Modification index in EFA PreviousNext
Mplus Discussion > Exploratory Factor Analysis >
Message/Author
 Jenny L.  posted on Sunday, June 02, 2013 - 10:41 am
Dear Drs. Muthen,

I was running an EFA with 5 items and only a one-factor solution was produced. Model fitness was mediocre so further modification was needed.

The modification index showed that one item had high MIs and EPCs with several other items. Does it mean that removing this item would lead to noticeable changes in other items and thus it should be the one considered for removal?
 Linda K. Muthen posted on Monday, June 03, 2013 - 10:57 am
You should get both a one- and two-factor solution with five factor-indicators. I would look at the two-factor solution. A residual covariance between two items suggests a possible need for another factor.
 Dodam Park posted on Wednesday, April 12, 2017 - 11:03 pm
Dear Drs. Muthen,
I performed EFA like below.
I added one modification index to improve the model fit. But it hasn't changed the fit at all, so I got just the same result without the modification index. Could you tell me what kind of a problem happened?


TITLE: EFA model of microaggression
DATA: FILE IS microEFA.dat;
FORMAT IS 18f8.2;

VARIABLE: NAMES ARE micro3 micro4 micro9 micro10
micro11 micro13 micro14 micro16 micro18
micro19 micro22 micro23 micro24 micro25
micro26 micro27 micro30 micro31;

ANALYSIS: TYPE = EFA 1 2; ROTATION = GEOMIN;

MODEL: micro16 with micro26;

PLOT: TYPE = Plot2;

OUTPUT: MODINDICES(0);
 Bengt O. Muthen posted on Thursday, April 13, 2017 - 4:04 pm
Please send your output and license number to Support.
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: