Which is the best statistical tool to conduct a comparative study of technological advancement in agriculture in two different districts? The study is based on block level secondary sources of data for technological equipments like number of tractors, thrashers, tiller, harvester, pump sets and thrower per thousand hectare of net sown area. I want to make a comparision between the two district about which distrcit is technologically more advance. Which tool can help me in figuring out this? Suggestions are welcome...
Dear Faiyaz,
In order to capture the impact of technology in agriculture it is good to have a dependent variable or an output, like farm income or kilograms of crop produced.
If you have such information you can apply two linear regressions (one for each district) using income or crop production as the dependent variable. The results will show you if technology has a significant impact on your dependent variable and if that impact is positive or negative.
Another option is to compare the efficiency between both districts, which can be done using Stochastic Frontier Analysis (SFA) or Data Envelopment Analysis (DEA). Again, your output should be something like farm income or crop production and your input variables would be the technological information you have.
These are my suggestions, hope it will help you.
The simplest method will be to use t-test for comparison of two means. This can be done in Excel or any other data analysis software such as STATA, SPSS, R, etc. However, you can calculate composite index of technological development for both districts and then compare the index numbers. You may consult my M.Sc. thesis "A Spatio-Temporal Analysis of Agricultural Development in Himachal Pradesh" for the latter case.
Hope it is helpful.
Thank you for your valuable suggestions @Gabriel Herrera
I have crop production data as output value and I'll incorporate the proceedure that u explained. Thank you for helping me out.
Thank you @Najibullah sir
One thing I also want to know is can f-taste be used for the study?
and
Where can I get your M.Sc thesis?
F test is used when you compare several means. If you are doing so, it can be used otherwise not. I guess you compare one variable for two district, which require t test.
You can get my master thesis from my research gate page or google its title mentioned above.
Yes sir, I'm comparing as much as 25 variables of two districts.
Thank you..
Thank you for your suggestion @ Haji Saediman sir. I just want campare the variables in two districts and than I want to state which one is more developed or sufficient in terms of that variable. So, how could I came to conclude about the comparision of more better and less better? What is the process to identify? I mean which value explains what in the t-taste?
T test is the most appropriate when comparing means from two groups regardless of the number of variables you have. In the event you have several groups, you may wish to consider ANOVA
Such comparisons are best done with t test with a normally distributed data and chi square when it is not. Overall means are best compared by looking at the data then choose the kind of test appropriately.
Use ANOVA for comparison among the different variables, then choose a post hoc (e.g. Duncan, LSD etc.) for mean separation. For comparison between the two locations, you should use student's independent sample t-test.
Good luck!
- Badges
- Science topic
- Similar topics
- Agriculture