Formulation of effective hypothesis

1.1 Introduction

What do we mean by formulation of hypothesis? And two, what describes an effective hypothesis?

1.2 Definition

Hypothesis formulation is the STEP ONE process of evaluating the merit of a hypothesis using a verifiability or falsifiability-oriented experiment or method to either prove or disapprove the hypothesis. The end result then is that, either the hypothesis is true hence accepted or it’s false hence rejected.

Hypothesis formulation is therefore quite paramount and establishing a good one is value adding for it helps in theory building. Therefore, a testable hypothesis is not just a simple hand jotted statement but one need to think about it in the proper manner. This takes us then to the scientific steps of formulating a hypothesis, i.e. “hypothesizing”


This means claiming about a population parameter. It entails setting of assumptions (Hypotheses) that will be pegged on when testing a theory. The hypothesis being formulated is unproven yet it acts as a proposition or supposition that tentatively explains certain facts or phenomena and it can empirically be tested.

Things to know in step one;

i). That it is not possible to start doing our investigation or inquiry in research unless we express a solution-based explanation of the issue causing the problem. The tentative (i.e. for the time being) explanations should be traceable from either the subject matter or researcher’s previous knowledge or experience. Therefore, the hypothesis (plural hypotheses) is a tentative solution of a problem and the aim of undertaking a research is not to provide a solution to the problem but aims at verifying the hypothesis.

ii) To be complete hypothesis must include three components –

  • The variables
  • The population
  • The relationship between the variables

iii) Hypothesis does not need to be correct, for its purpose is to predict what the researcher expects to see or discover.

iv) The role or goal of hypothesis, is to determine whether this guess is right or wrong. State the problem that you are trying to solve.

v) Hypothesis is decomposed in to two; hypo and thesis.



HYPO; Carries twofold meanings

One; Tentative or subject to confirmation. That is the suggestion or proposition given is awaiting approval or disapproval through a scientific method

Two; Composition of two or more variables, which is to be verified.

THESIS: It is a statement expressed to imply the solution of a problem. It therefore means the position taken by the variables aforementioned in the specific frame of reference.

Many PhD scholars complete their doctorate degree but they cannot tell you what was their “THESIS!”. BY the way let me tease your mind a little bit, you cleared your PhD and graduated? So what was your THESIS anyway?

If you are asked such a question, how do you answer that?

Hint: In this case, you are being asked about the position taken by your new variables that you introduced in your study which was based on the past studies you have reviewed in your chapter two of literature review or from any other source of your hypothesis. Position of your variables means what they have explained in addition to an existing relationship or body of knowledge.

Some researchers and scholars will cite or try to remember the topic of their PhD paper and refer to it as their THESIS. For example, they may say my thesis was on the “Effects of Drugs in Study Behavior of University Students

Oops! This is not a correct answer. Assume that you were the one who did that aforementioned study. You will agree with me that in your literature review, you will have reviewed past studies on the various behaviors amongst University students that has been instigated to them by drugs. And for sure you will have discovered many diversified behaviors from past empirical reviews and theoretical reviews.

So in your research efforts you also agree with me that you cannot write another research paper similar to the ones you have reviewed in literature and claim that you have added new knowledge in the existing body of knowledge!

For your information, the above topic, “Effects of Drugs in Study Behavior of University Students”, represents existing body of knowledge on the relationship between drug abuse and study behavior of University students. So, the question is; what new idea(s) are you bringing on board? Remember, if you propose a research, which is the same as that of your predecessor researchers, then the chances of your paper being thrown through the window, is high.

Now let me take you to the previous viewpoint of the term THESIS. This term means the position taken by the variables aforementioned in the specific frame of reference.

What kind of position is this? The position can be either;

Theoretical or Conceptual -Hence theoretical knowledge gap

Methodological-Hence methodological gap

Contextual-Hence contextual knowledge gap.


Those are the positions that the variables you propose to use in your study may assume for you to explain what your THESIS was in your PhD study or your research assignment. In the next explanation, the focus will consider three aspect only

Further explanation;

2.1.1 Conceptual Variable Position

Conceptual variable position refers to the variable position played by a variable based on the conceptual or theoretical role it plays in a model. You should note that if a variable assumes a certain conceptual characteristic such as being a predictor, dependent, moderating or intervening in nature, then this turns to be their conceptual or theoretical variable position. Those several conceptual positions are summarized below;

Independent variable, also known as the predictor or estimator or predator variable (i.e. factor)

Dependent variable, also referred to as criterion, response or output variable.

Intervening variable, also known as intermediating variable

Moderating variable

Control variable

Extraneous variable etc.

Each type of variable has a distinct role that it plays in any relationship in the natural phenomena. For example, an independent variable plays the role of influencing another variable referred to as the dependent or criterion variable. One key characteristic of an independent variable is that it can be manipulated or controlled. So, the researcher manipulates this variable to observe the changes on the dependent variable. The independent variable is used to predict whether the changes that will occur on the dependent variable will be significant or not.

On the other hand, the dependent variable plays the role of response to any influence emanating or originating from the theoretical predictor variable. It is therefore not subjected to manipulation by the researcher as it is in the case of the independent variable. The following examples details this matter using conceptual framework for that purpose.

Example 1

Researcher “one” assigned variable X the role of independent variable and Y the role of dependent variable. This is theoretically/conceptually represented by Figure 1.1


Note that this is the original relational phenomenon on how X relates with Y. This is what we will refer to as the existing body of knowledge about X and Y variables as they exist in the natural way.

Example 2

In another study, Researcher two assigned variable Y and X the role of independent variable and dependent variable respectively. In other words the roles of X and Y were reverse as compared to example one above. This is presented in Figure 1.2 below



Researcher two now proposes that Y should take the position of independent variable and the X to be the dependent variable contrary to the body of knowledge that already exists (refer to example 1 in Figure 1.1). If again it is empirically proven that Y significantly play the role or position of prediction in influencing X then this is new conceptual knowledge. Therefore, it represents researcher “two” THESIS for it is empirically proven that Y has played the position of a true predictor of X and not the other way round.


Example 3

Researcher “three” considered variable X to influence Y as it is in the first example in Figure 1.1. That is, X was assumed to be the independent variable while Y still played the role or position of dependent variable. However, he further introduced Z in the model to test its intervening effect on the relationship between X and Y as shown in Figure 1.3. 



If it is further empirically or scientifically proven that Z significantly intervenes the relationship between X and Y, this is then new conceptual knowledge added to the existing one and it represents researcher “three” THESIS.

NOTE that, if we consider other variables with either moderating effect in the same model or any other variable role, it will bring in additional conceptual knowledge as discussed in other theoretical/conceptual knowledge gap articles.

2.1.2 Methodological Variable Position

This refers to the variable position played by a variable/construct based on the method that was used to measure it. You should note that if a variable is gauged or measured in dissimilar ways, it will translate to dissimilar research findings as compared to past studies which used the same variable(s).


Example 1

Let us assume that X, Y and Z variables which were used in the previous discussions were measured by researchers in different ways. To be more comprehensive, consider;

 X=to represent factors that influence financial performance of a business. Hence this is the independent variable

Y=to represent financial performance of business which is the dependent variable. Then let’s look at several cases as indicated below so as to comprehend the aspect of methodological variable position in a better way.


Case one:

Assume researcher “one” carried a research on the factors that determine the financial performance of a business. He proposed Market share (i.e. X) as one of the factors and that it was measured using the number of customers that specific firm serves as compared to the total number of customers who buy goods of the businesses under that industry. So the ratio scale used was



In this case, as shown in Figure 1.4 above, market share is the independent variable and it plays a methodological variable role or position where by it explains the size of the market share of the firm by comparing the customers who buy goods of all firms in that particular industry.

Financial performance on the other hand, is the dependent variable and it plays a methodological variable role or position of expressing the level of profitability of the firm by comparing the net profit generated per annum to total equity of the firm.

Now, in this case above, we will assume that the model represents the existing methodological knowledge as far as measurements of market share is concerned. Now, let case one be assumed to be the benchmark in our next cases we are going to discuss.  

Case two

Researcher “two” carried out literature review and got case one research findings where by market share was measured using the ratio of number of customers to that of the whole industry but he was not convinced that this is the right method of gauging market share. Then he proposed a different method.


That is, in his argument, he felt that if Market share is gauged like this,



It will be more appropriate to explain real changes in financial performance of the business. The reason being that the number of the customers who buy goods of a specific firm(s) may not even be truly loyal to the products of that specific firm. So, researcher “two” focused on the “return customers” category only, who repeatedly buy goods from that particular firm. This may represent the size of the market share of the business in a better way for consistency in buying means significant influence on the financial performance of that particular firm.

The conceptual framework representing this relationship between market share and financial performance with an appropriate method of measuring the predictor variable is as shown below in Figure 1.5



The measurement used for market share in case two as per Figure 1.5 is more logical as compared to the previous one where all buyer of goods of the specific firm could either be permanent or temporal. The temporal ones may not have much impact on the financial performance of the business and may contribute to spurious (non-sense relationship).

So in conclusion, you can see that when market share variable is gauged differently the level of significance relationship or influence on the dependent variable will vary. Therefore, the same variable will have a different methodological position role to play. This applies to all aspects in research.


2.1.3 Contextual Variable Position

This refers to the variable position played by a variable/construct based on the context where that variable applies. You should note that if a variable is considered as per the context where it logically applies, then the same variable will assume dissimilar contextual variable position accordingly and this will translate to dissimilar research findings as compared to past studies which used the same variable(s).


You should note that context may vary from global, regional, local, industry or nature of the firm under investigation. In the third case we are going to focus on the nature of the firm under analysis. In our case three, we will focus on context in terms of nature of the firm in which the variable is used.


So, in this case the variable maybe used as per what that particular firm in that particular industry refers to be logical. For instance, for the listed firms in different stock markets, profitability of a firm is measured using various proxies some of them being, namely; Return on Equity (ROE), Return on Assets (ROA) and Return on Investment (ROI). This is contextually logical because this is their common practice in each industry with firms listed at the respective bourses. Secondly, the data is sufficiently available in the public domain for most of listed companies are required by the respective constitutions to publish their accounts. Let’s look at an illustration using example 3.


Example 3

Researcher “three” carried out a research of the factors that influence financial performance of Small and Medium Enterprises (SMEs). From literature review she came across the research findings of past studies were by market share influenced financial performance of firms listed at the securities market. However, she felt that the measurement of profitability of SMEs could not rely on ROE. This is because it is in the public knowledge that most of medium and small businesses do not keep proper books of accounts and no constitutional requirement by the respective government to have their accounts published. Therefore, the data for ROE is scanty and unavailable. Therefore, she proposed a different approach. For instance, in her argument, she felt that if profitability is gauged as rate of employee turnover expressed as;


So in summary the existing knowledge shows that market share influence financial performance (ROE) of listed firms as indicated in the conceptual framework in Figure 1.5, then he may re-consider using a different proxy to gauge the dependent variable if it is a case of medium and small firms (SMEs) such as the rate of employee turnover.

Therefore, if it is a case of medium and small firms, then the conceptual framework will appear as follows as depicted in Figure 1.6




ROEMPLO is the Rate of Employee Turnover

E0 is the number of employees at the beginning of the year

E1 is the number of employees at the end of the year


In this case, researcher “three” is measuring financial performance for SMEs in a more logical way if he uses rate of employee turnover instead of ROE. This is because medium and small enterprises do not keep proper books of accounts and even if they do, they are incomplete and the records are kept in secrecy such that one cannot access them for analysis. Therefore, replacing dependent variable position to explain performance in terms of rate of employee turnover is more logical or plausible for it contextually make sense to medium and small enterprise owners/operators.

Therefore, researcher “three” THESIS is portrayed if this proposition is proved in a scientific way.

 Therefore, with proposition or supposition that performance variable play the contextual position or role of explaining performance using employee rate of turnover, achieves that goal.


In summary, hypothesizing step or stage involves the setting of the hypothesis statement and shedding more light on the variable position assumed by the study variables. There are three mainstream variable positions, namely;

Conceptual/theoretical variable position

Methodological variable position

Contextual variable position

All the three categories of variable positions can be used to establish the thesis that the researcher wish to add to the body of the existing knowledge. 


Also the term hypothesizing comes from the common term hypothesis which is unpacked in to two words

Hypo- which means two things; hypothesis is tentative or subject to confirmation and comprises of two or more variables, which is to be verified.

Thesis-is the solution to the study problem. Hence the variable position assumed by the variable introduced by the researcher is the solution if at all it is scientifically proven to play the role suggested by the researcher.

After formulation of an effective hypothesis, the proceeding steps or procedure is that of hypothesis testing. This step entails; determination of level of significance and critical region, then actual data collection and then actual hypothesis testing whereby the appropriate hypothesis testing decision rule is used and then a conclusion based on hypothesis decision rule used is arrived at.

About the Author - Dr Geoffrey Mbuva(PhD-Finance) is a lecturer of Finance and Accountancy at Kenyatta University, Kenya. He is an enthusiast of teaching and making accounting & research tutorials for his readers.