In the initial stages of a research project, an important first step is to prepare a diagram that illustrates the research hypotheses and displays the variable relationships that will be examined. Specifically, a conceptual model is a diagram that connects variables/constructs based on theory and logic to visually display the hypotheses that will be tested. Preparing a conceptual model early in the research process enables researchers to organize their thoughts and visually consider the relationships between the variables of interest. Conceptual models also are an efficient means of sharing ideas between researchers working on or reviewing a research project.
Two examples of conceptual models are shown in Exhibit 6-1. The top model has two measured variables connected by a single-headed arrow and is an example of a model of a bivariate regression. The measured variables are represented as rectangles and the arrow indicates that advertising budget is related to or predicts sales. The model at the bottom of the exhibit is more complex and represents a multiple regression model, much like the one you are asked to prepare for your Research Interest Overview. There are four measured variables all represented as rectangles. The three measured variables on the left are independent variables and the single measured variable on the right is a dependent variable. Each independent variable has an arrow that indicates that it is related to or predicts the dependent variable sales. Thus, the model indicates that sales is predicted by the advertising budget, the number of salespersons, and the amount of website traffic. All of these variables are measured with a single question and we think of them as individual variables. Also, no signs (+/-) are shown on the arrows to represent the orientation of relationships between the independent and dependent variables. Therefore, our hypotheses only show there is a relationship between the variables, but not whether the relationship is positive or negative (directional).
AD Budget
Sales
# Salespersons
Sales
AD Budget
Website Traffic
Exhibit 6-1 Conceptual Models of Measured Variables
The conceptual model in Exhibit 6-2 illustrates a more sophisticated analysis of relationships. There are four variables shown as ovals. In this situation, we refer to them as constructs because each variable is considered as latent and not directly measured. Instead, there are several questions that measure the construct indirectly. Note that researchers sometimes use the terms variable and construct interchangeably but constructs are always measured indirectly by indicator variables.
To understand the model in Exhibit 6-2, the variables/constructs must be defined. First, the construct “Technology Acceptance Climate” is the extent to which employees’ ongoing use of information technology is rewarded, supported and expected within the organization. Although not shown in the model, it is measured indirectly with several questions (measured indicator variables). Second, the construct “Shared Values” represents the feelings and beliefs that form an organization’s culture and provide a basis for individuals to understand the organization’s functioning and norms for behavior
.
Examples of shared values include: customer orientation, entrepreneurial values, adaptive cultural expectations and information sharing norms. Thus, the shared values construct is measured with several questions as well. Third, “Technology Implementation” is how quickly and completely the new information technology is integrated into ongoing
operations and “Productivity” is higher output per day. The Technology Implementation and Productivity constructs are both measured with several questions as well.
Technology Implementation
Productivity
Technology Acceptance Climate
Shared Values
+
+
+
Exhibit 6-2 Four Construct Conceptual Model
The constructs “Technology Acceptance Climate” and “Shared Values” are independent variables/constructs in the conceptual model. In contrast, the construct “Productivity” is a dependent variable. The construct “Technology Implementation” is more complicated in the model because it is both a dependent and an independent variable. In other words, it is a dependent variable because it is predicted by two independent variables (Implementation Climate and Shared Values) but it also is an independent variable because it is shown as predicting the construct Productivity.
Three hypotheses are illustrated by the arrows in this conceptual model. They include: "Technology Implementation is positively related to the Technology Acceptance Climate”, "Technology Implementation is positively related to the organization’s Shared Values", and “Productivity is positively related to Technology Implementation.” There is a plus sign (+) by all three arrows so all three relationships are represented as directional and positive.
Note that constructs in all conceptual models are represented in a sequence based on theory, logic or practical experiences observed by the researcher. The sequence of the constructs is illustrated from left to right, with independent (predictor) constructs on the left and dependent (outcome) variables to the right. That is, constructs to the left side are assumed to precede and predict constructs to the right. When there are more than two sets of constructs represented in a conceptual model constructs on the right are always assumed to be predicted by constructs on the left. Moreover, constructs considered as dependent in a conceptual model often are referred to as endogenous variables. Any construct that has an arrow pointed into it is an endogenous variable. Constructs that operate as both independent and dependent variables in a model also are considered endogenous. Finally, constructs that are always only independent variables are generally referred to as exogenous variables. Exogenous variables (constructs) only have arrows that point out of them and never have arrows pointed into them. In Exhibit 6-2 Technology Acceptance Climate and Shared Values are exogenous constructs. In contrast, Technology Implementation and Productivity are endogenous constructs.
The order of the constructs in the conceptual model shown in Exhibit 6-2 is based on the following assumptions. In an organization the climate of acceptance for new technology and the shared values both influence the extent to which new technology is implemented. Thus, implementation of new technology is dependent on and predicted by the climate of acceptance and shared values. Moreover, the Technology Acceptance Climate and Shared Values constructs are referred to as antecedents of technology implementation. In addition, implementation of new technology is expected to lead to higher productivity so the technology implementation construct is an independent variable that predicts productivity, a dependent or outcome construct.
When the sequence of the constructs has been decided, then the connecting arrows representing the hypothesized relationships must be drawn. The arrows are inserted with the arrow pointed to the right, which indicates the sequence and that the constructs on the left predict the constructs to the right. The predictive relationships are sometimes referred to as causal links, if the theory supports a causal relationship. If theory does not support a causal relationship then the link between constructs is considered a correlation. In connecting the constructs with arrows all the possible connections may not be included. For example, the model in Exhibit 6-2 does not have an arrow between Technology Acceptance Climate and Productivity, even though one could be drawn between the two constructs. An arrow is not drawn there because theory does not support such a relationship. Thus, arrows are drawn on conceptual models only where theory or logic supports a hypothesized relationship.
When you prepare your literature review you should include a written description of your conceptual model as well as an actual drawing of the model. The section of your literature review that describes your model typically is called a conceptual framework. The written description integrates all the information about the problem/opportunity in a logical manner, describes the relationships among the variables, explains the theory underlying these relationships, indicates the nature and direction of the relationships, and includes a conceptual model. Exhibit 6-3 provides guidelines on how to prepare a good conceptual framework.
Exhibit 6-3 Guidelines for Preparing Your Conceptual Framework
· The variables/constructs considered relevant to the study are clearly identified and defined.
· The sources of constructs are clearly identified. If new constructs are developed for the study the process for developing the constructs is explained, and their validity and reliability reported.
· If published constructs are used, their validity and reliability is reported for both the published study and for your own research.
· The discussion states how the variables/constructs are related to each other, i.e., which variables are dependent (endogenous) or independent (exogenous).
· If possible, the nature (positive or negative) of the relationships as well as the direction is hypothesized on the basis of theory, logic, previous research or researcher judgment.
· There is a clear explanation of why you expect these relationships to exist. The explanation cites theory, business practice or some other credible source.
· A conceptual model or framework is prepared to clearly illustrate the hypothesized relationships.