Agriculture Reference
In-Depth Information
multivariate analysis), the researcher tries to
infer or draw conclusion about the population
characteristics from the sample behaviors studied
so far
following few sections, we shall discuss the
descriptive statistics up to bivariate levels.
Multivariate and inferential statistics will be
taken up in the subsequent chapters.
through inferential
statistics.
In the
Analysis of data
Descriptive Statistics
Inferential Statistics
Estimation
Testing of Hypothesis
Point estimation
Univariate
Bivariate
Multivariate
Parametric
Nonparametric
Interval estimation
Central tendency
Simple Correlation
Multiple correlation
Dispersion
Simple Regression
Partial Correlation
Asymmetry
Multiple Regression
Association of Attributes
Path analysis
Index No
Two way ANOVA etc.
Multiple Discriminant Analysis
Time series
Canonical Analysis
Factor analysis/ Cluster analysis etc.
One way ANOVA etc.
In an attempt to summarize the information/
data, a researcher is always in search of certain
value(s) that can represent a set of information
given in a big table or otherwise. Thus, we
are always in search of such a measure, which
can describe the inherent characteristics of a
given set of information. Generally, a researcher
is in search of the two types of measures, one to
find out the central value around which the
observations are supposed to lie and another
measure for the spread/dispersion/scatteredness
of the observations.
For example, one may be interested in knowing
the 50th percentile value, that is, the value below
and above which there are half the population,
or the 3rd quartile value, that is, the value below
which there are 75% of the population and so on.
That means we are in search of certain locational
values.
According to Yule, a good measure should
have the following characteristics:
(a) It should be defined rigidly without any
ambiguity
(b) It should be based on all observations
(c) It should be easy to calculate
(d) It should be easy to understand
(e) It should be readily acceptable to mathemati-
cal treatments
(f) It should be the least influenced by sampling
fluctuations
8.2.1 Measures of Central Tendency
and Locations
Given a set of data, we are in search of a
typical value below and above which the
observations tend to cluster around. Thus, the
tendency of the observations to cluster around
a central value is known as central tendency.
Sometimes, researchers want to have an idea
about specific locations below and above which
there is a certain percentage of population.
The Different Measures of Central Tendency
Are:
(a) Mean
(b) Median
(c) Mode
(d) Midpoint average, etc.
