------------------------------------------------------------
a1 | 482 + 0.5449 0.3210 -.0034536 .
a2 | 483 + 0.6176 0.4162 -.0067417 .
a3 | 483 + 0.5873 0.3808 -.0053486 .
a4 | 483 + -0.2135 -0.4384 .0323812 0.3330
a5 | 483 + 0.0125 -0.2482 .0221584 0.2528
a6 | 483 + 0.1395 -0.1234 .0156097 0.1906
a8 | 483 + 0.1806 -0.0838 .0137681 0.1717
a9 | 483 + 0.6287 0.4284 -.0073476 .
a10 | 483 + 0.3115 0.0463 .0077008 0.1035
a11 | 483 + 0.6548 0.4596 -.0086483 .
a12 | 483 + 0.6528 0.4629 -.0084055 .
a13 | 483 + -0.4087 -0.5991 .0433993 0.4061
a15 | 482 + 0.1731 -0.0769 .0132409 0.1649
-------------+-----------------------------------------------------------------
Test scale | .0083316 0.1188
-------------------------------------------------------------------------------
alpha a1 a2 a3 a4 a5 a6 a8 a9 a10 a11 a12 a13 a15, item std
Test scale = mean(standardized items)
average
item-test item-rest interitem
Item | Obs Sign correlation correlation correlation alpha
-------------+-----------------------------------------------------------------
a1 | 482 + 0.6057 0.5186 0.3424 0.8620
a2 | 483 + 0.6559 0.5766 0.3363 0.8587
a3 | 483 + 0.7203 0.6525 0.3283 0.8543
a4 | 483 - 0.6771 0.6014 0.3336 0.8573
a5 | 483 - 0.5657 0.4729 0.3474 0.8646
a6 | 483 - 0.5269 0.4291 0.3522 0.8671
a8 | 483 + 0.4589 0.3535 0.3605 0.8712
a9 | 483 + 0.7204 0.6526 0.3283 0.8543
a10 | 483 + 0.5958 0.5072 0.3436 0.8627
a11 | 483 + 0.6680 0.5908 0.3348 0.8579
a12 | 483 + 0.7204 0.6526 0.3283 0.8543
a13 | 483 - 0.7274 0.6608 0.3275 0.8539
a15 | 482 - 0.4845 0.3819 0.3573 0.8697
-------------+-----------------------------------------------------------------
Test scale | 0.3400 0.8701
-------------------------------------------------------------------------------
. alpha a1 a2 a3 a4 a5 a6 a8 a9 a10 a11 a12 a13 a15, item
Test scale = mean(unstandardized items)
average
item-test item-rest interitem
Item | Obs Sign correlation correlation covariance alpha
-------------+-----------------------------------------------------------------
a1 | 482 + 0.6051 0.5182 .2781408 0.8621
a2 | 483 + 0.6551 0.5771 .2737861 0.8588
a3 | 483 + 0.7181 0.6523 .2680839 0.8546
a4 | 483 - 0.6766 0.6001 .2705528 0.8573
a5 | 483 - 0.5680 0.4730 .2809156 0.8648
a6 | 483 - 0.5268 0.4289 .2857697 0.8672
a8 | 483 + 0.4594 0.3535 .2923981 0.8713
a9 | 483 + 0.7200 0.6527 .266834 0.8544
a10 | 483 + 0.5973 0.5069 .2779143 0.8627
a11 | 483 + 0.6688 0.5912 .2713449 0.8579
a12 | 483 + 0.7199 0.6534 .2673543 0.8545
a13 | 483 - 0.7300 0.6607 .2633163 0.8537
a15 | 482 - 0.4811 0.3833 .2915311 0.8692
-------------+-----------------------------------------------------------------
Test scale | .275996 0.8701
-------------------------------------------------------------------------------
Thanks for any guidance you can provide.
alpha (varlist), item
alpha (varlist), item asis
alpha (varlist), std
I had been using alpha (varlist), item asis because I had taken steps to reverse code my items before analysis and thus wanted the variable to be taken as it was coded but for one of my scales I get very odd results when I run alpha with item asis. I get values that are much more expected for both alpha (varlist), item and alpha (varlist), item std.
I have included below the results I get when I run with each option and am looking to understand the significant variance to better understand how to treat my data.
alpha a1 a2 a3 a4 a5 a6 a8 a9 a10 a11 a12 a13 a15, item asis
Test scale = mean(unstandardized items)
average
item-test item-rest interitem
Item | Obs Sign correlation correlation covariance alpha
-------------+-----------------------------------------------------------------
a1 | 482 + 0.5449 0.3210 -.0034536 .
a2 | 483 + 0.6176 0.4162 -.0067417 .
a3 | 483 + 0.5873 0.3808 -.0053486 .
a4 | 483 + -0.2135 -0.4384 .0323812 0.3330
a5 | 483 + 0.0125 -0.2482 .0221584 0.2528
a6 | 483 + 0.1395 -0.1234 .0156097 0.1906
a8 | 483 + 0.1806 -0.0838 .0137681 0.1717
a9 | 483 + 0.6287 0.4284 -.0073476 .
a10 | 483 + 0.3115 0.0463 .0077008 0.1035
a11 | 483 + 0.6548 0.4596 -.0086483 .
a12 | 483 + 0.6528 0.4629 -.0084055 .
a13 | 483 + -0.4087 -0.5991 .0433993 0.4061
a15 | 482 + 0.1731 -0.0769 .0132409 0.1649
-------------+-----------------------------------------------------------------
Test scale | .0083316 0.1188
-------------------------------------------------------------------------------
alpha a1 a2 a3 a4 a5 a6 a8 a9 a10 a11 a12 a13 a15, item std
Test scale = mean(standardized items)
average
item-test item-rest interitem
Item | Obs Sign correlation correlation correlation alpha
-------------+-----------------------------------------------------------------
a1 | 482 + 0.6057 0.5186 0.3424 0.8620
a2 | 483 + 0.6559 0.5766 0.3363 0.8587
a3 | 483 + 0.7203 0.6525 0.3283 0.8543
a4 | 483 - 0.6771 0.6014 0.3336 0.8573
a5 | 483 - 0.5657 0.4729 0.3474 0.8646
a6 | 483 - 0.5269 0.4291 0.3522 0.8671
a8 | 483 + 0.4589 0.3535 0.3605 0.8712
a9 | 483 + 0.7204 0.6526 0.3283 0.8543
a10 | 483 + 0.5958 0.5072 0.3436 0.8627
a11 | 483 + 0.6680 0.5908 0.3348 0.8579
a12 | 483 + 0.7204 0.6526 0.3283 0.8543
a13 | 483 - 0.7274 0.6608 0.3275 0.8539
a15 | 482 - 0.4845 0.3819 0.3573 0.8697
-------------+-----------------------------------------------------------------
Test scale | 0.3400 0.8701
-------------------------------------------------------------------------------
. alpha a1 a2 a3 a4 a5 a6 a8 a9 a10 a11 a12 a13 a15, item
Test scale = mean(unstandardized items)
average
item-test item-rest interitem
Item | Obs Sign correlation correlation covariance alpha
-------------+-----------------------------------------------------------------
a1 | 482 + 0.6051 0.5182 .2781408 0.8621
a2 | 483 + 0.6551 0.5771 .2737861 0.8588
a3 | 483 + 0.7181 0.6523 .2680839 0.8546
a4 | 483 - 0.6766 0.6001 .2705528 0.8573
a5 | 483 - 0.5680 0.4730 .2809156 0.8648
a6 | 483 - 0.5268 0.4289 .2857697 0.8672
a8 | 483 + 0.4594 0.3535 .2923981 0.8713
a9 | 483 + 0.7200 0.6527 .266834 0.8544
a10 | 483 + 0.5973 0.5069 .2779143 0.8627
a11 | 483 + 0.6688 0.5912 .2713449 0.8579
a12 | 483 + 0.7199 0.6534 .2673543 0.8545
a13 | 483 - 0.7300 0.6607 .2633163 0.8537
a15 | 482 - 0.4811 0.3833 .2915311 0.8692
-------------+-----------------------------------------------------------------
Test scale | .275996 0.8701
-------------------------------------------------------------------------------
Thanks for any guidance you can provide.
