Nursing Research: Measurement and Scales

The focus of measurement is the quantification of a characteristic or attribute of a person, object, or event. Measurement provides for a consistent and meaningful interpretation of the nature of an attribute when the same measurement process or instrument is used. The results of measurement are usually expressed in the form of 

numbers. Measurement is a systematic process that uses rules to assign numbers to persons, objects, or events which represent the amount or kind of a specified attribute. However, measurement also involves identifying and specifying common aspects of attributes for meaningful interpretation and categorization, using a common conceptual perspective. Ambiguity, confusion, and disagreement will surround the meaning of any measurement when it is undefined.

Scale

The measurement relevancy can be determined only when an explicit or implicit theory structures the meaning of the phenomenon to be studied. Qualitative assessments apply measurement principles by providing meaning and interpretation of qualitative data through description and categorization of phenomena. Thus, measurement may not result in scores per se but may categorize phenomena into meaningful and interpretable attributes. Therefore, measurement is also basic to qualitative analysis. Measurement is a crucial part of all nursing settings. Nurses depend on measuring instruments to determine the amount or kind of attributes of patients and use the results of measurements such as laboratory and physical examination results to determine patient needs and their plan of care.
Nurse researchers use a large array of clinical laboratory, observational, and questionnaire measures to study phenomena of interest. Nurse educators depend on measurement instruments and test scores to help determine a student’s mastery. Measurement is central to all that nurses do. The rules used for assigning numbers to objects to represent the amount or kind of an attribute studied have been categorized as nominal, ordinal, interval, and ratio. These types of measurement scales are common in nursing. Measurements that result in nominal-scale data place attributes into defined categories according to a defined property.
Numbers assigned to nominal-level data have no hierarchical meaning but represent an object’s membership in one of a set of mutually exclusive, exhaustive, and unorderable categories. For example, categorizing persons in a study as either female or male is measurement on the nominal measurement scale. In ordinal-scale measurement, rules are used to assign rank order on a particular attribute that characterizes a person, object, or event. Ordinal-scale measurement may be regarded as the rank-ordering of objects into hierarchical quantitative categories according to relative amounts of the attribute studied. The categorization of heart murmurs in grades from 1 through 6 is an example. In this ordinal measure, a Grade 1 murmur is less intense than a Grade 2, a Grade 2 less intense than a Grade 3, and so forth. The rankings in ordinal-level measurement merely mean that the ranking of 1 (for first) has ranked higher than 2 (for second) and soon. Rankings do not imply that the categories are equally spaced nor that the intervals be-tween rank categories are equal.
Interval-scale measurement is a form of continuous measurement and implies equal numerical distances between adjacent scores that represent equal amounts with respect to the attribute that is the focus of measurement. Therefore, numbers assigned in interval-scale-measurement represent an attribute’s placement in one of a set of mutually exclusive, exhaustive categories that can be ordered and are equally spaced in terms of the magnitude of the attribute under consideration. However, the absolute amount of the attribute is not known for a particular object because the zero point is arbitrary in an interval scale. The measurement of temperature is a good example of an interval-level measure because there is no true zero point. For example, the zero point is different based on whether the Fahrenheit or Centigrade measurement approach is used, and one cannot say that an object with a temperature of 0°For0° C has no temperature at all. Ratio level measures provide the same information as interval-level measures; in addition they have absolute zero points for which zero actually represents absence of the attribute under study. Volume, length, and weight are commonly measured by ratio scales.
There is controversy about the level of measurement scales and the type of statistical procedures that may be appropriately used for data analysis. There are researchers and statisticians who believe that only nonparametric statistical procedures can be used for data analysis when data are nominal or ordinal and that inferential statistics can be properly applied only with interval and ratio data. There is controversy about whether Likert scaling (which is often used in nursing with measures of attitude or opinion) is in actuality ordinal-level measurement for which only nonparametric statistics should be used. Likert scaling involves having subjects rank their responses to a set of items on a range of numbers, such as “1” to represent lack of agreement to “5” to represent complete agreement. It has been the accepted practice for investigators to use scores generated with Likert-type scales as interval-level data.
Nurses have typically borrowed many measures from other disciplines. This reflects the fact that nursing is a field that considers the biological and psychosocial aspects of care and is based on knowledge generated by many fields of inquiry. Therefore, many instruments developed by other disciplines are consistent with nurses’ measurement needs. However, the heavy dependence on borrowing instruments from other disciplines reflects the trend in the 1970s for nurses to pursue doctoral education in related fields, such as education, psychology, sociology, and physiology. Nurses became familiar with instruments from other fields during their graduate studies and were encouraged to use them in the nursing context.
By the mid 1970s nurses became more cognizant of some of the limitations in borrowing certain instruments from other disciplines. It is not unusual for instruments developed to measure psychosocial variables in other fields to be cumbersome and inefficient for use in the clinical settings of nurse researchers. Often the instruments needed by nurses to measure attributes in populations such as children, frail patients, the elderly, and the culturally diverse, instruments that measure important variables from the nursing perspective, do not exist. Nursing studies of families, communities, and organizations and systems have been hampered by the lack of effective measures to address group and system variables from the nursing perspective (Strickland, 1995).
The movement in nursing to develop more rigor in the use and development of measurement instruments gained prominence in the1970s. In June 1974, a contract was awarded to the Western Interstate Commission for Higher Education by the Division of Nursing, Bureau of Health Manpower, and Health Re-sources Administration to prepare a compilation of nursing research instruments and other measuring devices for publication.
With Doris Bloch as project officer, a two-volume compilation of instruments, titled Instruments for Measuring Nursing Practice and Other Health Care Variables, was published in 1978. Priority was placed on compiling instruments dealing with nursing practice and with patient variables rather than nurse variables. This was an important milestone for nursing measurement because it was the first effort that placed a large number of clinically focused instruments developed or used by nurses in the public domain.
During the late 1970s and early 1980s, two groups of nurse scientists focused their work on developing measurement as an area of special emphasis in nursing. At the University of Arizona–Tucson, Ada Sue Hinshaw and Jan Atwood focused their efforts on refining and further developing instruments for clinical settings and for clinically focused research. The first postdoctoral program in nursing instrumentation and measurement evolved at the University of Arizona, and annual national conferences on nursing measurement were offered. These conferences provided nurses a forum in which to discuss measurement issues and problems and to present information on instruments used in studies. Ora Strickland and Caroyln Waltz at the University of Maryland at Baltimore focused on defining measurement principles and practices to build rigor in nursing research. Careful assessments of nursing research published in professional journals were conducted. The assessments revealed that nurse investigators were not giving adequate attention to reliability and validity issues when selecting and developing instruments. In addition, nurse investigators tended to rely too heavily on paper-and-pencil self-report measures and did not give adequate attention to selecting biological measures as indicated by the conceptual frameworks of the studies (Strickland & Waltz, 1986).
The Maryland group published the first measurement textbook for nurses, Measurement in Nursing Research, and developed and implemented a measurement project funded by the Division of Nursing of the Department of Health and Human Services. This project prepared over 200 nurse researchers in clinical or educational settings to develop and test instruments for use in nursing and resulted in an award-winning four-volume series of books, Measurement of Nursing Outcomes, which compiled instruments developed by project participants. In 1993, Ora Strickland initiated and edited the Journal of Nursing Measurement with Ada Sue Hinshaw as coeditor. This journal brought nursing measurement to a new level of focus, responding to the need for continuing development and dissemination of nursing measurement instruments and providing an identifiable forum for the discussion and debate of measurement concerns and issues of interest.

The nursing profession has developed nursing measurement to a great degree between the late 1970s and the present. Nurses have developed and tested instruments for use in a variety of settings. In addition to creating new instruments, nurses have further developed instruments designed in other disciplines for use in nursing studies. Although much has been done, much remains to be done in nursing measurement.


EmoticonEmoticon