A new scientific orientation
After the confrontation with the work of Coleman, it was clear to me in which direction I wanted to continue with respect to my academic life. Of course, I had to continue my teaching task but besides that I planned to study as much statistics and mathematical modelling as I could. To do so I first looked whether it was possible to study mathematics and statistics next to my work but that was not possible. Then I decided to convince my colleague, Marius de Pijper, to join me in studying these topics.
For unclear reasons Marius had chosen to study Sociology. In general he did not understand the processes that were studied in Sociology, but he understood much better mathematical and statistical issues and later he liked very much programming. He was my ideal companion in my search for a more fundamental approach to social science research. Our cooperation went on for more than 25 years.
For unclear reasons Marius had chosen to study Sociology. In general he did not understand the processes that were studied in Sociology, but he understood much better mathematical and statistical issues and later he liked very much programming. He was my ideal companion in my search for a more fundamental approach to social science research. Our cooperation went on for more than 25 years.
Learning mathematical modelling
First of all we studied the book of James Coleman “Introduction to Mathematical Sociology”. This book forced us to repeat what we once had learned about differential and integral calculus because Coleman formulated mathematical models for the moves of people from one situation to another. For example people can change from member to non-member of an organization and the other way around.
Besides this interesting book, we also studied “Theory Construction” by Hubert Blalock. He tried to do the same as Coleman but now for variables like opinions which are in principle continuous, from very negative to very positive.
For me it was extremely interesting because these books showed that one could formulate processes described in the Sociology in mathematical form. Then one could derive new conclusions based on these models which were not possible in the verbal form and it was also clear how one could test such theories.
Marius liked this activity because he could concentrate on the models and the possible derivations and tests without worrying about the link with the social reality. That part he left to me.
First of all we studied the book of James Coleman “Introduction to Mathematical Sociology”. This book forced us to repeat what we once had learned about differential and integral calculus because Coleman formulated mathematical models for the moves of people from one situation to another. For example people can change from member to non-member of an organization and the other way around.
Besides this interesting book, we also studied “Theory Construction” by Hubert Blalock. He tried to do the same as Coleman but now for variables like opinions which are in principle continuous, from very negative to very positive.
For me it was extremely interesting because these books showed that one could formulate processes described in the Sociology in mathematical form. Then one could derive new conclusions based on these models which were not possible in the verbal form and it was also clear how one could test such theories.
Marius liked this activity because he could concentrate on the models and the possible derivations and tests without worrying about the link with the social reality. That part he left to me.
A suggestion for present researchers
Looking again at these books for this story, I realized that we normally study data at one point in time, ignoring that the processes occur through time. This assumes that we study processes that have reached a stable situation otherwise all measures may be biased. Studying the obtained values of the coefficients, one can determine if this condition is fulfilled (see the picture). This is hardly ever done. I really recommend researchers to look at the small book of Blalock. It is still very valuable. The models of Coleman have been substituted by what is called “loglinear models”.
Looking again at these books for this story, I realized that we normally study data at one point in time, ignoring that the processes occur through time. This assumes that we study processes that have reached a stable situation otherwise all measures may be biased. Studying the obtained values of the coefficients, one can determine if this condition is fulfilled (see the picture). This is hardly ever done. I really recommend researchers to look at the small book of Blalock. It is still very valuable. The models of Coleman have been substituted by what is called “loglinear models”.
The Summer school of the ECPR in Colchester
At some moment in that year I saw an announcement that in Colchester (Essex) a summer school in research methods was organized by the European Consortium for Political Research. This was an extra opportunity to learn more. I applied and was again accepted. In this way a long lasting relationship with the Summer school started as we will see.
The course of the Summer school was quite different from what I had studied during the last year. The emphasis was on basic statistics and statistical procedures to estimate and test the so called regression models. This was an approach that was very popular in Econometrics to test economic theories. The course was an interesting extension of my knowledge. I concluded that the course was very good but too short to become completely familiar with this topic. So I decided to spend the next year with Marius and another assistant on this topic.
Let me concentrate here on a phenomenon that is typical for Europe. The course took place at the Campus of the University of Essex which is at 5 kilometers from Colchester. We lived in high towers with in each floor six rooms and a kitchen and two showers. The food was included in the price for the course and was served in the restaurant of the campus.
All who have been for some time in England and were served British food without much choice will admit that this food is not in line with all peoples´ taste. Especially the French did not like the food at all. Therefore they went to the town every day to buy their own food and prepared their own meals. Later they required to get their money back for the meals. But another issue was that they missed the courses because they had to make these purchases and to cook of course. So they were also complaining that they had no time for the courses. Within a short time the ECPR decided to organize also a Summer school in France for the French speaking community.
To be honest I am not so sure whether the problem was indeed the food or their limited understanding of English. They also organized their own parties in the evenings while the people from all other countries came together every evening to have their drinks and made very nice parties.
At some moment in that year I saw an announcement that in Colchester (Essex) a summer school in research methods was organized by the European Consortium for Political Research. This was an extra opportunity to learn more. I applied and was again accepted. In this way a long lasting relationship with the Summer school started as we will see.
The course of the Summer school was quite different from what I had studied during the last year. The emphasis was on basic statistics and statistical procedures to estimate and test the so called regression models. This was an approach that was very popular in Econometrics to test economic theories. The course was an interesting extension of my knowledge. I concluded that the course was very good but too short to become completely familiar with this topic. So I decided to spend the next year with Marius and another assistant on this topic.
Let me concentrate here on a phenomenon that is typical for Europe. The course took place at the Campus of the University of Essex which is at 5 kilometers from Colchester. We lived in high towers with in each floor six rooms and a kitchen and two showers. The food was included in the price for the course and was served in the restaurant of the campus.
All who have been for some time in England and were served British food without much choice will admit that this food is not in line with all peoples´ taste. Especially the French did not like the food at all. Therefore they went to the town every day to buy their own food and prepared their own meals. Later they required to get their money back for the meals. But another issue was that they missed the courses because they had to make these purchases and to cook of course. So they were also complaining that they had no time for the courses. Within a short time the ECPR decided to organize also a Summer school in France for the French speaking community.
To be honest I am not so sure whether the problem was indeed the food or their limited understanding of English. They also organized their own parties in the evenings while the people from all other countries came together every evening to have their drinks and made very nice parties.
Now we have really to learn statistics
Being back again my proposal was to study mathematical statistics and estimation and testing of the model that was taught in Colchester. Now we had a group of three because we had an extra assistant, Jaap de Visser. We started to study the book of Hoel on “Introduction to Mathematical Statistics” and we did all exercises. If one of us could not solve a problem, the other could and so we learned statistics.
After finishing the book on statistics we moved to study the regression model and used our new statistical knowledge to check the proof under what conditions the OLS estimator is the best unbiased estimator for that model. That proof became a large part of the master thesis in ”sociology” of Jaap de Visser.
Still hungry for more
This year was very successful and we all three worked at it with pleasure but I was still not satisfied. I wanted to learn more and see more of the world. My plan was to go to the USA for 6 months. I wanted to participate during 2 months in the Summer school in An Arbor (Michigan) and study three months at a university in the USA. I had heart that Coleman accepted only students with a mathematics degree which I did not have. So I decided to write to Hubert Blalock if he was willing to accept me as a student for three months. He wrote quickly back that he would welcome me at his university which was at that time in Seattle, completely at the other side of the USA. So this was the plan for our next trip in March 1973.
Being back again my proposal was to study mathematical statistics and estimation and testing of the model that was taught in Colchester. Now we had a group of three because we had an extra assistant, Jaap de Visser. We started to study the book of Hoel on “Introduction to Mathematical Statistics” and we did all exercises. If one of us could not solve a problem, the other could and so we learned statistics.
After finishing the book on statistics we moved to study the regression model and used our new statistical knowledge to check the proof under what conditions the OLS estimator is the best unbiased estimator for that model. That proof became a large part of the master thesis in ”sociology” of Jaap de Visser.
Still hungry for more
This year was very successful and we all three worked at it with pleasure but I was still not satisfied. I wanted to learn more and see more of the world. My plan was to go to the USA for 6 months. I wanted to participate during 2 months in the Summer school in An Arbor (Michigan) and study three months at a university in the USA. I had heart that Coleman accepted only students with a mathematics degree which I did not have. So I decided to write to Hubert Blalock if he was willing to accept me as a student for three months. He wrote quickly back that he would welcome me at his university which was at that time in Seattle, completely at the other side of the USA. So this was the plan for our next trip in March 1973.