Grad Pack missing values module compatibility with AMOS
Locked
Grad Pack missing values module compatibility with AMOS
 
|
HI, I am currently using the Grad Pack version of spss
v.17. It does not have the missing values module, and as a student in
Australia I cannot purchase this add on module. I am currently wanting to run an AMOS SEM but I have missing
data – so ideally need to impute data using Bayesian imputation.
Can anyone tell me: ·
if I run the imputation module through the
missing values module (externally – using another computer with the full version
of spss 17) to create an imputed file, can I then use this imputed file in my
grad pack version of SPSS v17 with AMOS (without the missing values
module)? Before someone asks why I don’t just use the ‘other’
computer for my analysis – I am an external student (2 1/2 hrs from
campus – which has the full version) and need to be able to work on the
SEM models long term from home (months) with the Grad Pack version only. Any advice, or others experience would be appreciated. Regards, Paola “Ours has become a time-poor society, fatigued by
non-physical demands and trying to compartmentalize daily living tasks.
It is small wonder that physical activity is discarded in this
environment” p126 (Steinbeck, 2001) P Please consider the
environment before printing this email. |
Re: Grad Pack missing values module compatibility with AMOS
|
|
Hi Paola,
I've never used the Missing Values Analysis pack with Amos, only with SPSS. Once you use MVA, you retain the original data set and get x number of new sets (depending on how many imputations you want). You can use that data with any version of SPSS. However, SPSS with MVA will give you analysis for each imputation and pooled tests. If you run SPSS without MVA, you would need to filter out the original data and then your tests would all look like you have a lot more data than you do as it would just combine all the imputed data sets. I do not know if it is the same case with Amos or not.
Hope that helps,
Jill
On Tue, May 19, 2009 at 4:34 AM, Paola Chivers <[hidden email]> wrote:
|
Classification: LPA and cluster analysis
|
|
|
Re: Classification: LPA and cluster analysis
|
|
With hierarchical cluster analyhsis
(CLUSTER command in SPSS) there is no single solution: the procedure starts
with N clusters of 1 member each, and finishes with one cluster including all N
cases as members. Therefore it is not surprising that you found hierarchical
cluster to come up with a “one cluster solution”: that is just the
last step in the procedure, not “the solution”. What you have to do
next is examine the various “solutions”, with 1 to N clusters
including all the intermediate results (the penultimate one was a solution with
two clusters) to see whether any of them is of your liking. Remember, in all
this, that clustering is not a parametric but a heuristic procedure. There is
no “correct” solution. You can check, externally, which clustering
solution is better for your particular purposes. For instance, if you are
interested in some particular criterion, and we seek forming clustering that
are maximally homogeneous internally, and maximally distinct between them, in
some other variable, you can use one-way ANOVA with different clustering
solutions to see which is best for that purpose. Likewise, if you want to have
a moderate number of clusters, from 2 to six say, you can restrict yourself to
those “solutions” and try to choose the one you judge the best. As each procedure uses a different
algorithm to include or exclude cases in/from clusters, it is not surprising
either that solutions are not necessarily coincident case by case. Even within
the same procedure, say Hierarchical or quick cluster, using different criteria
may end up with different clustering decisions for specific cases. Such is the
nature of clustering. Hector From: SPSSX(r)
Discussion
|
Re: Classification: LPA and cluster analysis
|
In reply to this post by Dale Glaser
All forms of clustering are exploratory. In my experience, different
proximity measures and different agglomeration methods yield different
results. This has been true whether clustering types of classrooms,
poverty of county pops, congressional districts, schizophrenics, music
preferences, etc. That is why since the early 70's I have used multiple
clustering methods. I treat cases that are clustered together by
several methods as "cores". That often leaves some cases unclustered.
I then iteratively use the classification phase of DISCRIMINANT to find
cases that are far from the center of the assigned cluster or assigned
to a cluster other than the expected cluster. Those cases are then
considered unclassified going into the next round of DFA.
I. Some hierarchical techniques show a measure of "error" which sometimes shows a jump when very disparate cluster are joined. In the past those jumps and interpretability were the criteria for the ballpark number of clusters. These days the AIC and or BIC from TWOSTEP can also be used to ballpark the number of cores to retain. WRT the single link procedure, it will show in your instance all clusterings from 111 down to one. Looking at the plots often gives a guess about the number to retain. IIRC about 20 or so years ago someone at a Classification Society meeting pointed out that Latent Profile Analysis was the same as some other clustering approach but I do no recall which. If not in details, it still shares the goal of other non-hierarchical approaches of finding a single nominal level variable that describes a set of profiles were cases have similar values within a cluster and dissimilar values between the centroids of the clusters. Methods differ on how much they emphasize shape, elevation, or scatter of profiles. They also differ on the degree to which they tend to produce stringy or compact clusters. The Classification Society is for people interested in such issues. http://thames.cs.rhul.ac.uk/~fionn/classification-society/ to ask from more info you might want to post to this list. http://lists.sunysb.edu/index.cgi?A0=CLASS-L Art Kendall Social Research Consultants Dale Glaser wrote:
Art Kendall
Social Research Consultants |
| Free forum by Nabble | Edit this page |