Verfasste Forenbeiträge
-
Beiträge
-
Onyx currently has no means to compute indirect or total effects (i.e., multiplying and summing loadings). We will consider your request for future releases.
thanks for using Onyx,
AndreasHello Andreas,
thank you for the prompt reply and for fixing the problem so quickly!
Kind regards,
MaikeThe problem is fixed now. Please download the latest Onyx version (at least V1.0-1010). Thanks again for reporting this problem!
Maike,
thanks for letting us know about this.
Likely, you have encountered a bug in Onyx. From the values you report, I guess that you 10 observed variables? In this case, Onyx gets confused about the presence/absence of the mean structure or, I believe, does counting the dfs in a faulty way. We will check asap.
Thanks,
AndreasHi Sebwin,
these are the „restricted degrees of freedom“; they are sadly sometimes, not fully precisely, also just called „degrees of freedom“. The k(k-1)/2 are the number of entries in the covariance matrix, so these would be the degrees of freedom of the „saturated model“, i.e., the most complex possible model which estimates every entry of the covariance matrix separately. The difference of k(k-1)/2 to q is the number of degrees of freedom that your model has less than the saturated model. Note that it doesn’t make any difference if you compare two models, because the „k(k-1)/2“ part appears on both sides of the minus sign and goes away anyway.
Cheers,
Timo
Hi Timo
Thanks, once more, for answering.
I’m not sure, though. According to http://davidakenny.net/cm/basics.htm#Degrees, df = k(k – 1)/2 – q where k is the number of variables and q is the number of free parameters.
What do you think of this?Hi Sebwin,
What you are looking for are the „estimated parameter“, the second rof from top.
To add yet another A‘ (sorry 😉 : You can let Onyx do the work of deterimining the difference of degrees of freedoms between two models, just drag a line from one model to the other, and Onyx will give you the difference in chi^2, the difference in degrees of freedom, and the p-value determined from these two values.
Thank you, Timo, for your always elaborate answers.
It always takes me a while to process when you suggest an alternative approach to what I am trying to accomplish because I am not really „literate“ in these matters and, thus, lack the necessary flexibility to do (and understand) A‘ when I was set on doing A.
Your advice is greatly appreciated all the same — just with a lag of several days.
My thanks also to Robin, who suggested the same earlier.Hi sebwin,
the path coefficients per se cannot be significant, but I guess you mean significantly different from zero, which makes a lot of sense of course. Robin’s way is certainly quick. You just check the z-value whether it is above 1.96 or below -1.96, respectively, and if it is, the path coefficient is significantly different from zero. Advantage of this method of displaying the result (contrary, say, to giving the p-value for this test, which some other programs do) is that you can also check if the z-value is, say, above 2.96; if it is, the path coefficient is significantly different from 1 instead of 0 (which sometimes can be more interesting than the test against zero, although 0 is of course more frequently useful).
If you want the „best possible p-value“, which in this case would mean one that also includes the cross-information from other parameters, I suggest to set up a likelihood ratio test; clone your model, and in the clone, fix the path you are interested in to zero (or one or any other value). You can then connect the two models (by dragging a path from one to the other) and check in the little ball on the edge between the models (by hovering over it with the mouse) what the p-value for this comparison is. This is a Likelihood Ratio test, which is provably the best test asymptotically for normal distributed data.
BTW (I see we cross-posted, and you asked about modification indices), you can also use the LR between these two models as a modification index if the path you restricted is a factor loading.
Cheers,
Timo
Really, I need to find out which coefficients are significant.
I would suppose that mostly everyone is interested in knowing which ones of the path values in a model’s estimate are significant, but I have yet to find this information displayed in onyx.Hi Chang,
Thanks for using Onyx! You can turn off the „=0.0“ by choosing „Customize Model -> Change Path Style“ and select „label only“. I use the same workaround that you suggest here by typing some numbers in the label if I want to show students / readers something specific. You can, however, also turn off standardized estimates to show automatically, by selecting „Show Standardized Estimates“ on a path.
Cheers,
Timo
Standardizes estimates are displayed as NaN if one of the involved variance estimates are negative (Heywood case). This likely occurs when your model is misspecified.
