Agriculture Reference
In-Depth Information
In literature one can find three types of validity:
(a)
The concept of development relates to
understanding the conceptual framework of a
study to be taken up by researchers. In applied
research, the conceptual frameworks are mostly
obtained as a result of previous theoretical works.
As such conceptualization is a concern of mainly
theoretical research study. Having a fixed concep-
tual part of a study, it is now the work of a
researcher to find out the periphery of a research
work, precisely to find out the dimensions of a
research study. For example, in a market study,
the dimensions may be the reputation of the
manufacturers, the quality of the product, the
acceptance of the product to the customers, the
leadership of the manufacturing organizations,
their social responsibilities, etc. Fixing the
concepts and dimensions of a research study
leads in finding out the measurement criteria.
Indicators for measuring the concepts and
dimensions identified in the previous stages must
be worked out. Indicators are the means of mea-
suring the knowledge, opinion, expectations, satis-
faction, etc., of the respondents. In doing so, a
researcher must have alternative options for judg-
ing the stability and verification of the responses
received. Generally, the studies are related to mul-
tiple dimensions; as such, it is now essential to
have a particular numerical figure taking the vari-
ous indicators into consideration. The simplest
way of getting an index is to sum up the scores
of each and every response. Thus, the formation of
index plays a vital role in bringing the responses of
the different characteristics of a particular study
from a particular respondent into numerical values
by which respondents can be compared. It is better
to have indicators measured in terms of unit-less
measures so that the ultimate index is also a unit-
less quantity.
content validity
, (b)
criteria validity
, and (c)
construct validity
. Content validity is related to a
measure of adequate coverage of a topic under
study. The coverage of a study is one of the most
important criteria in evaluating a research proj-
ect. On the other hand, criteria validity relates to
an ability to predict or estimate the existence of
some current conditions. It is mainly concerned
with the fact of the power of measuring of some
other empirical studies. However, a research
work is able to predict the correlation, and other
theoretical propositions are measured through
construct validity. Construct validity compares
the result of a particular research work with
other works. Test for reliability refers to consis-
tency when measurements are taken repeatedly.
Reliability doesn't contribute to validity; simi-
larly, a reliable instrument may not be a valid
instrument. Reliability is measured in terms of
stability and equivalence aspects. Stability refers
to the consistency of results under repeated
measurements, while equivalence aspects con-
sider the amount of error that may be introduced
by the different investigators or different samples
of the items being studied. The test of practica-
bility is measured in terms of economy, conve-
nience, and interpretability. Economy refers to a
trade-off between the budget allocated for a par-
ticular project and the budget for an ideal project.
While convenience suggests the ease of handling
an instrument or measurement, the interpretabil-
ity consideration is supported or supplemented
by detailed interaction for administering the test.
The key scoring points and evidence about the
reliability are the guide for interpreting results.
5.2.3 Stages of Techniques for
Developing Measurement Tools
The measurement tools can be developed by
following systematic stages of development. By
and large, four stages of development in mea-
surement tools can be identified:
(a) Stage one: development of concept
(b) Stage two: specifying the dimensions of the
concept developed
(c) Stage three: indicator selection
(d) Stage four: indexing
5.3
Scaling and Its Meaning
Quantitative or measurable indicators have the
privilege of being considered as mathematical
manipulations or treatments. But most of the
socioeconomic, psychometric studies are being
measured through subjective or abstract
indicators. So, it is very difficult to aggregate
