Re: st: ordered logistic integration problems

["JVerkuilen (Gmail)" <jvverkuilen@gmail.com>]

On Thu, Mar 21, 2013 at 2:14 PM, Bontempo, Daniel E <deb193@ku.edu> wrote:
> Jay, I agree a lot is not being captured.
>
> As Nick Cox wrote:
> " There also seems nothing unusual in the idea that different proportions arise from different combinations. 1/1 of cars in our household have four seats and 3/3 cars in my friend's household. The same fraction, different situations, some information loss on data reduction."
>
> ... what is being lost is the "situation", specifically how many past tense verbs the kid attempted. I would love to try jointly predicting #attempted and fraction correct.

So both the number attempted and the number correct are random
variables? Not simple and I'm not surprised that the variable is
behaving oddly as it's forcing together multiple types of behavior.


>
> Also, unlike the original research that forced a list of 10 verbs, this analysis of spontaneous speech, is also not capturing the "difficulty" of the verbs the kid attempts and/or gets wrong.
>
> I would love to have the dis-aggregated data for each attempt, coded 0/1 for correct. At the attempt level, they could code easy/difficult. At the person level, the could code #attempted and fraction of difficult attempts.
>
> I think they will only be able to look at the predictions across group and occasion in these models and qualitatively judge whether the trajectories look different form the prior work when things were experimentally controlled. This qualitative view may induce new hypotheses about the kids development or self-knowledge.
>

Definitely. This would be a challenging model, that's for sure. What
department are you in at KU?

Jay

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/