Page 81 - Academic and Research Book_WEB

told by the respondent, for implicit as well as explicit meanings, it also helps in identifying themes, a
process that is associated typically with the grounded theory.

Reliability, Validity, and Standardization

Reliability, validity, and standardization are essential for the replicability of a research.

Reliability: is used to examine the accuracy or precision of a measurement instrument, questionnaire
or a focus group / qualitative interview-guide. Reliability is also an indicator of consistency in
measurement. For example, if a weighing scale consistently measures the same weight for the same
object, the weighing scale is said to be reliable.

Validity: is used to examine the accuracy, with which a test measures the construct it is intended,
planned, or required to measure. Validity is categorized into four types:

1. Construct validity: the extent to which an instrument measures, what it claims to measure;
2. Content validity: the extent to which the questions or constructs of the instrument are
representative of the concept or research questions;
3. Criterion validity: the extent to which the results from the data collected using the instrument
matches the results from the data collected using already tested or well-established and widely-
used instruments; and
4. Face validity: the extent to which the contents of the instrument are suitable to the research
question. Face validity is similar to content validity but is relatively informal and subjective.
Validity is also categorized as: (1) internal validity, which refers to the context in which the results of
a study are applicable; and (2) external validity, which refers to the extent of generalizability of the
results from a study.

Standardization: refers to the uniform procedures to ensure that the instrument is administered and
constructs measured, in the same way each time they are used or measured.

Fallacies in Quantitative Research Approaches

Quantitative research is based on a positivist paradigm and it makes use of self-administered
questionnaires or standardized interviews to generate data. These data are normally, quantitative in
nature, as they are measured on attitude scales or using metric measurements (like the age of the
respondent) and therefore, statistical methods are used to examine the hypotheses. These processes of
data collection and statistical analysis are not free from misconceptions, called fallacies. These fallacies
require to be understood and corrective actions be taken to ensure correct treatment to the data,
statistical analysis, and in turn, to drawing inferences about the research hypotheses.

Qualitative research is based on naturalistic paradigm and their data collection, analysis, and
interpretation approaches are different compared to those of quantitative research. The processes of
qualitative research are also not free from fallacies.
In the context of quantitative research, contextual variable fallacies, measurement error fallacies, and
missing data imputation fallacies are discussed here.

A contextual variable is one that is dependent on the predictor and it influences the outcome. There
are many misconceptions associated with this type of variable. The most common misconception
or fallacy with contextual variables is that it should not be confused with the interaction effect; a

70 | University Grants Commission