Validity measurement plays a central role in most of the controversial social issues. It
also ties up the decision making the process and the IQ controversy. The argument supports
that genetically based differences in intelligence presume that an IQ test is used to measure
knowledge possessed in measurement validity. Understanding validity and potential threats in
research validity are vital in performing excellent experimental research. Reliability is a
prerequisite for research validity. Assessing the reliability of research results involves the use
of statistical analysis to answer research questions or address the research hypothesis.
Experimental research validity is the extent to which research findings, interpretations, and
inferences are accurate, reasonable, and supported by empirical data. A range of validity
types is considered in the research methodology for quantitative analysis. They include
external and internal validity, construct, and statistical validity.
External validity relates to the generalizability of the inferences made based on
experimental findings to others with similar characteristics. It is often associated with the
importance of the problem being addressed by an empirical study. It shows concern since we
would like to assess whether the review could be useful to other people. It is a key to research
funding that says if a study could only be generalized to a particular learner group within a
setting and time, it is unlikely that funding could be secured for the review to take place for
Internal validity is a fundamental research validity concerned with the logic of the
causal-like relationship between the independent and dependent variables under examination.
It relates to the extent to which other confounding variables influence the research outcomes.
It is a prerequisite for performing external validity. Numerous groups of threats face
VALIDITY IN EXPERIMENTAL RESEARCH 3
experimental research from internal validity. Threats posed are treats related to research
participants and risks related to research instruments and procedures. A good design allows
us to satisfy conditions of causality effectively and includes association, temporal order, and
non-spuriousness. The base of achieving internal validity has an excellent solid research
Sampling can be defined as the process where a few are studied to learn about the
many. Its success solely depends on how good a job we do at selecting samples that
accurately represent the larger group. Random sampling is the method of choosing
participants that are representative of the target learner population. Several other common
types of validity include criterion-related, statistical, predictive, content, construct, and face
validity. Content validity is the extent to which sample behaviours or abilities are relevant to,
and representative of the construct is defined. It relates to the construct validity of the
measurement, shows a construct of interest in transformation into something measurable yet
representative of the construct.
Construct validity is the degree to which the construct of interest is validly defined,
measured, and inferred. Levels in experimental research for construct validity include the
construct validity of research instruments and the construct validity of empirical studies.
They are interrelated in that if a research instrument lacks construct validity, it is difficult to
discuss on the construct validity of the study. The validity of a research instrument is always
with the extent to which a tool measures what it is intended to measure. It is critical to
evaluate whether we agree or disagree with other researchers or theorists on the construct of
interest. It is a necessity to have an understanding of how to propose ways that define a
construct theoretically and operationalize it in research. A construct definition supported by
several empirical studies from various researchers is likely to be more valid than construct
newly introduced with little practical support.
VALIDITY IN EXPERIMENTAL RESEARCH 4
Statistical validity is about making inferences about an experimental study via the use
of statistics, such as comparing performances of two or more groups exposed to different
treatments. It is vital to ensure that we have performed sound statistical analyses that lead to
inferences and conclusions. It evaluates whether an observed causal-like relationship between
the dependent and independent variables are more likely to exist. Predictive validity relates to
the level of predictability of current test scores, and reported behaviours to future scores.
Criterion-related validity is linked to constructing validity in the sense that an instrument
should have a stronger relationship with other tools that measure the same or similar
construct. This is where we tend to find a moderate correlation between reading and writing
scores since some shared language abilities are useful for both writing and reading. Face
validity relates to the appearance of a research instrument being used or an experiment being
conducted. Experimental studies got face validity when it has both experimental and control
groups and uses a random assignment method.
It is essential that we carefully and reflectively consider a range of validity types and
the threats involved so that we can limit our ability to infer a casual-like relationship.
Research paradigms play vital roles in research validity discussions at an epistemological,
ontological, and methodological level. In experimental research, asking ourselves whether the
study has followed required methodological procedures, data has been properly collected and
analyzed to the acceptable standards. It is inevitable to consider research validity without
mentioning research reliability. Reliability is closely related to validity in that experimental
study cannot be valid if it uses unreliable data to analyze and answer the questions. It
ascertains the correctness and legitimacy of data. It is crucial to ensure our research findings
are based on trusted data and analysis.