SOC
206
Winter 15
Second
Assignment
Part One:
In Part One you will
estimate a path model of your choice without latent variables.
Answer the following questions:
- What did you learn from just looking at the
direct effects?
- Is there a variable where the indirect effect is
considerable?
- How much of the variation in your dependent
variables are you explaining? (Pay special attention to the final
dependent variable.)
- What happened to your model after you set some of
the paths to zero? Did the other paths change as well? How good is the
fit?
- Look at the two variables at the two ends of the deleted path. Go back to the first model with the path in. Write out all the different ways those two variables are related using the path coefficients from the first analysis. Then redo the same thing using the path coefficients from the second analysis, where the direct effect is set to 0. What is the total association in the first and in the second case?
HAND IN YOUR OUTPUT with your answers..
Part Two
In Part Two you do a latent variable analysis.
If you wish you may use your own data set and your own latent variable(s) for this part of the assignment. Hand in your output.
Evaluate the overall fit of your
model. Evaluate your measurement model. Explain your substantive model. What do
we learn?
If you don't have data of your own that you can use (e.g., because your
variables are categorical or you don't have multiple indicators to make latent
variables) you can download a data set from the
The question we want to answer is how
happiness is influenced by individualist ideology. Are people who subscribe to
a more individualistic ideology happier? Both happiness and individualistic
ideology are hard to capture and will be modeled as a latent variable. There is
a third, control variable, gender.
happiness, will have three indicators. Each scored on a scale
from 1-10 where 1= completely dissatisfied and 10=perfectly satisfied.
:
HAPHOME (V180) Overall, how satisfied are you
with your home life?
HAPFIN (V132) How satisfied are you with the
financial situation of your household?
HAPLIFE (V96) All
things considered, how satisfied are you with your life as a whole?
individualism, will have four indicators. Each is scored on a ten
point scale of agreement.
1
2
3
4
5
6
7
8
9
10
PRIVOWN (V251)
Government ownership
of
Private ownership of
business and industry should be
increased
business and industry should be increased
INDRESP (V252)
The state should take
more
responsibility
Individuals should take more responsibility
to ensure that everyone is provided
for
to ensure that everyone is provided for
COMPETE (V254)
Competition is harmful. It brings out
the
Competition is good it stimulates people to
worst in
people
work hard and develop new ideas
HARDWRK (V255)
Hard work doesn't generally bring
success
In the long run, hard work usually brings a
-- it is more a
matter of luck and
connections
better life
GENDER
1=Male and 2=Female
Create the proper model.
What you need to hand in is the print out of
the diagram with the standardized coefficients. Using the usual asterisk
notation, indicate which paths are significant at the .05, .01 or .001 level.
(You may use the infinity row of the t table in your book and treat C.R. as
t-values.) Also hand in the overall fit of the model (Chi-square and its degree
of freedom).
Answer the following questions:
- What is the effect of gender on happiness
controlling for individualism?
- What is the effect of individualism on happiness
controlling for gender?
- How well are we explaining why people are happy?
- Comment on why you think we have found these
results?
- What are the strongest indicators of happiness?
What are the strongest indicators of individualism? Does that make sense?
- How good is the overall fit?
HELP WITH STATA SEM:
HOW TO RUN STRUCTURAL EQUATION MODELS (SEM) IN STATA
Get into Stata.
Click on Statistics on the top bar.
Click on SEM(structural equation modeling) and choose Model building and estimation.
You can now start drawing your path model
1. From the side you can pick the variables (squares). Click on the square, move the cursor over the canvas (drawing area) and click to drop it.
2. Move to the bar above the canvas and you will see Variable written in the left corner. Next to it there is a drop-down menu. Pick the variable name and you will see it appear in the square.
3. Move the cursor again and click to drop the next square. Repeat until you have all the squares you need.
Suppose you forgot to put in the variable name or you want to make a change.
Move the cursor over the square and click. The square will be selected. (On the
left bar the fat arrow tip pointing up-left should be on. This is the Select
icon.) Now you can make the change on the bar above the canvas.
4. To draw the arrows, pick the horizontal skinny arrow on the left, click on it, take the cursor to the variable of origin. Click but hold down the clicker, extend the arrow to the variable of destination and release. (You will see that the error variable will be immediately added.) Draw in all single-headed arrows.
5. To draw double-headed arrow, find it on the sidebar under the single-headed arrow. (Hint: if you want it to curve out, on the left, start the arrow at the variable below and connect it up.)
If your model involves latent variables, you choose the ovals on the left. The only difference is that in the Variable window you will now have to type in a name of your choosing. (Latent variables are not in your dataset, they are created by the model.)
6. To estimate the model, click on Estimation, choose Estimate. You will get a panel with eight tabs. Make sure that under the Reporting tab you checked Display standardized coefficients and values.
7. Click on OK. You will see the path coefficients on your diagram. If you minimize the canvas you will see your output file.
8. To get additional output. Click on Estimation. Choose Testing and CIs (CI= Confidence Interval). Then choose Direct and indirect effects. From the panel that pops up choose Decomposition of effects into total, direct, and indirect (teffects). Then check Report standardized effects. Submit.
. estat teffects, standardized
9. To see how you reproduced the correlation go to Other after clicking on SEM. Choose Report model framework. Choose Display estimation results in modeling framework (framework) and click on all three boxes and Submit. What you need is the matrix that says ‘Fitted covariances of observed and latent variables (standardized)’. (Remember, the correlation is a standardized covariance.) Ignore the correlations that belong to the latent variable(s). This is the correlation calculated from the paths and correlations STATA estimated.
. estat framework, standardized compact fitted
10. To get the observed (as opposed to the reproduced or fitted) correlations, just use the correlation function
. correlate
11. To get the R-squared for each endogenous variable, go to Goodness of fit then Equation-level goodness of fit. Submit.
. estat eqgof
12. To get the overall goodness of fit measures choose Overall goodness of fit and in the Statistics to be displayed window select ‘all’
. estat gof, stats(all)