Week 4 Discussion: Levels of Measurement

Research Question

What is the impact of high patient-to-nurse patient ratios on nurse dissatisfaction and burnouts,
nursing errors, and patient outcomes, including morbidity and mortality rates in in-patient care?
Level of Measurement of Variables, Rationale of Selection, Considerations for Data

Analysis, and Advantages and Challenges

I. Dependent Variables
a. Patient-to-nurse ratios, nursing errors, patient morbidity, and mortality
Patient-to-nurse ratios, nursing errors, patient morbidity, and mortality will be measured
using the ratio scale. The rationale for choosing the ratio scale is that it can quantify and compare
values. For example, the scale can allow us to determine how many patients died (mortality),
how many errors were committed, the rate of disease in an in-patient facility (morbidity), and the
total number of patients under the care of a nurse (patient-to-nurse ratio). The ratio scale is the
fourth and highest level of measurement. Besides producing the order of variables, it can make
the difference between variables known and provide the “true zero” (Boslaugh, 2012).
Both descriptive statistics (such as mean, median, mode, and percentages) and inferential
statistics (like ANOVA, chi-square, and t-test) can be used to analyze patient-to-nurse ratios,
nursing errors, patient morbidity, and mortality. Without question, this is the advantage that the
ratio scale holds over the rest of the levels of measurements: it is possible to quantify, make a
comparison, and make use of the “zero point.” The scale provides the most accurate and exact
values because of this ‘absolute zero’ characteristic (Navarro, 2013). However, the challenge with
this measurement is that it requires some expertise to run ANOVA or t-test, especially when
using software applications like SPSS.

LEVELS OF MEASUREMENT 3
b. Nurse dissatisfaction and burnout
However, nurse dissatisfaction and burnout will be measured and analyzed either using
the ordinal or interval scales. The preferred method, though, is the interval scale because it offers
the possibility of measuring the order of the variables. For example, the Likert Scale can
determine the extent of burnout and dissatisfaction among nurses. A question like “rate from 0 to
10 your dissatisfaction or burnout” will be posed to nurses. Unlike ratio scales above, the only
analysis option for interval ratios is descriptive statistics (mean, median, mode, and percentages).
This is perhaps the advantage the interval scale holds over the nominal and ordinal scales (Gray
& Groove, 2020). However, it is challenging to divide and multiply variables, making using
inferential statistical analytical methods like t-test or ANOVA difficult.
II. Independent Variables
a. In-patient care facility
The in-patient care facility falls under the first level of measurement: nominal scale. The
reason is that an in-patient care facility is a non-numerical categorical variable that denotes a
distinct class (facility or special type of care). “Labels” or “tags” (such as health facilities) are
only used to identify objects but cannot be quantified on a nominal scale. The challenge in using
this level of measurement lies in the statistical analysis of the variable. Since the an-patient
facility is a “label” or “tag” and not a number, it is challenging to quantify the variable
statistically. The only way to collect and analyze data is to include a multiple-choice question,
wherein two or more facilities will be compared. For example, patients can be asked which in-
patient facility they prefer or which offers the best services. The advantage is that this scale is
easier to understand and apply.

LEVELS OF MEASUREMENT 4

References

Boslaugh, S. (2012). Statistics in a nutshell. O-Reilly Media, Inc.
Gray, J. R., & Grove, S. K. (2020). Burns and Grove’s the practice of nursing research:
Appraisal, synthesis, and generation of evidence (9 th ed.). Elsevier.
Navarro, D. (2013). Learning statistics with R. Lulu.com.