Sunday, June 18, 2006

Summary (part 1)

This is summary of everything happened in last few weeks:

Finally, I presented and defended my thesis on Thursday June 1st 2006, in front of my supervisor and one external examiner. It was take place for more or less one hour, started at 12.30 in the afternoon. After presentation for about 15 minutes (from 12 minutes that allocated), I received a lot of questions from the external examiner and two or three questions from my supervisor. Several important points that I had chance to make note on the questions I presented below:
  1. The main problem is on the understanding the statistical model used. In my study case, interpreting and presenting the results of logit regression model in a proper ways.
  2. What is the practical implication of my study?
  3. Expectedly, there is strong correlation between population density variable and regional variable, especially Java island variable. This issue was not yet clearly anticipated and controlled.
  4. Regarding unit of analysis: observably, there are more than one individual in one household that being analyzed. Thus, there is a big possibility of correlation among household member which not yet clearly anticipated and controlled in my regression model.

I had answered the above questions/problems as:

  1. Given the study that I did in my thesis, it was only an initial for further study in farm household theory. To continue improving the method and techniques, I need to learn more about another regression model or improvement of given model used by past studies.
  2. Regarding practical implication, I believe that my study have a general implication on the development of rural (farm) household potential in Indonesia. It is expected that by understanding the behavior of labor market in rural area in Indonesia will gives us hints for promoting any policies to improve economic condition in those area. Specifically, this study was not yet supplied or analyzed particular policies in Indonesia.
  3. I already suspected that the correlation between population density and Java island variable was exists. However, I still apply the two variables to represent impact of supply of labor (population density) and regional differences (by region/island). Due to time constraint, it is expected that those condition could be improve in future studies.
  4. About unit of analysis, my observation with the data found that there is a condition where each farm household have numerous member where they have activities not only in farm/agriculture but also non-farm/non-agriculture. Previous studies mostly used household head or farm operator and their spouse as unit of analysis, ignoring other household member. However, for Indonesia case that would not be able to capture the real situation of labor allocation within household in Indonesia. That is my main argument for using individual level as unit of analysis. Regarding the correlation among household members, definitely I need to control them. However, I could not found any literature with clear explanation of such condition particularly for topic off-farm work participation. Further improvement need to be done in the future.

Despite of all questions arise during the defense, my own note regarding my thesis is as follow:

  1. The data that I utilized, IFLS, has not yet being used optimally. The data contains plentiful information, but it is quite difficult to understand and to prepare the raw data to be able to use directly for analysis. Another self-study or training might be very beneficial to explore the benefit of IFLS data.
  2. The farm household model itself also still needs to explore further. From what I have found so far from various journals, the model has much potential to be used for different purposes. So, it is really beneficial to master this model indeed.


After the defense, I got a call from my best friend Paksi and he congratz me in here. Thank you very much, Bos! Looking forward to see you very soon. Well, that is all for now…

No comments: