*by*

*Paul J. Dejillas, Ph.D.*

*Professor of Anthropology*

**T**he chapter on theoretical considerations consists essentially of four sections: review of related literature, theoretical framework, hypotheses, and the conceptual model. The presentation may not necessarily proceed in the order presented here, but I find this order more useful. I want to go to the literature review immediately in order to know what the other authors are saying about the research problem and what theoretical frameworks and conceptual models they are using. In this manner, I can be assured that, when building my own framework and model, I am not duplicating their works.

**Review of Related Literature**

There are many good reasons why a good literature review is necessary. First, we want to document any existing studies that deal, partly or wholly, about our present research problem and objectives. In this manner, we do not run the risk of “re-inventing the wheel.” If, in the process of our literature review, we discover that the problem has already been treated exhaustively, then, we can go back to Chapter 1 and restate our problem, by focusing on an issue which existing literature have not yet dwelt fully or by emphasizing an untreated or unresolved question in their works.

A good literature review thus demonstrates the researcher’s knowledge and familiarity on the topic being investigated. It reflects the researcher’s ability and depth to undertake research activities. From the viewpoint of the reader, the more comprehensive and exhaustive the review, the greater is the confidence of the reader both on the competence of the researcher and on the quality of the research.

There are several types of doing literature review. Many authors identify the following types: self-study, context, historical, theoretical, methodological, and integrative review (Bieger and Gerlack 1996, Nachmias et al. 1997; Heiman, Gary W. 1998). Seasoned researchers agree that using most or all of these types ensures one with a good literature review.

*Self-study review*, which is so designed as to stock the researcher with all available knowledge about the topic that he is researching on.

*Context review*, which places the researcher’s study in the context of a bigger picture. It enables the researcher to establish how the current research continues the work that other researchers have begun and stopped. Thus, context review serves also the purpose of introducing the researcher’s work and establishing the significance and relevance of the present research question.

*Historical review, *which traces the development of the researcher’s chosen topic over time. It provides additional knowledge to the reader how the topic evolved and how it developed over the years.

*Theoretical review, *which tries to present a comparison of how different theories discussed by many authors address the same topic being investigated. It may then emphasize how these different theories have not able to directly and fully address the present problem being investigated by the researcher. In view of this, it opens an opportunity for the researcher to develop or improve existing theories that allow the treatment of the research problem being investigated. Because of its importance in building theories and models, this type of a review is usually reserved and presented under the section on “theoretical framework.”

*Methodological review, *which seeks to discuss, and even assess, the various methodologies---research design, sampling design, data collection instruments and methods, analytical measurements---utilized by other authors or studies in responding to their respective research problems and objectives.

*Integrative review, *which seeks to gather all available details and information about the research topic and problem being investigated, then, analyze all these information and data. It, then, summarizes what other authors have already done in terms of their significance and relevance to the present study, evolution and development of the topic, theories and models used, hypotheses established (if any), research methodologies adopted, and even the findings, conclusions, and recommendations drawn from their studies.

In literature review, the researcher notes down every observation drawn from other authors or studies within the context of how this relates to his/her present study. Some literature may address only one or a few aspects of the researcher’s problems, while others may not have dwelt at all on the researcher’s problems. This observation needs to be noted down and expressed in the literature review since these are limitations which may justify the researcher’s need to pursue the chosen study. In particular, this will justify the researcher’s attempt to develop a new theoretical framework and conceptual model, or improve existing ones. This is one reason why literature review ought to come first before even one builds his/her own framework and model.

Again, there are no rigid and fixed rules in formulating the review of related literature, but some authors recommend the presentation to be between 2,000 and 2,500 words or the equivalent of 8-10 pages typed in single space and using font 12. In addition, some institutions expect a minimum of 25 related studies to be reviewed in the case of master’s theses and 50 in the case of doctoral dissertations. Doing more than this is even better. One ought to note, however, that some topics may still be new or are rarely studied.

**Review of Related Studies**

Review of related studies is different from review of related literature. Related studies refers to other works, like thesis, dissertations, applied research studies undertaken by research agencies, and the like that are related to the topic, subject, and problems that you are doing. In reviewing these studies, it is necessary to mention their problems, hypothesis, methodologies, findings, and conclusions. The review of related studies is necessary in Chapter 2 because almost always you are not the first who is doing a study on the subject that you are researching.

**Theoretical Framework**

*Inductive-theory building* is characterized by a strictly empirical approach to finding generalizations and relies on the repeated observation of reality, after which summary statements are made to explain and classify what is observed.

*Deductive-theory building* is a form of inference that derives its conclusions by reasoning through premises, which serve as its proof; its emphasis is distinctly on the conceptual structure and its substantive validity; and focuses on conceptual development prior to empirical testing. While this method of theory building has been criticized for its general lack of reference to reality, it has resulted in useful results in a number of fields of interest.

*Functional-theory building* is marked by a continual interaction of conceptualizing and subsequent empirical testing; this method requires a constant interaction between theory and facts. With functional theories, numerous smaller conceptual models are usually constructed and tested until a more grandiose scheme can be built on empirical evidence.

*Model-based theory building*, the best examples of which come from the field of economics, specifies highly mathematical models that are subjected to rigorous statistical testing to see if they perform in the hypothesized fashion. If these models perform, they are often refined and integrated into existing theoretical structures within the field. If they do not, they are discarded or modified for future testing. Under this method of construction, the emphasis up front is on defining a conceptual model, then subjecting it to empirical testing. Essentially, substantive validity concerns are not of interest in the initial stages of theoretical development. The only concern in the initial stages in the model's testing is whether or not the model performs in the fashion desired by the researcher.

In building the theoretical framework, one needs to highlight or pay particular attention to the following:

· The variables or set of variables---whether endogenous or exogenous---included in the model.

· The relationships between and among these variables.

Variables can be distinguished into dependent and independent. The *dependent variable* is the variable that the researcher wishes to explain. The *independent variable* is that variable the researcher expects will explain changes in the dependent variable.

The independent variable is also called the *explanatory variable*; it is the presumed cause of changes in the values of the dependent variable. The dependent variable, also called *criterion variable,* is expected to be caused or influenced by the independent variable, also called *predictor variable*.

A good theoretical framework identifies and labels the important variables in the situation that are relevant to the problem(s) identified. It logically describes the interconnections among these variables: the relationships between the independent and the dependent variables.

A *relation* in research always refers to a relation between two or more variables. When we say that variable X (e.g. education) and variable Y (e.g. income) are related, we mean that there is something common to both variables, i.e., one affects the other. The relationship may be that "individuals with higher education have higher incomes."

The relationship of two variables can be:

*Positive* (e.g. education, measured in years of schooling, tends to be higher when incomes are rising). In this example, the relationship between education and income are moving in the same direction; the relationship is direct.

*Negative* (e.g. the demand for a given product---assuming a normal product---decreases as its price increases). Here, the relationship between demand and price is moving in the opposite direction; the relationship is what economists term as "inverse."

It can also happen that there is no relationship between an identified independent (or explanatory) variable and the dependent variable (criterion). If this is known beforehand, then, it may no longer be necessary to include this particular independent variable in the model. For example, "mountains" have no direct relationship with "education." But other variables like: distance, the presence or absence of electricity, number of children in the family, etc. can have direct impact on "education." If the distance between the school and the place where the pupil or student resides (measured in terms of kilometers) is extremely far that the latter have to cross mountains and rivers daily, then, this may have a significant bearing on the quantity and quality of education of the student.

Here, one sees the need to carefully identify the variables that are relevant to the study. One way of doing this is through a good literature survey.

One should also realize the need to clearly formulate the relationships of the variables that are identified in the study, and state whether the relationship is positive (direct), or negative (inverse).

**Hypotheses**

Just as the literature survey sets the construction of a good theoretical framework, a good theoretical framework, in turn, provides the strong logical base for developing testable hypotheses. There is no need to over-emphasize the point that the identification and formulation of testable hypotheses is still part of the development of a good theoretical framework.

Hypotheses are drawn from the theoretical framework, in particular, from the nature, extent, and direction of relationships of the identified variables. In quantitative research, the extent of relationship is determined by the value of the beta coefficients; meanwhile, the direction of the relationships is indicated by the signs of the beta coefficients.

The hypothesized relationships can be tested through appropriate statistical measures. Such measures enable the researcher to accept or reject a given hypothesis at some level or degree of confidence and significance.

A good theoretical framework enables us to develop good, testable hypotheses, where the conjectured relationships would hold.

But even if testable hypotheses are not necessarily generated (as in some applied research project), there is still a need to develop a good theoretical framework in order to systematically examine the problem under investigation.

A *hypothesis* predicts a particular relationship between two or more variables. If we think that a relationship exists between two or more variables, we first state it as a hypothesis and then *test the hypothesis* in the field.

By *test*, we mean either to confirm it to our satisfaction or to prove it wrong. Webster (1968) defines hypothesis as "a tentative assumption made in order to draw out and test its logical or empirical consequences… Hypothesis implies insufficiency of presently attainable evidence and therefore a tentative explanation."

**Conceptual Model**

The conceptual model operationalizes the theoretical framework; it translates theory into reality through some mathematical/statistical measures or qualitative analytical tools.

It is normally a schematic diagram identifying the various variables the research intends to study. These variables may be categorized into exogenous and endogenous.

Variables need to be identified also in terms of whether they are dependent or independent variables.

More importantly, the conceptual model expresses the relationship (hypotheses) between and among the various variables under study, in terms of the signs affixed in the beta coefficients in the case of statistical or econometric models.

Thus, a conceptual model contains the following elements:

1. A functional model that makes possible the translation or transformation of the theoretical framework of the study into some statistical/ mathematical measures or other qualitative analytical measures. For example:

*Education is a function of family income, distance between the school and the place where the student/pupil resides, presence or absence of electricity, availability of transportation, etc.*

2. A symbolic (mathematical/statistical or qualitative) expression of the theoretical framework that intends to apply or translate the latter to reality. This symbolic expression highlights the variables identified in the theoretical framework. We say symbolic here because variables are converted into specific symbols of Y, X_{1} and X_{2}, etc. For example:

*Y = f (X _{1}, X_{2}, X_{3}, X_{4})*

Where:

Y = education

f = is a function of

X_{1} = family income

X_{2} = distance between school and home of pupil

X_{3} = presence or absence of electricity

X_{4} = availability of transportation

3. Mathematical model that expresses the relationships (also hypotheses) of the identified variables, given as signs. For example:

*Y = a + b X _{1} - cX_{2}, + dX_{3} + eX_{4}*

The variables X_{1}, X_{2}, X_{3}, X_{4 } are the independent variables, while Y is the dependent variable.

One of the hypotheses expressed in the above mathematical equation expresses the negative relationship between education (Y) and distance X_{2}. It is thus hypothesized in the study that the greater the distance between the school and the residence of the student/pupil, the lesser the quality and quantity of education (measured in terms of number of schooling) of the student becomes. This could be justified in the fact that the number of hours consumed in traveling eats up the time that should have been consumed for studying the lessons learned in school.

Of course, this particular hypothesis will still be tested. If the sign of the beta coefficient (c) that will come out in the regression runs will indeed be negative, then, the hypothesis is validated to be true.

But if the sign is positive, then, the hypothesis that education varies inversely with distance is to be rejected. What could this mean? This could mean that those students/pupils who live far away from school are more motivated to study harder, are forced to find more time to study their lessons, or spend more of their travel time productively by studying along the way. But this is only one of the many interpretations that can be advanced if the sign of the *beta coefficient* is positive.

The conceptual model is expressed in terms of mathematical or statistical models and measurements (see example given above). These models and measurements are quantifiable expressions of the relationships between the dependent and independent variables, given as signs (positive or negative) in the equations. This is true when the research being undertaken requires a quantitative approach.

It is on these mathematical equations that raw data, gathered from actual cases or situations, are plugged in and fed into the computer for processing.

In the case of qualitative researches, a descriptive presentation of the nature and relationships of the variables you want to study is sufficient.