A Note on Proxy Variables and Instrument Variable Regression

This is a note about proxy variables and instruments for identification of structural parameters in regression models. We have experienced that in the econometric textbooks these two issues are treated separately, although in practice these two concepts are very often combined. Usually, proxy variables are inserted in instrument variable regressions with the motivation they are exogenous. Implicitly meaning they are exogenous in a reduced form model and not in a structural model. Actually if these variables are exogenous they should be redundant in the structural model, e.g. IQ as a proxy for ability. Valid proxies reduce unexplained variation and increases the efficiency of the estimator of the structural parameter of interest. This is especially important in situations when the instrument is weak. With a simple example we demonstrate what is required of a proxy and an instrument when they are combined. It turns out that when a researcher has a valid instrument the requirements on the proxy variable is weaker than if no such instrument exists

The importance of patient involvement in stroke rehabilitation

OBJECTIVE: To investigate the perceived needs for health services by persons with stroke within the first year after rehabilitation, and associations between perceived impact of stroke, involvement in decisions regarding care/treatment, and having health services needs met. METHOD: Data was collected, through a mail survey, from patients with stroke who were admitted to a university hospital in 2012 and had received rehabilitation after discharge from the stroke unit. The rehabilitation lasted an average of 2 to 4.6 months. The Stroke Survivor Needs Survey Questionnaire was used to assess the participants' perceptions of involvement in decisions on care or treatment and needs for health services in 11 problem areas: mobility, falls, incontinence, pain, fatigue, emotion, concentration, memory, speaking, reading, and sight. The perceived impact of stroke in eight areas was assessed using the Stroke Impact Scale (SIS) 3.0. Eleven logistic regression models were created to explore associations between having health services needs met in each problem area respectively (dependent variable) and the independent variables. In all models the independent variables were: age, sex, SIS domain corresponding to the dependent variable, or stroke severity in cases when no corresponding SIS domain was identified, and involvement in decisions on care and treatment. RESULTS: The 63 participants who returned the questionnaires had a mean age of 72 years, 33 were male and 30 were female. Eighty percent had suffered a mild stroke. The number of participants who reported problems varied between 51 (80%, mobility) and 24 (38%, sight). Involvement in decisions on care and treatment was found to be associated with having health services needs met in six problem areas: falls, fatigue, emotion, memory, speaking, and reading. CONCLUSIONS: The results highlight the importance of involving patients in making decisions on stroke rehabilitation, as it appears to be associated with meeting their health services needs.