Measurement, Statistics, and Appraisal
Introduction
The final components of the research process are data collection/measurement, data analysis
and interpretation of the results.
Quantitative Measurement
Measurement is a process for assigning numbers to objects, events, or situations in a consistent
manner. Measurement must begin by clarifying what is to be measured. Each variable in a
quantitative study must be able to be reduced to a number.
Instrumentation is a part of measurement. Instruments are developed according to specific
standards to examine specific variables. The purpose of instrumentation is to produce evidence
that is trustworthy in evaluating research outcomes. Instruments need to be reliable and valid.
Reliability refers to consistency or precision, whereas validity indicates that the instrument
actually measures what it is supposed to measure.
Measurement Strategies
Strategies must be developed to measure the object, element, or characteristic being studied.
Direct measurement is simply measuring concrete factors, such as height and weight. Indirect
measurement measures attributes that represent a concept. When the object to be measured
is abstract, a conceptual definition is necessary to clarify what is being measured and to select
appropriate means of measurement (Cooper & Schindler, 2003).
Measurement Error
There will always be errors in measurement, because there is no perfect measure. The
difference between what exists and what is measured by the research tool is called
measurement error, and is found in both direct and indirect measures. Two types of
measurement error are random and systematic.
Random error is a fluctuation in the statistics in measured data due to limitations of the
measurement device. This type of error usually results from the researcher's inability to make
the same measurement exactly the same way to get exactly the same results. Factors that can
result in random error include transient personal factors, such as hunger and motivation;
situational factors, such as distraction and environment; variations in administration of the
measurement, such as the use of different wording in interviewing; and processing of data,
such as recording errors.
The systematic error occurs when a problem persists throughout the research study and
inaccuracies are reproduced. Some systematic error will occur in almost every measure.
Therefore, closely linking the abstract theoretical concept with the instrument, and refining the
instrument, will greatly decrease systematic error. The use of more than one measure of a
concept will also decrease systematic error (Simon & Francis, 2001).
Levels of Measurement
Data generated by an instrument may be at nominal, ordinal, interval, or ratio level of
measurement, depending on the instrument. Nominal is the lower end of the scale, and ratio is
the higher end. Nominal-scale measurement is used when data can be organized into
categories, but the categories cannot be compared. Examples of nominal data are gender and
race. Numbers are assigned to nominal categories as labels, not for mathematical calculations.
Limitations and Assumptions
Limitations are restrictions that may decrease the generalization ability, extension of the study
sample to a larger population, or the findings of the study. Two types of limitations are
conceptual and methodological. Conceptual limitations stem from the limits within the
theoretical framework due to unclear concepts or relationships and limit the abstract
generalization of the study findings. They are reflected in the frame of reference, conceptual
definitions, and operational definitions. The methodological limitations limit the credibility of
the findings and restrict the population to which the research findings can be applied. These
limitations result from factors such as weak design and inaccurate use of statistical analysis.
Data Collection
Data collection is the gathering of relevant information in a precise, systematic manner. Data
collected for quantitative studies are numerical. Numerous methods exist for data collection
including observation, scales, and questionnaires.
Data Analysis
Data analysis is the process of applying statistical techniques systematically to describe data.
Data analysis is conducted to organize, reduce, and give meaning to data.
The level of measurement is associated with the type of statistical analysis. Therefore, the
highest possible level of measurement should be used. Data can be analyzed by hand, by using
computers, or by statisticians. Data analysis techniques depend on the research design,
researcher expertise, and type of data collected (Brink & Wood, 1998).
Research Outcomes
Data collection and data analysis are action-oriented, and when they are completed, activity
moves to abstract thinking through introspection and reasoning to synthesize and evaluate the
entire research process, organize the meaning of the results, and forecast the usefulness of the
findings. During this process, the researcher forms conclusions from the data, implications for
further research, and implications for nursing practice (Brink & Wood, 1998).
Qualitative Data Collection Methods
Qualitative research requires a simple data collection method because the complexity of the
research is in the data itself. If the researcher is too complex in his or her method, the reaction
between the complex data and the complex method could be disastrous. Common data
collection methods include observation, interviewing, and examining historical documents or
written texts. A qualitative inquiry or interview should include understandable, and rather
simple, questions.
Qualitative research should start with asking simple questions that evolve into complex
answers. Using qualitative interviews allows the researcher to gather detailed, rich information.
When doing interviews, it is necessary to reach the saturation level, which means the
information is repeated. Robson (2002) offers some helpful advice when doing an interview for
a qualitative study. The interviewer should encourage the participants to talk freely. This can be
encouraged in the following ways:
Interviewers must listen more often than they speak, and should not interject their personal
opinions.
Questions should be delivered in a clear, concise, and nonthreatening way. Interviewers should
not put participants on the defensive nor try to confuse them.
The researcher should not lead the participant in answering the questions a certain way or in a
way the researcher wants the questions answered.
Data Analysis
Unlike quantitative data analysis, qualitative data analysis actually begins when the researcher
is still collecting data. The researcher will usually gather large amounts of textual data in some
fashion. There are three stages of data analysis:
•
Descriptive: The researcher becomes immersed in the data, and finds it necessary to read
and reread the information collected. If the researcher misses anything during this period of
time, important information may be lost.
•
Analysis: During this phase, themes and patterns from the data collected are identified.
These themes and patterns are relationships that the researcher feels are present, based on
what the participant has said.
•
Interpretation: The researcher offers his or her explanation on what is happening with the
study. Researchers develop tentative propositions and form them into a tentative theory.
The validity of the theory is then tested by predictions, which means the theory must be
tested with a similar or same-sample population.
Qualitative data may be hard to assemble and organize because of the voluminous amount of
information. One challenge of qualitative research is data reduction because everything may
look more important than it is. Thousands of pages must be reduced to a short report. Bogdan
and Biklen (1998) suggest that "the process of data analysis is like a funnel: Things are open at
the beginning (or top) and more directed and specific at the bottom" (p. 7).
Data Presentation
There are certain strategies for data presentation:
•
Natural, in a way that resembles the phenomenon being studied in a sequential order
•
From most simple to most complex
•
According to themes or patterns
•
By the researcher's theory or theories regarding the phenomenon
•
Arranged for storytelling
•
From the most important to the least important (major to minor)
•
Dynamic, where the researcher will save discoveries or surprises for the end of the study
Conclusion
Understanding the components of quantitative studies is critical to advance nursing knowledge
and improve patient outcomes. By mastering quantitative research, nurses will be able to
implement change and improve the care that they provide to patients, families, and
communities.
Qualitative research is an investigative process in which the researcher attempts to make sense
of a social phenomenon by replicating, cataloguing, contrasting, or classifying the object of the
study. Therefore, the researcher is entering the participants' world by trying to understand
their perspectives and meanings.
References
Bogdan, R. C., & Biklen, S. K. (1998). Qualitative research for education: An introduction to
theory and methods. Boston: Allyn & Bacon.
Brink, P. J., & Wood, M. J. (1998). Advanced design in nursing research (2nd ed.). Thousand
Oaks, CA: Sage.
Cooper, D. R., & Schindler, P. S. (2003). Business research methods (8th ed.). Boston: McGrawHill Irwin.
Robson, C. (2002). Real world research: A resource for social scientists and practitionerresearchers (2nd ed.). Oxford, United Kingdom: Blackwell Publishing.
Simon, M. K., & Francis, J. B. (2001). The dissertation and research cookbook (3rd ed.).
Dubuque, IA: Kendall/Hunt.
© 2012. Grand Canyon University. All Rights Reserved
Purchase answer to see full
attachment