According to Santayana (2008, p.1), “measurement is at the core of doing research”. Generally, after conducting research, researchers have to reduce their finding to numbers in order for people to understand their study findings. Such measurements are exceptionally essential in the measurement of hypotheses. Researchers therefore must come up with sound measurement tools for both independent and dependent variables.
We will write a custom Essay on Process Involved in Choosing a Measurement Tool for Variables in Research specifically for you
301 certified writers online
Measurement is not as easy as it sounds. It is composed of four basic principles. The first two, conceptualization and operationalization are a bit simple. The other two are more advanced and complicated. They include determining the levels of measurement and measuring validity and reliability. Conceptualization involves defining a concept.
Thus, researchers must define that particular concept in a manner that is understandable by the scientific community. Researchers are allowed to borrow conceptualization from their predecessors and other researchers. On the other hand, operationalization involves making the conceptual definition more precise. This often involves linking the conceptual definition to one or more indicators.
The indicators chosen must reflect observable reality in numbers. For example, in the parent involvement study, the conceptual definition is ‘parents’ involvement’. However, the operational definition is ‘parents’ role construction’. It is also important when giving an operational definition to establish the unit of analysis.
In addition, a researcher needs to evaluate the level of measurement. Santayana (2008, p.1) indicates that “the level of measurement is the precision by which a variable is measured and there are four levels of measurement namely nominal, ordinal, interval, and ratio”. The four levels are named from the lowest to the highest. On the other hand, Stevens (1951, p. 1) states that “everything that can be measured falls in the aforementioned levels and the higher the type, the more the precision in measurement”.
Variables that are categorical in nature usually fall in the nominal level while those that have significant importance fall in the ordinal level. Variables that have significant distances between their ranks fall in the interval level. On the other hand, variables that have fixed intervals and a fixed reference point are grouped in the ratio level.
It is for these reasons that Creswell (2011) holds the opinion that researchers must define their variables conceptually and operationally while putting into account the levels of measurement since these factors eventually determine how the data collected is analyzed. In the parent involvement study, ‘parents’ involvement’ falls in the ordinal level.
Creswell (2011) also notes that accurate researches are accurate and valid. Validity can be measured using four methods namely face, content, criterion and construct. On the other hand, reliability can be estimated using the following methods test retest, multiple forms, inter rater, and split half.
Therefore, in order to locate an instrument that is to be used to measure a given variable, a researcher must put into account a number of factors. First and foremost, according to Slavin (1984), the researcher must define his concept conceptually and operationally. After that, the researcher must establish the level of measurement in which his or her variable falls.
In addition, the researcher must detect the rationales he or she will use to estimate validity and reliability. The achievement of this goal relies on having proper conceptual and operational definition and the establishment of the level of measurement of the variable.
Cresewell, J. (2011). Educational Research: Planning, Conducting, and evaluating quantitative and qualitative research. New York: Pearson.
Santayana, G. (2008). Measurement, Reliability, and Validity. Web.
Slavin, R. (1984). Research Methods in Education: A Practical Guide. New Jersey: Prentice Hall.
Stevens, S. (1951). Handbook of Experimental Psycholog. New York: Weley.