Scientific Program > Topic 4 > Session 4H >
Presentation 4H1. Using Statistics as a Tool in Political Research: The Case of Electoral Behaviour

Theodore Chadjipadelis (Greece)


Presentation Abstract

The demand by the society for qualitatively controlled information absolved from "noises" which intentionally or without purpose are included, is especially obvious in a returning, from time to time discussion, about the control of the public measurements related to the Mass Media audience, political parties and persons, educational parameters, economical and social indexes and factors, etc.

In this paper we will refer more extensively in the problems of using statistical techniques in the political sciences.

Prior to electoral processes the civilians are bombarded with information, graphs, tables and findings of opinion polls which are referred in indexes, figures such as the expected power of parties participating in elections, the publicity of persons, evaluative classification of issues and politicians. After the elections the results are explained and analyzed with the use of advanced statistical techniques aiming to support the pre-electoral statements and attitudes of political parties.

By following and commenting the phases of observation of electoral behavior we will refer to some problems.

While publishing the results, many analyzers and journalists do not understand that the values that are counted for the figures (usually percentages or flows) are not accurate values but estimations. They do not understand that a confidence interval means that the real size (e.g. the percentage of a political party is found with exactly the same probability between the upper and lower limit, considering that it is found nearer to the center of the confidence interval. Yet, the possibilities of getting results about subpopulations of the electorate are limited as they depend on the design of the sample.

All political parties are interested in the analysis of behavior of subpopulations of people. However, if the sample can be designed so that to represent the general population, this does not mean that each part of it represents the subpopulation from which it happened to come from. The supporters of some party found in the sample are not a representative sample of the total supporters of the party, and the youth between 18 - 25 found in the sample are not necessarily a representative sample of all the youth, etc. The use of methods of analyzing two-way tables (such as correspondence analysis) can lead to errors and misunderstandings.

The development of technology and computers led to an outbreak of gathering information data and fed the need of the evaluation and use of the gathered data. Information as a property product and as a means of policy making replaced in a great degree the material products leading to a new form of concentrating power.

Statistics today, in the era of information outbreak can be generally defined as the "Science that is concerned with the gathering, evaluation and processing of information".


Download in Adobe Acrobat format (107 Kb).


Back to top

ICOTS-6, The Sixth International Conference on Teaching Statistics - International Program Committee (IPC) Website.
Copyright © 2001 by the IASE. All rights reserved. This information is subject to change without notice. This page was last modified on July 2, 2002.
For questions or comments, contact the Webmaster, Dagan Ben-Zvi.