norcag: Attitudes Towards Sex Data from the 1989 General Social...

In mixor: Mixed-Effects Ordinal Regression Analysis

Description

Each subject provided ordinal responses on three items concerning their opinion on early teens (age 14-16) having sex before marriage (Item1), a man and a woman having sex before marriage (Item2), and a married person having sex with someone other than their spouse (Item3). Data are provided as frequencies by response pattern.

Usage

data(norcag)

Format

A data frame with 105 observations on the following 6 variables.

ID

a numeric vector indicating unique patient identifier

SexItems

ordinal item response coded as 1 = always wrong; 2 = almost always wrong; 3 = wrong only sometimes; 4 = not wrong

int

a numeric vector of ones; used in the stand-alone MIXOR program to indicate the intercept

Item2vs1

attitude towards premarital vs teenage sex

Item3vs1

attitude towards extramarital vs teenage sex

freq

frequency weight of the pattern

Source

Agresti A. and Lang J.B. (1993) A proportional odds model with subject-specific effects for repeated ordered categorical responses, Biometrika 80, 527-534.

References

Hedeker D. and Mermelstein R.J. (1998) A multilevel thresholds of change model for analysis of stages of change data, Multivariate Behavioral Research 33, 427-455.

Examples

library("mixor")
data("norcag")
# random intercepts model assuming proportional odds for differences in item responses
Fitted.norcag<-mixor(SexItems~Item2vs1+Item3vs1, data=norcag, id=ID, 
     weights=freq, link="logit", nAGQ=20)
summary(Fitted.norcag)
# random intercepts model assuming non-proportional odds for differences in item responses
Fitted.norcag.np<-mixor(SexItems~Item2vs1+Item3vs1, data=norcag, id=ID, 
     weights=freq, link="logit", nAGQ=10, KG=2)
summary(Fitted.norcag.np)
# SCALING model 
Fitted.norcag.scale<-mixor(SexItems~Item2vs1+Item3vs1, data=norcag, id=ID, 
     weights=freq, link="logit", nAGQ=10, KS=2)
summary(Fitted.norcag.scale)

Example output

Loading required package: survival
Warning message:
In model.matrix.default(mt, mf, contrasts) :
  non-list contrasts argument ignored

Call:
mixor(formula = SexItems ~ Item2vs1 + Item3vs1, data = norcag, 
    id = ID, weights = freq, nAGQ = 20, link = "logit")

Deviance =         2436.854 
Log-likelihood =  -1218.427 
RIDGEMAX =         0.1 
AIC =             -1224.427 
SBC =             -1236.917 

                    Estimate Std. Error  z value   P(>|z|)    
(Intercept)        -2.081101   0.202849 -10.2594 < 2.2e-16 ***
Item2vs1            3.807599   0.268844  14.1628 < 2.2e-16 ***
Item3vs1           -0.570918   0.197825  -2.8860  0.003902 ** 
Random.(Intercept)  5.139592   0.959090   5.3588 8.377e-08 ***
Threshold2          1.134229   0.098389  11.5280 < 2.2e-16 ***
Threshold3          2.783234   0.186590  14.9163 < 2.2e-16 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
Warning message:
In model.matrix.default(mt, mf, contrasts) :
  non-list contrasts argument ignored

Call:
mixor(formula = SexItems ~ Item2vs1 + Item3vs1, data = norcag, 
    id = ID, weights = freq, nAGQ = 10, link = "logit", KG = 2)

Deviance =         2401.609 
Log-likelihood =  -1200.805 
RIDGEMAX =         0.1 
AIC =             -1210.805 
SBC =             -1231.621 

                   Estimate Std. Error z value   P(>|z|)    
(Intercept)        -1.81905    0.19270 -9.4398 < 2.2e-16 ***
Item2vs1            3.15542    0.26733 11.8035 < 2.2e-16 ***
Item3vs1           -0.59966    0.19905 -3.0125  0.002591 ** 
Random.(Intercept)  4.40011    0.80989  5.4329 5.543e-08 ***
Threshold2          1.59227    0.18510  8.6020 < 2.2e-16 ***
Threshold3          3.21192    0.33989  9.4499 < 2.2e-16 ***
Threshold2Item2vs1  0.96062    0.20618  4.6591 3.176e-06 ***
Threshold3Item2vs1  1.06468    0.36491  2.9176  0.003527 ** 
Threshold2Item3vs1  0.26318    0.26612  0.9889  0.322690    
Threshold3Item3vs1 -0.61730    0.65391 -0.9440  0.345168    
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
Warning message:
In model.matrix.default(mt, mf, contrasts) :
  non-list contrasts argument ignored

Call:
mixor(formula = SexItems ~ Item2vs1 + Item3vs1, data = norcag, 
    id = ID, weights = freq, nAGQ = 10, link = "logit", KS = 2)

Deviance =         2418.607 
Log-likelihood =  -1209.303 
RIDGEMAX =         0.2 
AIC =             -1217.303 
SBC =             -1233.957 

                   Estimate Std. Error z value   P(>|z|)    
(Intercept)        -2.22116    0.33600 -6.6107 3.826e-11 ***
Item2vs1            4.50668    0.69147  6.5175 7.149e-11 ***
Item3vs1           -0.62054    0.36986 -1.6778  0.093394 .  
Scale.Item2vs1      0.63125    0.19665  3.2100  0.001328 ** 
Scale.Item3vs1     -0.01540    0.31656 -0.0486  0.961200    
Random.(Intercept)  6.78761    2.47036  2.7476  0.006003 ** 
Threshold2          1.39174    0.23492  5.9242 3.138e-09 ***
Threshold3          3.61363    0.57930  6.2379 4.435e-10 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

mixor documentation built on May 2, 2019, 5:56 a.m.