These are the things that have not been completed:Classify each variable as independent or dependent.In 3-5 paragraphs, explain your work. Include the name of the statistical tool you used to calculat

Inzinerine Ekonomika-Engineering Economics, 2013, 24(1), 28-35 The Quantitative Content Processing Methodology: Coding of Narratives and Their Statistical Analysis Sigitas Vaitkevicius, Lina Kazokiene Kaunas University of Technology K. Donelaicio st.

73, LT-44029, Kaunas, Lithuania e-mail:

sigitas. vaiktekicius@ktu. It Kaunas University of Technology Panevezys Institute Nemuno st. 33, LT-37164, Panevezys, Lithuania e-mail:

Una. kazokiene@ktu. It cn>ss»'ef http://dx.doi.org/10.5755/j01 .ee.24.1.2350 The paper deals with the quantitative analysis of the qualitative content, which enables converting the variables used in qualitative research into quantitative variables and comparing them to variables used in quantitative research. This methodology is based on the combination of Romanticist, Phenomenological, Dialectical, Critical, and Post-structural hermeneutic systems of interpretation described by Demeterio III (2001). The specially designed content systematization technique, which describes the logical sequence of content construct identification and is based on Wilkinson's (1991) Key and Hook technique and Merkys 's (2005) operationalization, was developed by using the dichotomous coding technique.

The paper describes the methodology and the areas of its application. It also gives a practical example of its application in a research study.

An integrated testing and development of the methodology was performed by investigating evaluation efficiency of business public relations. The paper introduces the context and scope of the applied procedure in the research. It also presents the data processing steps, the performed analysis and the key results. The presented research in this paper is discussed in three phases: input, implementation and impact assessment phases, where fragments of each are provided.

The article focuses on the narrative content.

Keywords: research methodology, quantitative and qualitative variables, comparability of qualitative and quantitative information, qualitative and quantitative data.

Introduction Currently a social research methodology is so advanced that can cover the entire research process starting from the identification of phenomenon itself and concluding with the investigation of the spread of typological features in the whole population (Bryman, 2007).

Qualitative and quantitative cognition technology allows to study the separate identities of social subject and to gain new sociological knowledge in formalized and transformed way by the latest modem mathematics and computing technique achievements (Leitch, 2007).

Depending on different methodological approaches, application traditions and possibilities that are determined by problematic situation and the nature of the phenomenon under study the way of cognition ean be both qualitative and quantitative (Creswell & Garrett, 2008; Vallejo- Martos, 2008, Carraher et al, 2009, Ciabusehi & Martin, 2012).

As evidenced by research practice and theory these methods effectively complement each other and in order to get to know the research object properly their interaction gives more comprehensive, objective and reliable results (Denscombe, 2008; Freshwater, 2007; Greene, 2007; Creswell et al, 2003) though sometimes these groups are shown as opposing and incompatible but in scientific literature there are attempts to look for these different methodological approaches coordination opportunities (Reichardt & Rallis, 1994; Tashakkori & Teddlie, 1998; Moon & Moon, 2004; Johnson & Onwuegbuzie, 2004).

Currently MIX - method direction gains popularity and treats traditional research approaches as an integrated methodology of quantitative and qualitative research (Denscombe, 2008; Freshwater, 2007; Greene, 2007; Creswell et al, 2003). (Plano et al, 2007; Johnson et al, 2007; Creswell et al, 2011) describe the mixed methods research as a research design with philosophical assumptions as well as methods of inquiry, where:

• As a methodology, it involves philosophical assumptions that guide the direction of the collection and analysis of data and the mixture of qualitative and quantitative approaches in many phases in the research process; • As a method, it focuses on collecting, analyzing, and mixing both quantitative and qualitative data in a single study or series of studies. Its central premise is that the use of quantitative and qualitative approaches in combination provides better understanding of research problems than either approach alone.

Analyzing application examples of the qualitative and quantitative research approaches (Chipunza & Gwarinda, 2010; Golicic & Davis, 2012), it was noticed that their inherent properties condition the application consistency:

-28- Sigitas Vaitkevicius, Lina Kazokiene. The Quantitative Content Processing Methodology: Coding of Narratives...

a) Identification of phenomena. Fact-finding and data analysis according to the grounding theory.

b) Qualitative research. Description of the identified phenomenon and identiflcation of its typological features.

c) Quantitative research. Identifying occurrence of the phenomenon and distribution of its typological features in the population.

Such an application seems completely rational because in the process of social cognition most oñen it is impossible to perform all the process contemporaneously.

Due to big data volume and big quantity of carried actions it is likely that the research under all conditions will have to be carried out according to certain logical consistency therefore the researcher could potentially face some problems of data comparability.

We can distinguish comparison issues of ontological and methodological qualitative and quantitative research data. Ontological investigation problem is determined by complex, multidimensional, contextual, stochastic and constructional psychosocial analysis of social reality. This means that if a researcher performs a full-length research it can take several years to accomplish it starting with the identiflcation of the phenomenon and concluding with the description of its typological features present in the population. Another situation, which calls for altemative solutions, is comparison of data gathered during qualitative and quantitative parts of the study. Such situations are examined in the works of (Morse, 1991; Greene et al, 1997; Creswell et al, 2003; Eliot, 2005; Ivankova et al, 2006; Bryman, 2006; Creswell et al, 2007; Teddlie et al, 2009).

During the research studies qualitative and quantitative data more often are collected relatively separately and the results are usually compared at a qualitative level. Of course, one can use various verbal content quantification programs, such as T-lab, NVivo, Kokybis etc. However they are not powerful enough to quantify qualitative content from scientific point of view, because they create qualitative variables based on syntax of the sentence rather than on real authorial intention. As a result, even when verbal constructs have been recorded and coded in order to compare them with quantitative data, in most cases the interpretation of the psychosocial reality depends on the ability of the researcher to penetrate the phenomenon, its structure and its qualitative features.

Methodological research problem is also determined by psycho-social complexity which leads to the idea that cognition is impossible only with qualitative and quantitative means. In most cases, to understand the nature of the phenomenon completely it is not enough to employ only case study analysis, focus groups, non-stmctured interviews or questionnaire surveys. The limitations become especially apparent when only quantitative methods are used without combining them with qualitative research methods. Furthermore there is one more important issue that can arise due to a large number of questions, namely the leaming effect on the part of respondents.

When conducting various studies it was noticed that questions which respondents did not mention as important in the qualitative study, in quantitative research sometimes outweighed the ones they had mentioned because a quantitative questionnaire, due to its more comprehensive nature, tends to reflect respondent's social opinion rather than personal one. In the case of qualitative research respondent tends to talk about issues that concem them directly, failing to identify general social stimuli and factors of social environment, which influence them indirectly with their state without a special reminder.

In order to solve this methodological problem, the attempts were to integrate open-ended questions into close- ended questionnaires. This helped to shorten the questionnaire and get some new data about the phenomenon.

However, lack of methodology for comparing the results at quantitative level resulted in limited scope of the research.

Such integration helped to expand the structure of attributes but don't analyze their distribution in the population. Moreover, it was noticed that when an open- ended question expanded the scope of a close-ended question, when it was located after a set of questions it expanded upon, phrases used in close-ended questions dominated the answers of the respondents despite the fact that they had already answered such questions. This showed that while filling in the questionnaire respondents not only adapt to its content, but also start to imitate it developing an idealistic archetype of their identity. Summing up, the longer the questionnaire the greater the probability that questionnaire responses will show an idealistic model of behavior instead of a realistic one.

The research problem is a design of methodology, which would allow the using of the Hermeneutic systems for statistical coding of the authorial intentions extracted by the qualitative methods such as interviews and narratives or open-ended questions.

The aim of the paper is to familiarize the reader with the quantitative content processing methodology, the course of its compile and to give an example of its application.

This methodology must meet the following quality criteria: objectivity, representativeness, reliability, validity, efficiency and utility.

Novelty. So far, information was collected relatively separately using qualitative or quantitative cognition way and the results of these researches were compared most often at qualitative level. Prepared the quantitative content processing methodology enables to compare the information gathered in a qualitative and quantitative cognition way and to get reliable complex reality research results.

Object of the research - coding of narratives for statistical analysis.

Methods of research - theoretical modeling, quantitative and qualitative analysis of empirical data.

Description ofthe proposed Methodology The main idea of the quantitative content processing methodology is that the principles used for coding verbal content can be used for statistical coding of respondents' opinions, attitudes and feelings obtained using the hermeneutics systems described by Demeterio III, (2001).

In order to ensure repeatability of the results a special content systemizing sequence was employed for coding. It was based on Wilkinson's (1991) Key and Hook technique, which describe logical sequence of the content constmct identiflcation, and on Merkys's (2005) operationalization technique supplemented by dichotomous coding.

-29- Inzinerine Ekonomika-Engineering Economics, 2013, 24(1), 28-35 Combining all the pieces into a single structure produced the following content processing and coding sequence (Figure 1):

1.

Reading the text and understanding its overall meaning (based on romanticist, phenomenological, dialectical, critical, or post-structural hermeneutics systems of interpretation).

2.

Identification of text's logical structure (based on Wilkinson (1991) Key and Hook technique) 3.

Dividing the text into logical passages.

4.

Comparative analysis of the logical passages and their conflation into coherent logical sequence.

5.

Operationalization ofthe logical structure:

a. Interpretation of theoretical (extensive) concept T| used by the respondents; b.

Classification of T| into constrictive concepts t|, t2, ..., t„, according to the respondents' intention, which have a lower degree of abstraction than T,; c. Creation of ti, t2,..., tn empirical indicators E|, E2, ..., En; (Expressed by notions or by the logical construct) d. Development of measure scale and confirmation of E|, E2, ..., En metrological criteria ei, e2,..., en.

6. Dichotomous coding of the notions, which were, mentioned as 1 and those which were not mentioned as 0.

Extensive conœptj Constrictive concept, t] Constrictive concept, tj Empirical indicator (notion), E, Empirical indicator (notion]i^E„ Empirical indicator (notion), El Empirical indicator (notion)!, E„ El metrological criteria (mentioned notion), e E„metrologicalaiteria (mentioned notion), e El metrological criteria (mentioned notion), e E„ metrological criteria (mentioned notion), e j-Yes(l) '-Not(O) j-Yes(l) '-Not(O) j-Yes(l) '-Not(O) j-Yes(l) '-Not(O) Figure 1. The dichotomous coding model for concept T The use of the hermeneutic system for content processing enabled: first, to decode methodically the meaning of the text; second, to perform analogical actions with visual, non-verbal, emotional, audio and symbolic content (Demeterio, 2001); third, working experience with hermeneutics allowed to notice that the type of hermeneutic system not only determines content recognition, but also defines its logical structure - therefore this technique can be applied in wider research context.

Finally, the hermeneutic system can be related with the typological model of the research aims. For example, in social research of individuals' or groups' attitudes sometimes one needs to understand the real intention of their attitudes. This aims the application of romantic hermeneutics, which is especially valued in intercultural research. Such aims can be attributed to the first group of typological models. The second group of research aims involves understanding of respondents' or their groups' attitudes in research context. It is related to dialectical hermeneutics and especially useful in confirmative research. The third group involves understanding the interaction of opinions that forms the images in researcher's sub-consciousness and is related to phenomenological hermeneutic which is useflil in exploratory research. The fourth group involves the aims of diagnosing hidden pathology of opinion and at the efforts to remove ideological distortions. These aims are related to critical hermeneutics and they are especially important when one needs to eliminate conformist attitudes and political ideology ofthe respondents. Finally, the fifth group involves the aims of understanding the principles behind formation of one's opinion, which are closely related to sub-structural hermeneutics and especially useftil in exploratory research used for analyzing the reasons behind population's behavior and for assessment of the face validity of research tools.

Recognition ofthe logical structure ofthe text is used as a tool to ensure content validity. It enables identifying the statistical potential of the text and researcher's notes used for recording the opinions, attitudes and feelings. The methodology is designed in such a way that after coding the text it becomes possible to measure the statistical relationship between constructs. For this purpose, Wilkinson's (1991) method for identifying the logical structure of content is used. It enables integrating the insights gained at the stage of hermeneutic analysis into one logical structure which, depending on the hermeneutic system used, will reflect the logical setup of the text, respondents' opinions, attitudes and feelings, or researcher's insights.

Identification of logical passages of the text, their analysis and integration into one logical sequence is a preparatory step before operationalization. Its purpose is to identify the topics and sub-topics discussed in the text which qualitatively describe the subject of the study. The methodology is based on the view that mathematical testing is primarily a tool for statistical confirmation ofthe theoretical insights rather than an exploratory tool. For this reason, the structure of logical relationships has to be uncovered first and only then coded. Finally, the models of these relationships are examined.

Analysis of the logical passages of the text can be horizontal and vertical. In this methodology, vertical analysis means identifying the same logical passages in different parts of one respondent's text and conducting their comparative analysis. Meanwhile horizontal analysis means identification and comparison of the same logical passages across the texts of several respondents.

Operationalization of logical structure is performed as a background for concept's dichotomous coding (Figure 1).

During the operationalization, original concepts are extracted up to a conditional empirical indicator, whose presence in the text is assessed based on whether it is mentioned, i.e. if it is mentioned it is coded as 1, if not-as 0. Other strategies can be applied for concept T coding.

For example: when analyzing the intensity of relative use of certain concepts, the frequency of mentioning them can be coded instead ofthe fact of them being mentioned.

-30- Sigitas Vaitkevicius, Lina Kazokiene. The Quantitative Content Processing Methodology: Coding of Narratives...

The Context of the Proposed Methodology Application Testing of the methodology and development of methodological background was done due to Evaluation of Effectiveness of Public Relations in Business Organizations study in the doctoral dissertation written by the author and supervised by the other author of this paper. Saving the space of this paper only an illustrative example of this methodology is presented. Overall study stmcture is presented in Figure 2.

INTERVIEW SURVEY Input stage Asîesment of public relation activities Public relation activities' implementation stage Public relation influence stage Assesstnent of objectives Assessment criteria for implementation time Assessment criteria for characteristics ofthe target audience Assessment criteria for expected results Pressrelease assessment critaü Announcement's presentment assessment criteria Final results assessnient criteria Outcome assessment criteria Productivity assessment criteria Figure 2. The context and scope Research study had three stages: (1) Input stage.

Qualitative method was used at this stage due to its creative nature, relatively little knowledge available and the qualitative diversity of decisions that had to be taken.

(2) Public relations activity implementation and (3) Public relations influence stages, due to relatively clear and sufficiently researched typology and stmctural simplicity, were analyzed using notions-attitudes questionnaire with close-ended answer format. This enabled to assess the applicability of this method in the research.

Dichotomous coding of interview results on the subscale for assessment of the predicted productivity of public relations (PR).

The meaning ofthe text was identified using the system of dialectic interpretation (Demeterio, 2001). It was chosen because ofthe research aims. In this case, uncovering the ofthe methodology application unified meaning of the text was less important than identifying the meaning relevant to this particular research study.

While the texts were repeatedly listened to and re-read, a kind of discussion between the researchers' biased position, their presumptions, and the biased text with presumptions contained in it took place (Figure 2).

The reading cycle started with the projection of the researcher's biased position and presumptions onto the text and then allowing the text to reveal its position and presumptions in this context. Such reading mode enabled to maintain the integrity of the content without losing the track of the respondents' flow of thought. Next using Wilkinson's (1991) Key & Hook technique the logical stmcture of the text was analyzed (see the right column in Table 3.

Table 1 Logical structure and logical passages of text fragments Logical passages based on dialectic hermeneutics Question — What should the lower level goals cover?

Respondents audience size indicators (Fragment No. 1) •S ig Purpose of the goal > .E Answer. How big an audience I will reach having in mind that the circulation is 100 thousand. We can assume that 3-4 people have read the newspaper. And it is even bigger for a radio, this means how manv people actually noticed and positively remembered. Sooutof those 17 million positively remembered 4 and noticed such.

So let's summarize. The mostly who speaks is... and then more precisely, i.e. how manv articles and in which publication... This is my work tool. I need to know what are our weak points.... products, finance management, which speaker is stronger, where I need to put more effort, how we plan to reach the potential audience, what I can leave as it is.

Question - What is the usual purpose of public relations in your company?

Answer. In some cases the clients set goals, e.g. financial, increase in sales ... then one should view public relations as a whole, let's say. as one ofthe components of marketing strategy. Then very close cooperation is required, PR should very closely cooperate with other components of marketing. Let's say if the desirable number is about 20 customers PR professionals are asked to help improve relationship with these customers and, naturally, increase sales. Let's suppose PR professionals develop certain loyalty programmes but for example, if sales agents that are responsible for these customers are incompetent and don't perform their job well, sales will not increase. So 1 mean that PR is not a miracle or a magic word using which one can reach one's goals. Initially it is very important with the client to learn their goals and expectations. Let's say for example in our case how we work, PR is mostly. Not mostly but most often they irjfluence people's opinion.

Opinion of a certain target audience or ... sometimes there is .. there may exist a negative predisposition or no opinion when let's say it is a novelty. Then we can introduce quantitative goals.

Let's say product x ... nobody ...

still knows about it. such product doesn't exist, but we introduce a target that in .. that in one year target audience will learn: this target audience is formed, by .. well using demographic, social and other parameters and let's say 10 or 20 per cent of it will ¡earn.

If we see that it is truly possible to achieve that, then why not?

* Logical sequence ofthe text was identified using the Key & Hook technique (Wilkinson. 1991). where we used underlined words as keys (hooks) showing their logical sequence. Different underline types represent different key sequences.

Purposefulness ofthe goals Limiting factors Effectiveness of public relations Respondents audience size indicators (Fragment No. 2) -31 - Inzinerine Ekonomika-Engineering Economics, 2013, 24(1), 28-35 For the analysis of logical structure the special meaning of the text was used. It was identifled using the system of dialectic hermeneutic interpretation and related to the key mark (underlined and wave underlined) integrated into the question.

This was the starting point in the identification of text's logical structure. In the first question, the logical structure was related to the term "size" while in the second one it was related to the term "goal". The analysis revealed that the logical structure of "goal" is broader than that of "size" because in this study size was one of the goals. In this way, the hierarchy of the entire content in terms of meaning was underlined too.

Table 2 Horizontal and vertical analysis of logical passages o u ç ze u ••5 < c cm 2 A .c — >. ra g M ra _ V Fragment No.

1 (chronological sequence) How bis an audience I will reach having in mind that the circulation is IOO thousand. We can assume that 3-4 people have read the newspaper.

And radio has bigger audience that means how many of people really noticed and positivelv remembered So out of these 17 million positively remembered 4 and noticed such. So lets totalize. The mostly who speaks is...

and then more precise, i.e. how many articles and in which publication...

Determinants: Fragment No. 2 <- Horizontal analysis of fragment No.

1 and 2 -> o 0 .

0 • o • o o Potentially reached audienceor partof it 10 or 20 per cent of it ... 3-4 have read.

will learn 10 or 20 per cent of it will learn No.

of positive, neutral, negative articles about the company.

... how manv... positively remembered Potentially reached audience or part of it ... how manv... remembered and noticed Potentially reached audience or part of it ...outofthese...4... remembered, and noticed No.

of positive/ neutral/ negative articles about a company... positively remembered 4...

The next step was identification of logical passages ofthe text (the left column in Table 1). The purpose of this exercise was to create logical indexes, which would enable further horizontal and vertical analysis of text fragments.

Such analysis based on the "Audience size indicators" logical passages is presented in Table 2. This analysis identified eight statements falling within three logical definitions: the number of press releases, the potentially reached audience or part of it, and the number of positive / neutral / negative articles about the company. Analysis shows that although some ofthe responses are expressed in figures, they can only be regarded as providing qualitative rather than quantitative data. This is because when for example a respondent talks about circulation of the printed material he or she do not provide exact numerical information offering a rather general example instead. This means one cannot use the ratio scale for quantitatively defining this variable but a nominal scale can be used.

Table 3 Operationalization model for the "Productivity" concept Concept, T Question/examples of response fragments (concepts, t„) Empirical indicators E„ of concepts tn, E„ metrological criteria, e„ What should the lower level goals cover? / "Potentially reached audience or part of it" ... if the circulation is 100 thousand. We can assume that 3-4 people have read the newspaper. And radio has bigger audience that means how many of people really noticed and remembered positively. So out of 17 million positively remembered 4 and noticed such.

So let's totalize ; What should the lower level goals cover? / "Potentially reached audience or part of it" ... planned to reach a potential audience...; What is the usual purpose of public relations in your company ?/ "Potentially reached audience or part of it" ...Then we can introduce quantitative goals. Let's say about product x ...

nobody ... still do not knows, such product like doesn 't exist, but we introduce a parameter that in .. that in one year will learn from target audience; let's say this target audience is formed, by .. welt by the demographic, social and other parameters and thereat from it will ¡earn say ¡0 or 20 per cent... .

No.

of positive, neutral, negative articles about the company Topic where company is mentioned No.

of times the company was mentioned in the media.

Potentially reached audience or part of it ei - ...will read...

newspaper...; e2 -... plan to reach potential audience..; e3 - target audiences [part of it]...; Mentioning 1, Not mentioning 0 Next using the operationalization procedure the operationalization model ofthe "Productivity concept" was created (Table 3). It unifies the entire sequence of content interpretation and coding into one logical structure. This model differs from the more commonly accepted ones because in this case operationalization of the concept takes -32- Sigitas Vaitkevieius, Lina Kazokiene. The Quantitative Content Processing Methodology: Coding of Narratives...

place at the level of understanding rather than semantic terms.

This enables unification of the meaning of various definitions and analyzing them using mathematical methods.

Construction of the Predicted PR Productivity Index The indicator of the predicted PR productivity assessment was constructed by creating a mean of four primary variables: the number of positive/negative articles, topics, the number of times the company was mentioned in the media and the potential audience (Table 4).

The results of reliability analysis show representative capacity of the index. However, due to a relatively small number of mentions, it is more appropriate to use induction rather than deduction for assessing the indicator's quality.

Reducing the potential number of testing insights to the scientifically justified level it can be argued that the constructed indicator in similar conditions will definitely refieet the totality of its attributes. On the other hand, if we interpret the results of such tests from the empirical theory of truth perspective, the results confirm that the model of transformation of the primary variables into the index is correct, and that index construction models would also be suitable for defining the overall change in attributes in the larger scale research studies.

Table 4 Correlation of impact assessment indicator within variables PR impact assessment No.

of times the company was mentioned in the media.

Topic where company is mentioned Potentially reached audience or its part No.

of positive, neutral, negative articles about the company Corrected Item-Total Correlation 0,537 0,537 0,537 1,000 Construct validity parameters Inter-Item Correlations rmean 1* .

Tmax Cronbach's Alpha ,769 ,537 1,000 ,889 from the context of the real authorial intention as well.

Introduced methodology enables to solve information gathered through qualitative and quantitative way comparability issue and expands the boundaries of scientific understanding of the social researeh process.

o Gives to the researchers more possibilities to explore the coded text mathematically and to link obtained insights into solid logical structure by identifying and positioning in assessment system new not investigated attributes of the research object and instantly identifying its infiuence on the behavior of the object.

Conclusions During the compile procedure there was found, that the Hermeneutic analysis could be useful tool for quantitative content processing. The hermeneutic analysis allows the understanding of the real authorial intention and is usefiil for the preparation of it for the further coding.

Testing results of quantitative content processing methodology based on verbal narrative coding principles confirmed:

o The methodology creates premises to compare statistically the information received from the narratives not only from the context of words or their collocations but References Bryman, A. (2006). Integrating Quantitative and Qualitative Researeh: How is it Done? Qualitative Research, 6(1), 97- 113.

http://dx.doi.Org/10.l 177/1468794106058877 Bryman, A. (2007). Barriers to Integrating Quantitative and Qualitative Research. Journal of Mixed Methods Research, 1(1), 8-22. http://dx.doi.org/10.1177/2345678906290531 Carraher, S., Pamell, J., & Spillan, J. (2009). Customer Service-Orientation of Small Retail Business Owners in Austria.

The Czech Republic, Hungary, Latvia, Slovakia, and Slovenia. Baltic Journal of Management, 4(3), 251-268 http://dx.doi.org/10.1108/17465260910990975 Ciabusehi F., & Martin, O. (2012). Knowledge Ambiguity, Innovation and Subsidiary Performance. Baltic Journal of Management, 7(2), 143-166 http://dx.doi.0rg/lO.l 108/17465261211219787 Creswell, J. W. (2003). Research design: Qualitative, Quantitative and Mixed Methods Approaches. [2ed.] Thousand Oaks:

SAGE Publications, Inc, 246 Creswell, J. W., & Garrett, A. L. (2008). The „Movement" of Mixed Methods Researeh and the Role of Educators. South African Journal of Education, 28, 321-333.

Creswell, J. W., & Piano Clark, V. L. (2011). Designing and Conducting Mixed Methods Research. [2ed.] Thousand Oaks:

SAGE Publications, Inc, 488.

Creswell, J. W., & Tashakkori, A. (2007). Editorial: Differing Perspectives on Mixed Methods Researeh. Journal of Mixed Methods Research, 1(4), 303-308. http://dx.doi.org/10.1177/1558689807306132 -33- Inzinerine Ekonomika-Engineering Economics, 2013, 24(1), 28-35 Chipunza, C, & Gwarinda, S. A. (2010). Transformational Leadership in Merging Higher Education Institutions: A Case Study. South African Journal of Human Resource Management, 8(1), 1-10.

Demeterio, F. P. A. III. (2001). Introduction to Hermeneutics [Electronic version, http://www.geocities.com/ philodept/diwatao/introduction_to_hermeneutics.htm]. Diwatao, 1,1.

Denscombe, M. (2008). Communities of Practice: A Research Paradigm for the Mixed Methods Approach. Journal of Mixed Methods Research, 2(3), 270-283. http://dx.doi.0rg/lO.l 177/1558689808316807 Eliot, J. (2005). Using Narrative in Social Research: Qualitative and Quantitative Approaches. London: Sage.

Freshwater, D. (2007). Reading Mixed Methods Research: Contexts for Criticism. Journal of Mixed Methods Research, 1(2), 134-146. http://dx.doi.org/10.1177/1558689806298578 Greene, J. C. (2007). Mixed Methods in Social Inquiry. San Francisco: Jossey-Bass.

Golicic, S. L., & Davis, D. F. (2012). Implementing Mixed Methods Research in Supply Chain Management. International Journal of Physical Distribution & Logistics Management, 42(8/9), 726-741.

Greene, J. C, & Caracelli, V. J. (1997). Advances in Mixed-Method Evaluation: The Challenges and Beneflts of Integrating Diverse Paradigms. San Francisco: Jossey-Bass.

Ivankova, N. V., Creswell, J. W., & Stick, S. (2006). Using Mixed Methods Sequential Explanatory Design: from Theory into Practice. Field Methods, 18(1), 3-20. http://dx.doi.org/10.1177/1525822X05282260 Johnson, R. B., Onwuegbuzie, A., & Tumer, L. (2007). Toward a Definition of Mixed Methods Research. Joumal of Mixed Methods Research, 1(2), 122-133. http://dx.doi.org/10.1177/1558689806298224 Johnson, R. B., & Onwuegbuzie, A. J. (2004). Mixed Methods Research: a Research Paradigm whose Time Has Come.

Educational Researcher, vol.33, no.7, p.14-26. http://dx.doi.org/10.3102/0013189X033007014 Leitch, C. (2007). "An Action Research Approach to Entrepreneurship", in Neergard, H. and Ulhoi, J.P. (Eds). Handbook of Qualitative Research Methods in Entrepreneurship, Edward Elgar, Cheltenham, 144-169.

Merkys, G. (2005). Education Research Methodology. Unpublished Lectures Material. Kaunas University of Technology.

Moon, J., & Moon, S. (2004). The Case for Mixed Methodology Research: A Review of Literature and Methods [interaktyve].

Morse, J. (1991). Approaches to Qualitative-Quantitative Methodological Triangulation. Nursing Research, 40, 120-123.

http://dx.doi.org/10.1097/00006199-199103000-00014 Piano Clark, V. L., & Creswell, J. W. (2008). The Mixed Methods Reader. Thousand Oaks: SAGE Publications, Inc, 640.

Reichardt, S. S., & Rallis, S. F. (1994). Qualitative and Quantitative Inquiries are Not Incompatible: A Call for a New Partnership. In: Reichardt, C. S., Rallis, S. F. /Eds./. The Qualitativequantitative Debate: New Perspectives. San Francisco, CA: Jossey-Bass, 85-91.

Tashakkori, A., & Teddlie, C. (1998). Mixed Methodology: Combining Qualitative and Quantitative Approaches. Applied Social Research Methods Series. Thousand Oaks, CA: Sage.

Teddlie, C, & Tashakkori, A. (2009). Foundations of Mixed Methods Research: Integrating Quantitative and Qualitative Approaches in the Social and Behavioral Sciences. Thousand Oaks, CA: Sage.

Vallejo-Martos, M. C. (2008). The Organizational Culture of Family Firms as a Key Factor of Competitiveness. Journal of Business Economics and Management, 12 (3), 451-481.

http://dx.doi.org/10.3846/16111699.2011.599407 Wilkinson's, A. M. (1991). The Scientist's Handbook for Writing Papers and Dissertations. Engelwood Cliffs, NJ:

Prentice Hall.

Sigitas Vaitkevicius, Lina Kazokiene Kjekybinë turinio apdorojimo metodologija: naratyvq kodavlmas ir jq statistine analizè Santrauka Norint issamiai suvokti, apibOdinti ir aiskinti tiriam^ reiäkinj bei jj tobulinti, reikia derinti skirtingas metodologines prieigas (Reichardt, Rallis, 1994; Tashakkori, Teddlie 1998; Moon, 2004; Johnson, Onwuegbuzie 2004 ). Pastebeta, kad naudojamoms kokybinéms ir kiekybinéms pazinimo technikoms yra bûdingas tam tikras nuosekiumas:

1) fenómeno identifikacija; 2) kokybinis tyrimas; 3) kiekybinis tyrimas.

Is paziûros toks taikymas yra racionalus. Vis tik didelé duomenij apimtis ir tyrimo loginis nuosekiumas sudaro prielaidas atsirasti tam tikroms duomenij palyginimo problemoms. Galima jvardyti dvi problemij grupes: ontologinf ir metodologin^. Abiem atvejais problemos susidaro del komplikuotos, daugiadimensinés, kontekstualios, stohastinés, konstruktinés psicho-socialinés realybés. Ontologinés tyrimo problemos aveju viso ilgio tyrimui atlikti, gali prireikti ilgo laikotarpio. Jei tyrimo objektas yra jautrus laikui, tenka identifikuot^ fenomenq baigti tirti kitoje populiacijoje. Tokiu atveju gali kilti tyrimo rezultatij palyginimo klausimas. Siandien sukurtos kvantifikavimo programos (T-lab, Nvivo, Kokybis) negali savarankiskai kvantifikuoti kokybinj turinj.

Metodologinio pobOdzio tyrimo problemos siejamos su tyrimo ribotumais, respondentij ismokimo efektu, atsirandanéiu pasirinkus vien^ pazinimo technik^. Taigi, dazniausiai psicho-socialinés realybés pazinimas ir toliau priklauso nuo tyréjo gebéjimij jzvelgti tiek patj fenomen^, tiek ir jo Struktur^ bei kokybinés raiäkos charakteristikas.

[vardyti pastebéjimai paskatino ieäkoti hermeneutikos taikymo respondentij nuomoniij, nuostatij statistiniam kodavimui galimybiij.

-34- Sigitas Vaitkevicius, Lina Kazokiene. The Quantitative Content Processing Methodology: Coding of Narratives...

Straipsnio tikslas - supazindinti skaitytoj^ su kiekybinio turinio apdorojimo metodologijos sudarymo eiga ir pateikti jos taikymo pavyzdziq.

Paminetina, kad pateikiama metodologija pasizymi tokiais metodologines kokybés kriterijais kaip objektyvumas, reprezentatyvumas, reliabilumas, validumas, efektyvumas ir naudingumas. Objektas — naratyvij kodavimas statistinei analizei.

Tyrimo metodai - teorinis modeliavimas ir empiriniq duomenij kiekybiné ir kokybiné analizé.

Naujumas - iki §iol kokybiniu ir kiekybiniu pazinimo büdu informacija buvo renkama iä dalies atskirai, o tokii| tyrimij rezultatai buvo lyginami dazniausiai kokybiniu lygmeniu. Parengta kiekybinio turinio apdorojimo metodologija, leidzia atlikti kokybiniu ir kiekybiniu pazinimo büdu surinktos informacijos palyginim^. Tai leidzia gauti reliabilius kompleksiskos realybés tyrimo rezultatus, sumazinant tyréjo gebéjimt^ jtak^.

Pagrindiné kiekybinio turinio apdorojimo metodologijos idéja remiasi Demeterio (2001) isvada, kad verbalinio turinio kodavimo principai is esmés gali büti panaudojami ir hermeneutikos taikymo metu gautij respondento nuomoniij, nuostatq ir pojQciij statistiniam kodavimui. Uztikrinant duomenij pakartojamum^, sudaryta turinio sisteminimo technika ir dichotominio kodavimo technika isplésta operacionalizacija. Turinio sisteminimo technika susideda is teksto supratimo ir teksto loginés struktüros identifikacijos, teksto loginiq pasazq isskyrimo bei loginés teksto struktûros operacionalizacijos.

Sudarant metodologija, laikomasi nuostatos, jog moksliniai tyrimai gali turéti kelet^ tikslij (kartais net vienu metu).

Kiekybinio turinio apdorojimo metodologijos kompleksiniam testavimui ir metodinii| nuostatij formavimui buvo atlikti keli tyrimai. Pradiniai konstrukciniai darbai pradéti atliekant „Strategines analizas instrumentarijaus modeliavim^ mazóse ir vidutinése jmonése" (Vaitkevicius, 2006). Tçsiant kiekybinés turinio apdorojimo metodologijos formavim^, Zuperkienés (2008) „Vadovij asmeniniij ir dalykiniq savybiij ugdymo tobulinimo" disertacijoje panaudotas dichotominis kodavimo modelis. Isryskéjus metodologijos disfunkcijoms, kylancioms del metodinio neisbaigtumo, ji tobulinama ir pakartotinai testuojama disertacijos tyrime „Verslo jmoniij rysiij su visuomene efektyvumo vertinimas" (Kazokiene, 2010). Siame tyrime, tyrimo dizainas iapléstas struktürizuotu interviú, leido identiflkuoti disfunkciji^ priezastis ir sékmingai jas pasalinti. Kadangi 5io metodologijos testavimo etapo esminis privalumas, lyginant su ankstesniais, - tyrimo struktQra, -jos pristatymui straipsnyje skiriamas nemazas démesys.

Minétas tyrimas apémé tris etapus: rysiij su visuomene jvesties, jgyvendinimo ir poveikio vertinimo tyrim?. Atsizvelgiant j tai, kad rysiij su visuomene jvesties etapas yra maziausiai apibréztas ir aiäkus, stokojantis issamios pradinés informacijos, sio etapo empiriniam pagrindimui pasirinkta kokybiné prieiga. Kiekybinio tyrimo prieiga pasitelkta rysiij su visuomene jgyvendinimo ir poveikio vertinimo etapij pagrindimui. Sj sprendim^ lémé moksliniq tyrimij gausa, padéjusi suformuoti issami) teorinj pagrind^ rysii| su visuomene jgyvendinimo ir poveikio vertinimo klausimais. Tyrimo duomenij apdorojimas vyko trimis s^lyginiais etapais.

Pirmajame étape apdoroti apklausos rezultatai, panaudojant faktorinç bei reliabilumo analizes.

Antrajame étape, dichotomiskai uzkodavus interviú rezultatus, atlikta sustambintq kintamqjq stambinimo loginés sekos analizé.

Treéiajame étape atlikta kiekybinii) ir kokybiniq rezultatij sqveikos désningumq analizé.

Aptariant turinio sisteminimq, paminétina, jog teksto reiksmei atpazinti pasitelkta Demeterio (2001) dialektinés interpretacijos sistema. Kelis kartus perklausant ir skaitant tekstus, äaliäkos tyréjo pozicijos ir susiformuotos prielaidos kito, leidziant tekstui ji| kontekste atskleisti sav^^ pozicijq.

Daugkartinio skaitymo modelis sudaré galimybç islaikyti tyrimo turinio vientisum^. Toliau, atliekant teksto loginés struktüros analizç, pirmojo klausimo atveju loginé struktüra susieta su „apimties" terminu, antrojo klausimo atveju su „siekio" terminu. Analizes rezultatais nustatyta, kad „siekio" loginé struktüra yra platesné nei apimties ir yra isryskinta viso turinio reikSmés hierarchiné struktüra. Tolesniame, loginiq pasazq isskyrimo étape, suformuotos loginés nuorodos, leidziancios atlikti teksto horizontaliqj^ ir vertikali^q analizç. Straipsnyje analizuotame pavyzdyje suformuotos trys loginés apibréztys:

praneäimq spaudai skaicius, pasiekta potenciali auditorija (ar jos dalis), straipsniij (neigiamq, teigiamij, neutraliij) skaiêius apie jmonç. Nors kai kurie atsakymai iáreiksti skaiÉiais, jie informatyvumo poziüriu laikomi kokybiniais, o ne kiekybiniais duomenimis. Paskutiniame turinio sisteminimo étape atlikta operacionalizacijos procedura. Ji susiejo visq turinio aiskinimo ir kodavimo sek^ j vientis^ logiák^ Struktur^. Naudotas operacionalizavimo modelis nuo jprasto skiriasi tuo, kad s^vokos operacionalizacija vyksta ne reiksminiij terminq, o sampratos lygmeniu, kuri leidzia unifikuoti skirtingtj apibrézíiij reiksmes ir padaryti jas matematiäkai apskaitomas. Straipsnyje, pateikiant kompleksinio tyrimo rezultatus, eksplikaciniu pavyzdziu parinktas numatomo rysiij su visuomene produktyvumo vertinimo ir poveikio vertinimo suderinamumo tyrimo fragmentas.

Metodologijos testavimo rezultatai patvirtino, kad ]% naudojant, sudaroma galimybé uzkoduotq tekstq tirti matematiskai ir gautas jzvalgas susieti j vienodq loginç Struktur^. Kiekybiné turinio apdorojimo metodologija suteikia galimybç kokybinius duomenis transformuoti statistinei analizei, t.y.

tiriam^j objektg ir jo savybes analizuoti kompleksiskai, naudojant kiekybinius ir kokybinius metodus. Kiekybiné turinio apdorojimo metodologija pasiteisino tiriant kürybinio ir lengvai standartizuojamo proceso sgveikq. Toks metodologijos panaudojimas atvéré galimybes identiflkuoti bei vertinimo sistemoje pozicionuoti naujus iki to laiko neatrastus tiriamojo objekto pozymius ir iskart jvertinti jij poveikj objekto veiksenai. Pagrindinis sios metodikos ribotumas yra susijçs su kokybinio turinio kodavimu, kuris ypac jautrus tyréjo ekspertinio vertinimo jgûdziq lygiui. Kuo tyréjas yra labiau pazengçs ir kuo jis geriau pazjsta vertinimo turinj, tuo didesné tikimybé, kad tyréjas teisingai atpazins ir uzkoduos turinj.

ais ribotumas isaiskéjo disertacijq tyrimij metu, kai buvo paméginta turinj koduoti ne pagal jo tikr^^ prasmç, o pagal kitij autoriq sudarytus tipologiniíj konstruktq modelius. 15 esmés toks kodavimas irgi yra galimas, taciau tokiu atveju pokalbio metu uzduodami klausimai taip pat turétij büti orientuoti j tokio konstrukto turinio atskleidim^, priesingu atveju bûtij issikreiptas vertinimo turinys ir komplikuotijsi tiesos atpazinimas.

Tobulinant kiekybinç turinio apdorojimo metodologijq, tolesniuose etapuose svarbu koncentruotis j jos teikiamus privalumus bei egzistuojancius ribotumus ir pasiülyti adaptyvi^ priemonç tyréjams, nepriklausomai nuo jq ekspertinio vertinimo lygmens.

Raktazodziai: tyrimtf metodologija, kokybiniai ir kiekybiniai kintamieji, kokybinés ir kiekybinés informacijos patyginimas, kokybiniai ir kiekybiniai duomenys.

The article has been reviewed.

Received in August, 2012; accepted in February, 2013.

-35- Copyright of Engineering Economics is the property of Engineering Economics and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.