Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
st: Re: Midas univariable regression
From
NELE BRUSSELAERS <[email protected]>
To
[email protected]
Subject
st: Re: Midas univariable regression
Date
Tue, 7 Feb 2012 17:22:25 +0100
Hi,
I'm using the user-written midas command (Stata v11.2 - Mac), but I do
have a question considering the interpretation of the univariable
regression model...
Why is the p-value of the pooled sensitivity of var1 and var2 so high?
The 95% confidence intervals are clearly not overlapping, so I would
expect that this is 'strong' evidence that both sensitivities are
different....
I hope someone can help me out! Thanks!
Nele
. midas tp fp fn tn, reg(var1 var2 var3)
Estimating Covariate Effect Of: var1
Estimating Covariate Effect Of: var2
Estimating Covariate Effect Of: var3
Sensitivity and Specificity
| Parameter category nstudies Sensitivity p1 Specificity
p2 |
--------------------------------------------------------------------------------|
| var1 Yes 22 0.79 [0.74 - 0.85] 0.24 0.96 [0.94 - 0.98] 0.00 |
| No 25 0.65 [0.57 - 0.72] . 0.90 [0.86 - 0.95] . |
| var2 Yes 34 0.78 [0.73 - 0.83] 0.54 0.96 [0.94 - 0.98] 0.03 |
| No 15 0.61 [0.52 - 0.69] . 0.87 [0.82 - 0.93] . |
| Var3 Yes 11 0.80 [0.71 - 0.89] 0.20 0.97 [0.94 - 1.00] 0.10 |
| No 38 0.71 [0.65 - 0.77] . 0.93 [0.91 - 0.96] . |
+-----------------------------------------------------------------------------------------+
Joint Model
+------------------------------------------------------------+
| Parameter category LRTChi2 Pvalue I2 I2lo I2hi |
|------------------------------------------------------------|
| var1 Yes 19.94 0.00 90 80 100 |
| No . . . . . |
| var2 Yes 17.18 0.00 88 76 100 |
| No . . . . . |
| var3 Yes 3.53 0.17 43 0 100 |
| No . . . . . |
+------------------------------------------------------------+
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/