| Message/Author |
|
|
| Ismail S posted on Wednesday, March 02, 2016 - 9:03 am
|
|
|
Hi all,
I'm not sure if my data can be analyzed using a latent growth curve model as discussed by Barbara Byrne
I have yearly economic data on 34 countries for the 2006-2014 period. Most items are on a 1-7 Likert scale and I have the average score on each item from the surveys of firms conducted in each country every year. The yearly samples do not necessarily include the same firms from previous years. However, approximately a third of the 2014 Survey respondents are “repeat” respondents—that is to say, they are executives who have previously taken part in the Survey. This improves the comparability of data across years.
So, can I treat each country-aggregate as a single observation and use a latent growth curve model? Thanks for your help. |
|
|
|
|
That would be very approximate.
I would treat this as 2-level growth in Mplus (3-level analysis) with firms within countries. Keep track of which firms are present in different years using a multiple-cohort approach to growth (see UG ex 6.18) |
|
|
| Ismail S posted on Friday, March 04, 2016 - 4:43 am
|
|
|
I am going to use the Global Competitiveness Report data (http://reports.weforum.org/global-competitiveness-report-2015-2016/competitiveness-rankings/) and they only provide the overall average on each item for each country. So I don't have any firm level data and can't keep track of which firms are present in different years. So the set-up would be like this (with more variables than provided below):
Country Year Property rights Patents
USA 2006 5.1 4.8
... ....
USA 2014 4.9 3.5
Canada 2006 5.3 4.5
....
Canada 2014 4.6 5.6
Austria 2006.....
.....
Does this look right?
Thanks. |
|
|
|
|
| You may want to discuss this on SEMNET. |
|
|
| Ismail S posted on Saturday, March 05, 2016 - 5:21 am
|
|
|
| I apologize if my question was not appropriate. |
|
|
| Back to top |
