Dear Bijesh,

I hope you find some of my publications (you can find them on my profile) useful. I attach one of them.

Regards,

Alfredo J. Escribano.

Data Comparative Sustainability Assessment of Extensive Beef Catt...

Dear Bijesh Mishra,

My last thesis was similar to your research; so, I can give you some guides.

1/ You control  first practice as factor = 1 and others as factor =0; run logit analysis on SPSS

2/ Change to the the second practice as factor = 1 and others as factor =0; run logit analysis on SPSS

3/ do the same ways to the last.

4/ Compare the results: Choose the model:which has the betas of Xi  being the bigest.

Good luck.

Le Xuan Thai

Dear Bijesh,

I hope you find some ideias from my publications (its a pith that all my papers are in Portuguese). I will attach it for you at follow.

Regards,

Miguel Angelo Perondi.

What are the indicators to measure agriculture sustainability in farm level? - ResearchGate. Available from: https://www.researchgate.net/post/What_are_the_indicators_to_measure_agriculture_sustainability_in_farm_level [accessed Mar 3, 2016].

Conference Paper METODOLOGIA DE AVALIAÇÃO DAS POLÍTICAS DE DIVERSIFICAÇÃO EM ...

Article THE THEORETICAL APPROACH OF LIVELIHOODS DIVERSIFICATION

Dear Bijesh,

There seem to be two parts to your question: (i) what are useful indicators to quantify  sustainability aspects at the farm level, and (ii) how to combine results of these indicators into one statistic for a follow-up regression analysis to explain the variance among farms in their sustainability assessment.

A pragmatic approach towards quantifying sustainability would be to start from the three dimensions of sustainability (economic, social, and ecological) and work towards an integrated measure. The assessment then involves identifying meaningful sustainability attributes and finding a single metric of welfare that would allow combining them into an aggregate sustainability measure. The multi-attribute utility (MAU) approach is often used for this purpose. The additive aggregate utility function is commonly assumed because of its simplicity.

Thus, the construction of the aggregate sustainability measure involves identification of relevant sustainability attributes, transformation of individual attribute measures into commensurable units, development of a set of weights indicating relative impact of each attribute, and aggregation of the individual attributes into the aggregate measure.

The methodology developed in Sydorovych and Wossink, Agricultural Systems (2008) concentrates on the elicitation of relative weights of individual attributes for each of the three dimensions of sustainability.

One you have the aggregate measure for each of the three dimensions of sustainability you can run your follow up regression. This will be a Tobit model regression rather than a logit/probit model. 

[You could also incorporate the 3  separate aggregate measures into one overall sustainability index. It will be difficult to find weights for this and a non-parametric approach would likely be preferable (DEA)]  See attacahed paper by De Koeijer et al. (2002).

Thank you everyone for the answer and help. They are really helpful. Much appreciated.

Please, see this interesting framework:

http://mesmis.gira.org.mx/en/static/mesmis_framework

Dear Bijesh, you may also consider to use landscape metrics which may be calculated with the use of GIS, see for instance http://www.wildflower.org/lrp_client_docs/docs_edgworkshops/Webinar%202:%20Site%20Preparation/botequiha_2002_applying%20landscape%20eclogical%20concepts%20and%20metrics%20in%20sustainable%20landscape%20planning.pdf 

Good luck, Zdenek