Pradip Kumar Sahu

Research methodology : a guide for researchers in agricultural science, social science and other related fields

1.Scientific process and research.- 2.Research Process.- 3.Research problems.- 4.Research Design.- 5.Variables, Measurement and Scaling technique.- 6.Sampling Design.- 7.Collection of data.- 8.Processing and analysis of data.- 9.Formulation and testing of hypothesis.- 10.Analysis of variance and Experimental designs.- 11.Analysis related to breeding researches.- 12.Multivariate analysis.- 13.Instrumentation and computation.- 14.Research proposal and Report writing.- References.- Index.

Collection of Data

Research is a process of knowing the unknown things for the betterment of humanity. In the process, explorations of the different sources of information are carried out by a researcher to expose the so long unexposed truth. Thus, information is the prerequisite for achieving the objectives of any research program. Information may be qualitative or quantitative/numerical. Data refers to different kinds of numerical information. In any research program, a researcher is always in search of a suitable mechanism/process for the collection of information. Thus, data collection plays a vital role in any research program. Depending upon the research design, in particular, the objective of a research program and the types of information required are fixed. The next task is data collection. Depending upon the sources of information, data may be (1) primary or (2) secondary.

Methods of Estimation

In chapter one, we have discussed different optimum properties of good point estimators viz. unbiasedness, minimum variance, consistency and efficiency which are the desirable properties of a good estimator.

Likelihood Ratio Test

In the previous chapter, we have seen that UMP or UMP-unbiased tests exist only for some special families of distributions, while they do not exist for other families.

On Assessment of Sustainability of Crops and Cropping System—Some New Measures

Abstract Long-term experiments are the means to evaluate the sustainability status of crops/cropping systems (measured in terms of yield). Different nutrient management practices (developed by taking combinations of organic and inorganic nutrients in different proportions) are used in long-term experiments and such practices are evaluated in terms of their sustainability status. The existing measures are only a few in number and they have their own limitations. This paper considers the development of new measures on assessing sustainability of sole crops and also under cropping system. These measures are considered to be more useful and appropriate.

Sustainability of Different Nutrient Combinations in a Long-Term Rice-Wheat Cropping System

ABSTRACT Sustainability of any cropping system across years influences the food safety in a particular region, as cropping systems are mostly region/zone specific. The present study examined the yield sustainability of 12 nutrient treatments that were combinations of organic and inorganic sources (inorganic, organic or different combinations of both) in a rice-wheat cropping system, using information from experiments conducted across 17 years in the same field. Existing statistical measures along with a proposed measure of sustainability was used to compare efficiency of different nutrient treatments. Results revealed that highest productivity did not necessarily mean highest sustainability; mostly the treatments with combination of inorganic and 25-50% organic nutrients in rice not only yield more, but they were also moderately sustainable in nature. Clustering analysis showed that control treatment (T1) itself formed a group, whereas the treatment with only inorganic fertilizer at different proportions of recommended fertilizer dose during rice season formed another group and the treatments in combinations of both the types of nutrients at different proportions and levels formed a different group relative to their yield response in the component crops, as well as in the rice-wheat cropping system.

Analysis of Variance and Experimental Designs

In Chap. 9, a discussion has been made as to how two sample means can be compared using τ or t-test. Problem arises when one wants to compare more than two populations at a time. One of the possible solutions to this problem is to take \( ^{\mathrm{ m}}{{\mathrm{ C}}_2} \) no. of pairs of samples and test these using the τ or t-test as applicable. Another important procedure is to use the analysis of variance technique. The analysis of variance technique, in short ANOVA, is a powerful technique used in the field of agriculture, social science, business, education, medicine, and several other fields. Using this tool the researcher can draw inference about the samples whether these have been drawn from the same population or they belong to different populations. Using this technique the researcher can establish whether a no. of varieties differ significantly among themselves with respect to their different characteristics like yield, susceptibility to diseases and pests, nutrient acceptability, and stress tolerance and efficiency of different salesmen; for example, one can compare different plant protection chemicals, different groups of people with respect to their innovativeness, and different drugs against a particular disease, different programs of poverty reduction, performances of different business houses, and so on.

Basic Experimental Designs

Statistical tolls or techniques are used to extract information, which otherwise remain hidden, from a set of data. As has been mentioned earlier, data can be gathered/collected from existing population (through sample survey technique/census method) or can be collected by conducting experiment as per the objective of the experimenter. In the first case, the researcher has little choice of controlling the external factors while collecting information from the existing population; the maximum the researcher can do is to orient the collected data from a befitting sample so as to explain the objective in mind. This type of data collection is mostly used in social, economical, political, and other fields. On the other hand, in the second option, the researcher has greater control over the data to be collected for specific purpose through experimentation. The researchers can exercise control to the extraneous factors to some extent allowing the desirable factors to vary. To examine the performance of different varieties of paddy with respect to yield, the experimenter can select the varieties as per the objective of the program and put all the varieties essentially under the same protocol so that only the source of variation can be the varieties. In this chapter we are concerned about such experimental procedure, collection of data and their analyses toward meaningful inference about the objectives the experimenter has in mind.

Non-parametric Test

In parametric tests we generally assume a particular form of the population distribution (say, normal distribution) from which a random sample is drawn and we try to construct a test criterion (for testing hypothesis regarding parameter of the population) and the distribution of the test criterion depends upon the parent population.

Estimation and Inferential Statistics

1. Theory of Point Estimation.- 2. Methods of Estimation.- 3. Theory of testing of hypothesis.- 4. Likelihood Ratio Test.- 5. Interval Estimation.- 6. Non-parametric test.- 7. Statistical Decision Theory.