Evaluating IRT models with - gllamm -.


122 views
Uploaded on:
Description
Assessing IRT models with - gllamm - . Herbert Matschinger College of Leipzig – Office für Psychiatry email: math@medizin.uni-leipzig.de. "Wellbeing Related Personal satisfaction" (HRQOL) . 5 3 - unmitigated things to evaluate wellbeing related nature of lifi. dichotomized {1} {2,3} Versatility
Transcripts
Slide 1

Evaluating IRT models with - gllamm - Herbert Matschinger University of Leipzig – Department fã¼r Psychiatry email: math@medizin.uni-leipzig.de

Slide 2

“Health-Related Quality Of Life” (HRQOL) 5 3 - unmitigated things to survey wellbeing related nature of lifi. dichotomized {1} {2,3} Mobility Self-care Usual (Daily) exercises Pain/uneasiness Anxiety/dejection

Slide 3

Data European Study of the Epidemiology of Mental Disorders (ESEMeD) 2001 - 2003 6 European nations : N = 21425 Belgium 2419 France 2894 Germany . 3555 Italy 4712 The Netherlands 2372 Spain 5473

Slide 4

Research Questions Do these 5 things measure one single measurement ? Is the entirety of supports an adequate measurement ? Do the thing parameters contrast between nations ? Do the separation parameters contrast between nations ? The amount of measurements ought to be expected

Slide 5

Random Intercept Modell 2 sets of indicators : 1) X – altered impacts 2) Z – irregular impacts De Boeck,P. what\'s more, Wilson,M. (2004). Exploratory Item Response Models: A General Linear and Nonlinear Approach . New York, Berlin: Springer.

Slide 6

1 and 2-Parameter IRT η pi = log(ï€ pi/(1-π pi ) Logit[Pr(y pi = 1; θ p )] = -β i + θ p irregular capture î¸ p  θ p0 Z i0 fixe Effekte -β i  Logit[Pr(y pi = 1; θ p )] = - β i + λ i θ p Logit[Pr(y pi = 1; θ p )] = Xâ\' pi β i + θ p Xâ\' pi λ i

Slide 7

Preparation of the information egen design = group(eurod1-eurod5 countryn) drop if design ==. contract eurod1-eurod5 countryn pattern,f(wt2) reshape long eurod,i(pattern)j(item) for num 1/5:gen itemX=0 \ supplant itemX=1 if thing ==X

Slide 8

Structure of the information (long arrangement)

Slide 9

Frequencies

Slide 10

1 - Parameter m odel for every one of the 6 nations gllamm eurod item1-item5, nocons link(logit) fam(bin) i(pattern) w(wt) adjust dab Intercept (- β) SE item1 ; -3.528638 .043579 item2 ; -5.576689 .061522 item3 ; -3.851309 .045826 item4 ; -1.809436 .033516 item5 ; -4.449809 .050426 var( θ ) = 6.4802599 .16565955 log probability = - 31545.688

Slide 11

gllapred raschu,u - [ back means ] (means and standard deviations will be put away in raschum1 raschus1) Non-versatile log-probability: - 31522.338 -3.155e+04 - 3.155e+04 - 3.155e+04 - 3.155e+04 - 3.155e+04 Log-probability:- 31545.694 gllapred raschmu,mu - [ reaction probabilities ] (mu will be put away in raschmu ) Non-versatile log-probability: - 31522.338 -3.155e+04 - 3.155e+04 - 3.155e+04 - 3.155e+04 - 3.155e+04 Log-probability:- 31545.694

Slide 12

gr7 raschmu raschum1,s([item]) t2("1 - Parameter No Differential Item Functioning - no nation effect") ylab(0(0.1)1) yline(.5) psize(120) xline(0)

Slide 13

2 – Parameter (Birnbaum) model eq discrim: item1-item5 gllamm eurod item1-item5, nocons link(logit) fam(bin) i(pattern) w(wt) eqs(discrim) adjust dab e(b)[1,10] eurod: eurod: eurod: eurod: eurod: pat1_1l: pat1_1l: item1 item2 item3 item4 item5 item2 item3 y1 - 4.8818004 - 8.507911 - 7.8598477 - 1.6312255 - 2.9198215 1.1467799 1.562892 pat1_1l: pat1_1l: pat1_1: item4 item5 item1 y1 .55341432 .27273694 3.8073739 limitation def 1 [pat1_1]item1=1 gllamm eurod item1-item5, nocons link(logit) fam(bin) i(pattern) w(wt) constr(1) frload(1) eqs(discrim) adjust speck

Slide 14

2 - Parameter Mode l Intercept SE item1 ; - 4.885114 .1492511 item2 ; - 8.501162 .3370249 item3 ; - 7.840406 .3939964 item4 ; - 1.630332 .0350471 tem5 ; - 2.919229 .0394554 var(1): 1 (0) loadings for arbitrary impact 1 item1: 3.8137476 (.13478418) item2: 4.3643911 (.20426133) item3: 5.9397444 (.32311677) item4: 2.1078878 (.05010409) item5: 1.0385574 (.03638325) LL = - 30547.222

Slide 15

gr7 intmu intum1 ,s([item]) t2("2 - Parameter No Differential Item Functioning - no nation effect") ylab(0(0.1)1) yline(.5) psize(120)

Slide 16

1 –Parameter m tribute l/impact on θ Generate two arrangements of marker variables for two diverse reference classifications Estimate the model twice for distinctive reference classes Compare the two results as for the distinctions of the settled parameters (thing troubles)

Slide 17

1 –Parameter m tribute l/impact on θ burn countryn[omit] 1 or roast countryn[omit] 2 xi3:eq f1: i .nation gllamm eurod item1 item2 item3 item4 item5,link(logit) fam(bin) adjust dab i(pattern) w(wt) nocons geqs(f1) Caveat: The yield does not let you know what contrast you have employed

Slide 18

1 –Parameter m tribute l impact on θ (reference gathering is Belgium (1)) Intercept (- β) SE impact SE Reference = Belgium item1 | - 3.510938 .0803299 France . 54711696 (.0949265) item2 | - 5.55874 .0912916 Germany - .11129763 (.0931433) item3 | - 3.833834 .0815657 Italy - .19944865 (.08868224) item4 | - 1.788838 .0753648 Netherlands .14494306 (.10088877) item5 | - 4.432465 .0842277 Spain - .21903409 (.08672157) var(1): 6.4690974 (.16557056) LL = - 31485.737 26

Slide 19

1 –Parameter m tribute l impact on θ (reference gathering is France (2)) Intercept (- β) SE impact SE Reference = France item1 | - 2.96341 .0696799 Belgium - .54803974 (.09490968) item2 | - 5.01124 .0813064 Germany - .65890334 (.08620927) item3 | - 3.28631 .0709361 Italy - .74705435 (.08142842) item4 | - 1.24127 .0651058 Netherlands - .40267533 (.09432633) item5 | - 3.88495 .0737165 Spain - .76663319 (.07938433) var(1): 6.4690974 (.16557056) LL = - 31485.737

Slide 20

Differences between „fixed“ parameters ( β) France - 2.96341 - 5.011237 - 3.28631 - 1.241268 - 3.884948 Belgium Difference - 3.510938 - 5.55874 - 3.833834  0.548 for every thing - 1.788838 - 4.432465

Slide 21

Systematics in contrasts The „fixed“ parameter rely on upon the complexity utilized for the indicator. The „fixed“ parameter are the thing troubles for the reference class of the indicator. The distinction in challenges between the two appraisals are the contrasts between the two reference classes (nations) These distinctions are the same for all things

Slide 23

Modeling nation contrasts by means of „fixed“ impacts/consequences for β The „fixed“ impacts rely on upon the reference classification Choose classification 1 (Belgium) for reference Define all conceivable association impacts between the things and the 5 shams (France to Spain) Constrain all the 5 collaboration impacts to be equivalent for every thing

Slide 24

Interactions and requirements singe countryn[omit] xi3: i.countryn*item1 i.countryn*item2 i.countryn*item3 i.countryn*item4 i.countryn*item5 for An in num 2/5 \ B in num 1/4:constraint def A _Ico2Xit = _IBco2Xit for An in num 6/9 \ B in num 1/4:constraint def A _Ico3Xit = _IBco3Xit for An in num 10/13 \ B in num 1/4:constraint def A _Ico4Xit = _IBco4Xit for An in num 14/17 \ B in num 1/4:constraint def A _Ico5Xit = _IBco5Xit for An in num 18/21 \ B in num 1/4:constraint def A _Ico6Xit = _IBco6Xit

Slide 25

gllamm grammar gllamm eurod item1-item5 _Ico2Xit-_I4co6Xit, link(logit) fam(bin)i(pattern) w(wt) nocons adjust speck constr(2/21) gllamm model with limitations: ( 1) [eurod]_Ico2Xit - [eurod]_I1co2Xit = 0 ( 2) [eurod]_Ico2Xit - [eurod]_I2co2Xit = 0 ( 3) [eurod]_Ico2Xit - [eurod]_I3co2Xit = 0 ( 4) [eurod]_Ico2Xit - [eurod]_I4co2Xit = 0 ( 5) [eurod]_Ico3Xit - [eurod]_I1co3Xit = 0 ( 6) [eurod]_Ico3Xit - [eurod]_I2co3Xit = 0 ( 7) [eurod]_Ico3Xit - [eurod]_I3co3Xit = 0 ( 8) [eurod]_Ico3Xit - [eurod]_I4co3Xit = 0 ( 9) [eurod]_Ico4Xit - [eurod]_I1co4Xit = 0 (10) [eurod]_Ico4Xit - [eurod]_I2co4Xit = 0 (11) [eurod]_Ico4Xit - [eurod]_I3co4Xit = 0 (12) [eurod]_Ico4Xit - [eurod]_I4co4Xit = 0 (13) [eurod]_Ico5Xit - [eurod]_I1co5Xit = 0 (14) [eurod]_Ico5Xit - [eurod]_I2co5Xit = 0 (15) [eurod]_Ico5Xit - [eurod]_I3co5Xit = 0 (16) [eurod]_Ico5Xit - [eurod]_I4co5Xit = 0 (17) [eurod]_Ico6Xit - [eurod]_I1co6Xit = 0 (18) [eurod]_Ico6Xit - [eurod]_I2co6Xit = 0 (19) [eurod]_Ico6Xit - [eurod]_I3co6Xit = 0 (20) [eurod]_Ico6Xit - [eurod]_I4co6Xit = 0

Slide 26

Results item1 | - 3.510822 .0803441 item2 | - 5.558637 .0913007 item3 | - 3.833719 .081579 item4 | - 1.788679 .0753829 item5 | - 4.432352 .0842393 France Ico2Xit | .5468687 .0949383 Germany Ico3Xit | - .1115037 .0931599 Italy Ico4Xit | - .1996552 .0887003 Netherlands Ico5Xit | .1447117 .1009002 Spain Ico6Xit | - .2192302 .0867413 ........................................... ........................................... I4co2Xit | .5468687 .0949383 I4co3Xit | - .1115037 .0931599 I4co4Xit | - .1996552 .0887003 I4co5Xit | .1447117 .1009002 I4co6Xit | - .2192302 .0867413 Belgium Item 1 Item 5

Slide 27

var(1): 6.4695502 (.16557072) LL= - 31485.71012285193 Compare with results from slide 19 These two models

Recommended
View more...