1
The solution of a task №1. «Invent Yourself: Good guesses»
Team «12FM» Novosibirsk
Slide No.1. Hello, my name is Vladimir Iovets. Irepresentteam 12FMfromNovosibirsk. Let me introduce the solution to the task «Invent Yourself: Good guesses».
Slide No.2.Statement of the task presented in the slide.
Slide No.3. Atthebeginningoftheworkwehaveputforwardahypothesis:one of the most importantquantityfor Siberians is temperature. Todeterminehowpeoplefeeltemperature.
Slide No.4.Purpose: toresearchtheprocessofidentifyingthetemperaturebypeopleofdifferentages.
Tasks of the research
- To study the theoretical justification of the phenomenon under study;
- Select a parameter for the study;
- To conduct a survey of the inhabitants of Novosibirsk;
- To measure the selected parameter in the most accurate way;
- To compare the received statistical data with the measured value;
- Give an explanation of the results of the experiment.
Slide No.5.Let us consider the basic concepts underlying the solution of this problem.
Groups have outstanding collective intelligence and are capable of producing solutions that are much more insightful than the conclusions of the smartest of its participants.
To the crowd was wise, it is absolutely not required that leading roles in it played exclusively smart people. Even if the crowd does not include professionals, the group is still able to come to a wise collective decision.
An example of using the search for the best solution can be the Google search engine, which is capable in a matter of seconds to scan 106 Web pages and find one with the necessary information.
Another example of using the wisdom of the crowd is the mechanism of the stock market.
Slide No.6. Collective mind can be applied to solve a huge number of tasks that can be divided into three areas:
- Cognitive (problems that have or will have specific solutions, for example, with which score will this match win FC "Siberia");
- Coordination (which requires coordinating the behavior of an individual member of the group with the behavior of the others, considering that these others are trying to act in their own interests, for example, how to drive safely with heavy traffic);
- interaction (for example, the problem of environmental pollution).
In a reasonable group, participants are not encouraged to change their proposals to reach an acceptable solution for everyone. Instead, they gather together all the opinions and deduce from them the average collective judgments, which demonstrate not how a member of the group thinks, but actually what they all think together. Thus, the best way for a group to become reasonable is to allow each of its participants to think and act as independently as possible.
Slide No.7. The possibilities of collective intelligence were demonstrated in a series of experiments conducted by American sociologists and psychologists between the 1920s and the mid-1950s.
Hazel Knight (Columbia University Professor) in the 1920s asked students of her course to guess the air temperature in the room and then deduced the arithmetic mean of these assumptions. Then the tasks became more complicated, for example, it was necessary to solve the charade, the puzzle, the quest, and so on.
Sociologist Keith H. Gordon, in conducting similar experiments, determined the following accuracy parameters:
- when determining the weight of different objects, 94%;
- determination of the size of the canister, 94.5%;
- determination of the number of sweets in the jar with jelly beans, 97.6%.
The conclusions of the experiments were formulated by J. Teynor, professor of economics, they are presented on the slide.
Thus, we can distinguish four conditions that characterize the wise crowd:
1. diversity of opinions;
2. Independence of participants;
3. Decentralization;
4. Aggregation.
If all the listed conditions are met in the group, its general "judgment" will be very accurate with a high degree of probability.
Slide No.8. We will verify this statement using mathematical logic: when assessing the probability of occurrence of an event by a large number of people, we will find their "common" answer, in which the errors of the participants mutually exclude each other. Thus, in the experimental part of the study we learn the opinions of a large number of people of different ages and find their "common" answer. To do this, let's ask the residents of Novosibirsk the question "What air temperature do you feel?"
Slide No.9. All the answers received from passers-by were recorded in the tables. Thus in table 1 all received answers were recorded; In the second, the answers were distributed in two categories: junior schoolchildren, adult survey participants; In the third table there was a division into answers that relied on the sensation and answers, in which respondents remembered the seen or heard forecast from the mass media. Fragments of the tables are presented on the slide.
The true temperature was measured with an electronic thermometer.
Slide No.10. Imagine the results of the experiment in the form of a summary table. We draw the attention of those present to the results: the highest percentage of coincidence with the real temperature is obtained by interviewing independent adult respondents who rely on their feelings. The greatest error was shown by the survey of schoolchildren, which, in our opinion, is explained by a lack of life experience.
A small percentage of coincidences of the named values with reference to the weather forecast can be explained by the fact that in media with a mass of communication, the temperature interval characterizes the weather as a whole, but does not characterize it at any given time.
Conclusion on the experiment: with the variety of opinions of independent participants it is possible to obtain a sufficiently accurate value of the measured value.
Slide No.11. In the next experiment, respondents were asked to determine the temperature of the liquid placed in the calorimeter. The purpose of the experiment: using a large number of opinions of independent participants in the experiment to obtain a reliable value of the water temperature readings in the calorimeter.
The survey involved 28 teachers and 34 schoolchildren of 6-7 grades. The results of the measurements were recorded in tables, the fragments of which are presented on the slide. The answers of adult respondents (teachers) had an interval from 28 to 350С, for children the range of answers was wider: from 2 to 300С.
The result of processing the obtained values is presented in Table 3. We draw your attention to the average temperature value obtained among all the respondents, it is 76%, which is much lower than that obtained by us on a street poll (about 85%). In our opinion, this result can be explained by a 50% decrease in the number of respondents, on the one hand, and the ratio of the number of respondents by age, on the other.
If from the general answers to exclude sometimes ridiculous responses of children, the percentage of "guessing" increases to 97%.
Thus, by studying the opinions of independent respondents, it is possible to obtain reliable temperature information without the use of measuring instruments.
Slide No.12. Let's try to consider the situation when among a group of respondents there will be those who report deliberately false information to the rest of those present.
To do this, we will ask some members of the team of teachers (students) of our gymnasium to report a temperature value different from that measured at 80С, distorting the result by 75%. We ask those present to determine the temperature of the liquid in the calorimeter and call its meaning aloud. The results of the measurements are recorded in the table.
In the first group there were 34 teacher-respondents, of whom 6 misled the others. In the second - students of the 7th grade, in the number of 28 people, among which 5 leaders reported incorrect data.
Slide No.13.Excluding false testimony, let's summarize the experiment. On the slide, it is presented in the form of a table.
Analysis of the data obtained showed that if we exclude the property of independence of members of a group of people, then it ceases to be wise. Schoolchildren are more likely to influence, especially if they are influenced by those who are recognized as leaders in this group.
People of middle age in making a decision are based on their own feelings, comparing them with the opinions of others. At the same time, some "correct" their decision, especially if the results differ slightly.
Thus, if the opinions of the participants in the crowd become dependent on the opinions of others, then the solution of the problem does not become the best.
Slide No.14. Conclusions on the work.
1. In solving this problem, we selected a parameter important for Siberian inhabitants - temperature.
2. During the survey of residents of Novosibirsk, more than 100 independent opinions were received on the determination of the ambient air temperature by sensations. The obtained average value of the temperature by 85% reflects the value measured by the electron thermometer.
3. With a decrease in the number of respondents participating in the task, the average value obtained during the experiment differs from the true one by 24%.
4. If during the experiment to exclude the independence of opinions, then the error in solving the problem will increase.
Thus, in order for the crowd to show wisdom in the solution of the task assigned to it, only on condition that it consisted of a large number of independent participants and was decentralized.
Slide No.15. The slide shows the literature used in the course of solving the problem. Thank you for your attention, I am ready to listen to your questions.