An Evaluation on Modeling and Design of Political Researches


The critical point in a research is actually its modeling and design in a way capable of incorporating the mass who is the addressee of the subject researched. Throughout every phase of research, from general to specific, the factor “addressee-ness of the question” should be paid attention.

Stuart Sutherland, in his book “Irrationality”, defines the state of being more influenced by antecedent elements than alterior ones as “primacy error”. This finding plainly suggests how careful one should be during the phase of determining statements, which is an important component of research.

Difficulties that may arise in modeling and design of research…

As is known to all, there is a process commonly followed in works of research. This process starts with defining the problem of research. Then, the arrangement of statements is made, which can also be considered one of the preliminary works. Statements define what do we search, and through which means and techniques do we perform the search during implementation phase. Statements are also acceptances that facilitate wise use of research sources and limited time.

In order to answer the question of research, statements that enable us to inquire every kind of phenomenon that constitutes this question are expected to be built wisely, with effective question clauses present. At this point, oral and written background that we may roughly define as literature, that is, all preceding works the information gathered from those become our primary helper. Besides, regarding the subject of a research, particularly when making a social research; social dynamics, reactions of the community to a similarly antecedent phenomenon, and the way of interaction with this phenomenon should be counted among our primary information. Namely, in Sutherland’s words, ‘background elements’.

It is clear that a mistake made at the initial stage of a research, may affect the entire research. “Primacy error”, which is likely to be made at stage of determining statements and plotting of the research – and affects the entirety of research both content-wise and technique-wise -, may lead us to a conclusion far removed from the data we aim to find. This is especially the case when antecedent information regarding research subject is insufficient or we are not competent enough to evaulate it.

A design error may also lead to deviations at stage of determining inquiry headlines of the research. It is obvious that during work, we are going to be inclined to query the elements that we assume related to the phenomenon researched, and evaluate these together with the answers they will bring. However, we can make several “small” mistakes in the process of selecting these elements, which is one of important mistakes made during design phase. Provided that these mistakes are isolated to their own within a research, we might not face too dramatic results. Until when? Until when these elements, by the nature of research, create antagonist effects – at every turn – upon other factors that “should be actually included”.

Inaccurate elements involved in the process, starting from the phase of modeling and design – and even from the moment the idea of research has been shaped within researcher’s mind – , will continue to exist in an extremely hidden fashion, deeming themselves meaningful until the moment that they can become effective at occurence of illusion.

Virtually, the antagonist elements that should not exist in a research; may hinder the identification of research’s phenomenon, the ability to create consequences, keep respondents from focusing on the main phenomenon and come up with answers accordingly. Or worse, they “may not”. In this case, they may block both the identification of research’s phenomenon and the focusing on it with every element. In a circumstance like this, we may even fall under “illusory correlation”, which is another common mistake, made through irrationally combining the data we have in hand.

Yet, better to point out that there are ways to be rid of from both cases of mistake through research. By obviating the harms that come with previous data or wrong links, researches that enable the obtainment of data to form a ground to get real information about phenomenon, are made.

As a research phenomenon : Political elections…

The abovementioned mistakes of modeling and design are prevalent for almost every research. Exceptionally, in explanatory researches where researcher has not much antecedent information about phenomenon, the purpose is to define the main components of phenomenon, doing so independently from antecedent information. This is why, in this case, researcher holds his preventive reflex as heightened as possible towards the possibility that every element may create an antagonist effect.

Problems that I mentioned above can often be seen in political researches made at any sequence – that is, time period – in order to foresee the possible political preferences of community.

In this article, regarding political researches, it is good to suggest that we are not dealing with deliberate behavior that stems from intentional use of the manipulation effect resulting from public sharing of research results or researcher’s engagement with any political interest group or a party. It is always possible to evaluate this kind of mistakes – which are outside of research discipline and not methodological and technical, but criminal – in another article.

In political researches – particularly in pre-election researches – , it is a first-priority necessity that social provision of the next election should be defined, to which phenomenon of society’s past experiences it corresponds – when itself is evaulated as a phenomenon – should be known, and boundaries of the research should be assessed according to this sense. Besides, foreseeing the sequence that society is in and the political culture level – that is, the ability to unbundle elements of phenomenon –, is of critical importance in order to carry a reliable research.

As for misty spheres such as Turkey, where political development process is developed by all actors of politics in an accelerated fashion, researchers must make more qualified researches in this respect. Following plain handling of phenomenon, the direction of development that might be evaluated as antecedent input  and today’s dynamics  that enable correct determination of elements, should be thoroughly analyzed. In fact, the most correct way to approach for a researcher, is to genuinely evaluate his/her competence in him/herself and get support from theoretical research, which is considered remote from the practical.

A more effective way to overcome these difficulties,  – as I stated above – is to handle the phenomenon plainly. Thus, the research and the researcher can evade the inclusion of elements that may create highly antagonist effects, to some extent at least. At this point, even if the researcher may think he/she can get rid of these possible effects in the process of analysis – and even if he/she is actually capable of that – , he/she has no assurance that these effects won’t have any function during the conduct of research. Without underestimation, the possibility should be considered that these components, bearing the risk of deviating results to an abnormal state – and foreesen to be manageable later–, may show their effects starting from the addressee during the research. It is a theoretically proven acceptance that observation influences the event observed and the way it occurs. Therefore, the observer – the researcher, that is – should accept as a principle, to affect the observation as minimally as possible.

Particularly in political researches, during the observation – made independently from the instrument of research used – , work should be conducted clear of components that may cause the addressee to get away from phenomenon, or get misguided in a fashion that will have the addressee influenced by factors that should have no place in phenomenon’s way of realization.

As another mistake in political researches – in all sorts of research, actually –, having the basic analyses made by the addresse, assigning the addresse to provide a rational internal consistency with charges of question which form ground to analysis, will multiply the negative effects of the abovementioned “factors that should have no place”.

It should be re-stated that tendency of orthodox researchers – independent from technological modernity of the instrument they use – to lay the need for data required in analysis they are going to make on the addressee in a most rough way, is their most conspicuous and result-affecting issue especially in political researches.

Starting to work under the assumption that the addresse is rational and will keep his/her self-consistency unaffected from observation factors throughout the research, will cause the researcher to face a data set which he/she is going to have difficulty to understand. Yet, it would be better for a researcher to foresee the irrationality (not counting ideological preferences) and predilection to produce unorthodox data and adopt an instrument of plain design in a way to avert it.

As I stated above, in political researches or, more specifically, election researches; irrational and unorthodox tendencies that may trigger in the addressee with a small precipitation can be averted by counterbalancing them or through a concentration that will never provide a basis for it at all.

Providing the addressee with the information of what is being researched, having the answers focused on phenomenon’s core and basically purifying them of the analysis made through phenomenon’s subfactors which are the problems of the theoretician, will enable more clear results based on analysis of the data.

In Turkey, it has been seen in many researches made prior to Presidential election held in August 2014, that the effects of these modeling and design mistakes were experienced and estimations made my bany researchers fell far from realized values. In an example where the abovementioned effect of  “taking the addressee away from phenomenon” was characteristically cultivated; following a charge of questions in which the addressee was distracted from Presidential election and focused on the last local elections held; the estimated vote rates of candidates obtained by asking people’s opinions about Presidential election, are as follows: Recep Tayyip Erdogan 58% (reality: 51%, deviation: 7%), Ekmeleddin Ihsanoglu 30% (reality: 38%, deviation: 8%) and Selahattin Demirtaş 12% (reality: 10%, deviation: 2%).

Here, the researcher actually asked the addressee “You are familiar with the results of previous election. Based on that, how should the electors vote?”, instead of asking “Who are you going to vote for?”. Because the addressee is seeing the research as a whole and , rather than his/her own preferences, is seeking a way out for the irrationality of society and him/herself.

The result: Is it possible to predict political results through research?..

Short answer: Yes… Where it is known that even an atom, which is expected to adhere to basic laws of the universe, can show schizophrenic propensities and, according to physics’ famous double-slit experiment, can develop behavior through observation; a researcher, when making an observation regarding any phenomena, should make a modeling and design that will minimize his/her manipulation (technically speaking) on the addressee.

In case that a research is considered to be solely comprised of field work and the answers coming from the addressee and the statement is accepted as a fact whose rationality and analysis is not required, it is inevitable that the researcher and the research will yield deviational results.

Through a research work in which fundamental principles – of research discipline and other disciplines that it takes support from – that should be assessed at logical scale are utilized in a plain manner; it is possible to form a data set that will enable the prediction of political elections. At this point, however, the researcher must perceive the research – throughout background, conduct, and sequel (analysis) phases – as a holistic work.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: