2.4. Meta-Analysis: Method of analysis using diverse datasets
Meta-analysis is a potential statistical approach to analyze the
reaction of managements (BL/CL/GL/HL/PL) to control treatment (FL) from
differentdistinct studies evolving a universal trend. This study has
attracted researchers from all over the world for better understanding
of all the variables.
Two stage-based Meta-analyses (MetaWin 2.1) were used to analyse the
database and understand the comparative changes (Chakraborty et
al ., 2017; Rosenberg et al ., 2000). Under this the effect size
(ES) was calculated for individual parameter using the equation as
proposed by Hedges et al. (1999):
\begin{equation}
ES\ =\ lnR=\ln\left[\frac{X_{T}}{X_{C}}\right]\nonumber \\
\end{equation}Where,
XT-Average of response variables (SOC, SCS, yield and
other parameters) of the treatments (LU),
XC-Average of these variables in FL with control
Since, all studies were from variable conditions was the basis of
multiple replications, the standard deviation calculated were based on
the number of observations with simple statistically procedure.
ES from individual studies were then combined using a mixed-effect model
to calculate the cumulative effect size and the 95% confidence
intervals (CIs) through bootstrapping with 4999 iterations (Adamset al ., 1997). The mixed-effect model is a random-effect
Meta-analytic model for categorical data (Rosenberg et al .,
2000), assuming random variation among studies within a group and fixed
variation between groups. The cumulative effect was considered
significant if the CIs did not pass over zero.
Interpretation of results : Results were back-transformed and
presented as change in percent caused by treatments in relation to
control. Significant differences were considered only p value is less
than 0.05. The Meta-analysed value has been presented in graph to
clearly show the significant effect of LUC and INM over the inorganic
fertilizer alone.