Nurses' retention and hospital characteristics in New South Wales, CHERE Discussion Paper No 52

NURSES’ RETENTION AND HOSPITAL CHARACTERISTICS IN NEW SOUTH WALES

Table 3 presents a list of the variables used in the econometric model along with their means.
Variables include controls for personal characteristics (sex, age, country of birth, citizenship, post
basic education, years since first registered as a nurse); job characteristics (hours of work, activity,
and job classification), hospital type dummies, and hospital characteristics. The NRB data include
more information on the job than we include in this analysis. Some of the variables (e.g. job premise)
become almost synonymous with the hospital types and others are not informative once we restrict
ourselves to public hospital RNs (e.g. field of work). Dummy variables for regions are also not
included as they do not add explanatory power over and above the hospital types.

Table 3. Variable Means.

* refers to an omitted group in regressions.

Since the survey is self reported, missing data items are not uncommon. In some instances,
personal information such as age and year of first registration could be reconstructed from
responses provided in other years by the same nurse. For job characteristics this is not possible
since the job may change over time. Rather than exclude all observations with missing values for any
of the variables, we construct dummies controlling for missing information for the variable in
question. For categorical variables (activity, job classification, post-basic qualifications) this dummy
acts as a new group. For continuous variables (age and hours), the dummy is defined similarly as a
binary variable and the main variable (age and hours) is given a value equal to the mean over non-
missing values. Hence the dummy for missing information in this case represents a shift around the
mean value.

<<   8   9   10   11   12   13   14   >>

More intriguing information