Back to search

National Technical University of Athens (NTUA)

Ολοκληρωμένη μεθοδολογία και πληροφοριακό σύστημα παραγωγής πρoβλέψεων για τις δράσεις της ψηφιακής στρατηγικής στην Ελλάδα

Abstract

dc:description

The goal of the present doctoral thesis is the forecast of the success of policy implementation strategies (PIS) that has been announced in the frames of Digital Strategy in Greece. These programs mainly include Actions of the Information Society and the Special Secretariat of Digital Planning. For this scope, the Actions were organized as forecasting problems for which we used judgmental forecasting methods, while we also implemented an information system for judgmental forecasts.Important issues relating to public policy, judgmental forecasting methods and forecasting software were examined through literature review, which aimed at fully understanding them and then shaping the methodology proposed in this thesis as a research approach to the problem under study.The Actions studied in the present thesis took place either for the first time, or as re-introduction of the same Action after a small number of repetitions. Consequently, past data are few and for small depth of time, and thus statistical forecasting methods cannot be applied in the present case. For this reason, we chose to use judgmental forecasting methods in order to forecast their success.If a PIS, namely Action, is selected, then more complex and costly assessment tools could be applied, such as cost-benefit analysis. The forecasts are not intended to replace the more specialized techniques, but precede them, saving time and money.The main method of judgmental forecasting used in the experiments of the present thesis was the method of Structured Analogies. This method was created initially for the forecast of decisions in problems where opponent interests exists. In the research, the method of Structured Analogies was used lightly differentiated so that it suits in the present case, while it was also combined with other forecasting techniques. The goal was to examine if satisfactory forecasts can be produced with this method for the Digital Strategy as well and also to investigate whether the variations of the method can give better results than the method as such. The Delphi method that was used in our survey too, also gave good results and even better when it was used in combination with the method of Structured Analogies. Nevertheless, it was the use of Structured Analogies through interaction groups that seems to make better use of the capabilities of the method and gave the best results in the forecasts. The personal interaction on group meetings led to the satisfaction of the participants. It seems that the greater the diversity in the composition, in terms of the origin of the experts in the interaction groups, the more accurate the resulting forecasts. The present research thus shows that it makes sense to rely on group rather than individual experience.The thesis includes a series of experiments with different carrying out means, various methods of judgmental forecasts and different categories of participants. In the first two experiments the participants were pre-graduate students of the School of Electrical and Computer Engineering of NTUA. The experiments were carried out in the classroom and in the examination room of the course “Forecasting Techniques”. The students were separated in teams depending on the method that would use, after that questionnaires about the Actions were given to them and they were asked to give forecasts for each Action. In the first experiment the forecasts were individual while in the second both individual and group.However, because in the case of judgmental forecasts the forecasts are considered to be more important when they come from experts, two experiments with experts were also organized. The first experiment was carried out electronically. The experts were informed about the experiment via e-mail, the questionnaires were sent via e-mail, the same for the answers of experts. Two experiments followed where in each one of them four experts were assembled, out of the set of experts that participated in the first experiment. The experts produced group forecasts and afterwards discussion followed in the form of interview about the experiments that they took part, where the experts also formulated their opinions and proposals.The above research was summarized in corresponding web site with informative content. A web information system for the production of judgmental forecasts was also developed, incorporating different methods of judgmental forecasting. This system constitutes an innovative application, since nothing equivalent has been created for commercial or academic purposes, except for the Delphi method. The system also includes the well-known and reliable method of statistical forecasts Theta, which can be used in cases where past data exist. Various conclusions came out from these experiments. The most important are that group forecasts give better results than individual ones, that the structured methods of judgmental forecasts, and specifically Structured Analogies, have better performance than the simplest methods and that the combination of selected methods and experts can also perform better. Moreover, the system that we propose, according to relative evaluation, affects positively and supports the whole forecasting process. In summary, we can say that the contribution of the present thesis is found firstly in that for the first time is attempted to forecast the success of the Actions of Digital Strategy in Greece and in particular with judgmental forecasting methods, both individual and group ones. The use of Structured Analogies, with differentiated form, in order to suit in the present case but also its combination with other methods is also something that is attempted for the first time. Moreover, the web information system for the production of judgmental forecasts that we implemented is something innovative not only for the Digital Strategy but also in general.

Degree

thesis:*
Grantor dc:publisher
National Technical University of Athens (NTUA)
Year dc:date
2014

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Litsa, Akrivi
  • Λίτσα, Ακριβή

Subjects

dc:subject × 23

Rights

dc:rights
Statement dc:rights
  • BY
Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/39492

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Litsa, Akrivi; Λίτσα, Ακριβή. Ολοκληρωμένη μεθοδολογία και πληροφοριακό σύστημα παραγωγής πρoβλέψεων για τις δράσεις της ψηφιακής στρατηγικής στην Ελλάδα. National Technical University of Athens (NTUA), 2014. http://hdl.handle.net/10442/hedi/39492