Skip to main content
. Author manuscript; available in PMC: 2012 Oct 1.
Published in final edited form as: Am J Geriatr Psychiatry. 2011 Oct;19(10):859–870. doi: 10.1097/JGP.0b013e318202bf5b

Table 5.

The relationship between stimulus, environment, and personal attributes and refusal: results of GEE multivariate analyses

Refused
Odds
Ratio
Confidence
interval
Stimulus type Real human - refa 1.00
Inanimate social 11.69*** 5.80–23.53
Simulated social 10.85*** 5.43–21.69
Manipulative 9.09*** 4.59–17.99
Task/work 7.88*** 4.00–15.55
Reading 7.10*** 3.30–15.26
Pet 6.94*** 3.05–15.79
Identity 5.93*** 2.85–12.33
Music 6.62*** 3.24–13.49
MMSE <3 - refa 1.00
3–10 2.63*** 1.53–4.54
>=10 4.71*** 2.71–8.18
Speech clarity Clear speech - refa 1.00
Unclear speech 0.56* 0.35–0.89
No speech 1.81 0.32–10.21
# psychotropic Medications
0 - refa 1.00
1 1.28 0.72–2.26
2+ 1.78* 1.08–2.93

Note: P-values are based on Wald statistics with 1 degree of freedom for individual categories, and with k-1 degrees of freedom for the overall test of a variable with k categories.

*

p ≤ .05;

**

p ≤ .01;

***

p ≤ .001

Note: The following were not candidates for the multivariate analysis since they were not significant (P>0.2) in the univariate model: age, gender, marital status, race, count of past interest, vision, no. of medications.

The following variables entered the backward elimination procedure and were excluded from the final model due to non-significance: introduction, education, no. of diagnoses, ADL, hearing.

refa - the category is the reference value