If the model with the population variable was significant, we then added an interaction between treatment and that variable (Survey response = Intercept + β*SATI.Comb + β*Treat + β*First.Gen+ β*First.Gen*Treat) and tested the fit of this new model to the data.
As dichotomous variables, the survey responses were interpreted as the presence or absence of the organizational characteristics of interest.
Phenotypes (outcome variables) from the survey responses were analyzed using SPSS version 13.0 for Microsoft Windows SPSSS, Inc., Chicago, IL).
Vanderbilt Hospital night-shift nurses typically work a schedule that includes 3 d on 12-h shifts (7 pm to 7 am) followed by two to five days-off, then on for 3 d, etc. Outcome variables from the survey responses were generated using SPSS 13.0 and are defined as follows.
Other than medication use, the survey variables represent fall survey responses.
Overconsumption of alcohol is a dummy variable for the quantity of the alcohol more than the beer of a large-sized bottle per day As lifestyle variables, we converted the survey responses into dichotomous variables (yes = 1) for each of the following: regular physical activity (RPA), consumption of alcoholic beverages almost every day or every day, and current smoking.
Exploratory efficacy variables were ocular symptom survey responses by visit and change from baseline to week 8, adherence questionnaire responses at week 8, and patient and investigator global preference responses at week 8.
A binary variable was created based on survey responses where "1" indicates an instructor felt there was enough time and "0" indicates feeling there was no enough time.
Colleague and manager survey responses revealed variable levels of knowledge of how decisions were made about which physiotherapy services are provided.
Explanatory models were estimated to gain insight into relationships among survey responses, external variables and demographics.
