Irmtrauds research choice
Back from the course in Essex Willem applied for a travel grant for the USA at the Fulbright Foundation and a visitors grant for 6 months at the National Science Foundation (ZWO). We planned to go there from March till September 1973.
In the fall of 1972 Willem followed a course of the Mathematical Centre in Amsterdam on mathematical decision theory. He saw this more as a nice exercise in a different application of mathematics and statistics. This course turned out for different reasons to be more important for our later work than expected. First of all, he learned the basic concepts of decision theory and its procedures to select optimal solutions for different decision problems. Secondly, he was confronted with some basic operations in matrix algebra. Also this knowledge turned out to be very useful within a short time.
After Irmtraud had finished her degree in French she had to look for a new activity. We were speaking about different possibilities. At some point Willem suggested with his recent course in mind: “In your thesis you write about deliberations between Margret of Austria and the emperor Charles V about the different possible strategies for the peace discussions. Do they also mention possible consequences, the likelihood that they occur and the value of these different outcomes? If that is so, we could also investigate if they had chosen the optimal strategy using a mathematical model.”
In the fall of 1972 Willem followed a course of the Mathematical Centre in Amsterdam on mathematical decision theory. He saw this more as a nice exercise in a different application of mathematics and statistics. This course turned out for different reasons to be more important for our later work than expected. First of all, he learned the basic concepts of decision theory and its procedures to select optimal solutions for different decision problems. Secondly, he was confronted with some basic operations in matrix algebra. Also this knowledge turned out to be very useful within a short time.
After Irmtraud had finished her degree in French she had to look for a new activity. We were speaking about different possibilities. At some point Willem suggested with his recent course in mind: “In your thesis you write about deliberations between Margret of Austria and the emperor Charles V about the different possible strategies for the peace discussions. Do they also mention possible consequences, the likelihood that they occur and the value of these different outcomes? If that is so, we could also investigate if they had chosen the optimal strategy using a mathematical model.”
The first decision study on data of the 16th century
Irmtraud was immediately excited about this idea and looked at the texts and came back after some time with the result: “They use indeed such statements. Would that mean that we can check if they have chosen the best strategy?” Willem answered: “That is indeed the case but then we have to transform the verbal expressions you have found in numerical values. But that I can do using well known scaling techniques.” Willem recommended Irmtraud to collect the different statements and read some literature about decision making, especially the book of Fishburn.
She collected the verbal statements of the different strategies which these decision makers in the 16th century considered and their idea about the possible consequences, chances of success and the value statements about the different possible outcomes. Willem also was curious about the possibilities to apply the Expected Utility model on data of the 16th century. During a class on measurement he asked the students to evaluate the different statements about probabilities and utilities using the Thurstone scaling procedures.
Irmtraud was immediately excited about this idea and looked at the texts and came back after some time with the result: “They use indeed such statements. Would that mean that we can check if they have chosen the best strategy?” Willem answered: “That is indeed the case but then we have to transform the verbal expressions you have found in numerical values. But that I can do using well known scaling techniques.” Willem recommended Irmtraud to collect the different statements and read some literature about decision making, especially the book of Fishburn.
She collected the verbal statements of the different strategies which these decision makers in the 16th century considered and their idea about the possible consequences, chances of success and the value statements about the different possible outcomes. Willem also was curious about the possibilities to apply the Expected Utility model on data of the 16th century. During a class on measurement he asked the students to evaluate the different statements about probabilities and utilities using the Thurstone scaling procedures.
If Margret and Charles V made an optimal choice they should have chosen the strategy for which the "expected utility" (EU) would be maximal. The expected utility can be computed for each strategy as the sum of the products of the probabilities and the values. For example for strategy S1 the EU (S1)=0,1x7,31 +0,9x-9,55= -7,864 which was clearly a negative result. For strategy 2 the EU(S2)= -6.663 and for strategy 3 the EU(S3)= -0.35. These results suggest that strategy 3 had the best expected unitility and therefore should be prefered. Margret and Charles V indeed had chosen this strategy in the negotiations. We were very pleased with the result and Irmtraud was thinking that this type of research could be continued because there are many of such documents about decisions that could be studied.
We also wanted to publish these results but first we had to take care that we had enough money for our trip to the USA. From the Fulbright Foundation we rather quickly received a message that they would cover our travel expenses. However by the end of the year there was still no reaction of the Science Foundation. So, when Willem was offered an opportunity to do an evaluation study with respect to renovation in three areas in Amsterdam, we agreed to do this research. Willem would develop the questionnaire and Irmtraud would manage the data collection.
The first “information and choice questionnaire”
In areas where renovation of flats is needed the population has to be asked whether they agree with the renovation plan. Normally a show flat is created and the people in that neighborhood are asked to evaluate the plan for the new flats. Of course such show flats look very nice, well decorated and are in general larger than the flats in which the people live at present. The consequence is that in general the majority of the interviewed persons are positive about the renovation and the work can start.
Thinking about this procedure and having just done a course in decision making, Willem thought that this was not a fair procedure because the consequences of the increased costs for rent were not at all taken into account. Because the renovation took place in areas where people with low income were living and the cost of the rent was rather low, the increase in rent could have serious consequences. We informed about the rent of the new flats and the possibilities of subsidies if the costs would become too high.
In the evaluation questionnaire we not only asked the people whether they liked the new flats but also what their income was, how much they paid for rent now and gave them information about the rent in the new flat which they wanted. On the basis of this information we could determine how much less income would remain and if they could get any subsidy. Only after the evaluation of the flat and the financial consequences, we asked the people whether they were in agreement with the renovation. Irmtraud looked for interviewers and organized the field work.
Managing this survey was a new experience for her. Some inhabitants were so angry about the renovation plans that they did not want to participate in the survey because they had put a lot of money in the flat to make it as nice as possible according to their taste. Irmtraud went to them and they showed what they had done. She convinced them that they could also say that they were against the renovation. Some older inhabitants were afraid of the young long haired student interviewers.Then Irmtraud went with them. Nearly always she could convince these people that the interviewer was not dangerous, that he could be trusted, and they cooperated. For Turkish speaking inhabitants she had to find a translator. In the end she organized a response rate of 99%.
We applied this procedure in three areas needing for each study two weeks. Using this “information and choice questionnaire” in all three areas the population in majority was against the plan. The reason was quite clear: compared with the present situation the costs of the rent became so high that many would need financial assistance of the state while they did not need it at this moment. These results meant that the renovation could not be started in these areas.
We also wanted to publish these results but first we had to take care that we had enough money for our trip to the USA. From the Fulbright Foundation we rather quickly received a message that they would cover our travel expenses. However by the end of the year there was still no reaction of the Science Foundation. So, when Willem was offered an opportunity to do an evaluation study with respect to renovation in three areas in Amsterdam, we agreed to do this research. Willem would develop the questionnaire and Irmtraud would manage the data collection.
The first “information and choice questionnaire”
In areas where renovation of flats is needed the population has to be asked whether they agree with the renovation plan. Normally a show flat is created and the people in that neighborhood are asked to evaluate the plan for the new flats. Of course such show flats look very nice, well decorated and are in general larger than the flats in which the people live at present. The consequence is that in general the majority of the interviewed persons are positive about the renovation and the work can start.
Thinking about this procedure and having just done a course in decision making, Willem thought that this was not a fair procedure because the consequences of the increased costs for rent were not at all taken into account. Because the renovation took place in areas where people with low income were living and the cost of the rent was rather low, the increase in rent could have serious consequences. We informed about the rent of the new flats and the possibilities of subsidies if the costs would become too high.
In the evaluation questionnaire we not only asked the people whether they liked the new flats but also what their income was, how much they paid for rent now and gave them information about the rent in the new flat which they wanted. On the basis of this information we could determine how much less income would remain and if they could get any subsidy. Only after the evaluation of the flat and the financial consequences, we asked the people whether they were in agreement with the renovation. Irmtraud looked for interviewers and organized the field work.
Managing this survey was a new experience for her. Some inhabitants were so angry about the renovation plans that they did not want to participate in the survey because they had put a lot of money in the flat to make it as nice as possible according to their taste. Irmtraud went to them and they showed what they had done. She convinced them that they could also say that they were against the renovation. Some older inhabitants were afraid of the young long haired student interviewers.Then Irmtraud went with them. Nearly always she could convince these people that the interviewer was not dangerous, that he could be trusted, and they cooperated. For Turkish speaking inhabitants she had to find a translator. In the end she organized a response rate of 99%.
We applied this procedure in three areas needing for each study two weeks. Using this “information and choice questionnaire” in all three areas the population in majority was against the plan. The reason was quite clear: compared with the present situation the costs of the rent became so high that many would need financial assistance of the state while they did not need it at this moment. These results meant that the renovation could not be started in these areas.
The political impact
Based on these results and similar results in other renovation areas a new law was decided in the parliament that determined that the rents could not be increased so abruptly by the renovation but should slowly be increased through the years.
The last report about this research was delivered on Thursday March 1 thanks to the help of Marius the Pijper who did the analysis on the existing computer while Willem wrote the final reports. This was two days before we would leave for the USA. Now we were sure that we had enough money for the stay in the USA. We also had time to give a party for the people who helped us with this project on March 2. This was also the day that we received an official letter of the University of Washington that our application to stay three months at their university was rejected. What now?
Based on these results and similar results in other renovation areas a new law was decided in the parliament that determined that the rents could not be increased so abruptly by the renovation but should slowly be increased through the years.
The last report about this research was delivered on Thursday March 1 thanks to the help of Marius the Pijper who did the analysis on the existing computer while Willem wrote the final reports. This was two days before we would leave for the USA. Now we were sure that we had enough money for the stay in the USA. We also had time to give a party for the people who helped us with this project on March 2. This was also the day that we received an official letter of the University of Washington that our application to stay three months at their university was rejected. What now?