Know When to Use Top-Down and When to Use Bottom-Up Approaches

Market research is grounded in the branch of philosophy known as logic. Two logical reasoning approaches are basic to research design. These approaches are known as deduction and induction.

Deductive Research

Deductive reasoning is a top-down approach that works from thegeneral to the specific. In empirical research, that means that a market researcher begins a study by considering theories that have been developed in conjunction with a topic of interest. This approach lets a market researcher think about research that has been already been conducted and develop an idea about extending or adding to that theoretical foundation. From the topical idea, the market researcher works to develop an hypothesis. This new hypothesis will be tested by the market researcher in the process of conducting a new study. Specific data that has been collected and analyzed in the new study will form the basis of the test of the hypothesis. The specific data will either confirm the hypothesis, or it will not. [It is important to note that an hypothesis that is not confirmed has not been proven false.]

Deductive Research Steps

 

• GENERAL – Literature Search & Theories
• Topic of Interest
• Theory-related Idea
• Hypothesis
• Data Collection
• Data Analysis & Hypothesis Testing
• Confirm the Hypothesis or Not
• Disseminate Findings

 

Inductive Research

Inductive reasoning is a bottom-up approach that moves from the specific to the general. In this case, specific refers to an observation made by the market researcher that eventually leads to broad generalization and theory. [It might be important to note – for discussions with colleagues or in public – that the term is bottom-up and not bottoms-up. Bottoms-up is a sort of toast for drinking, something that may seem entirely appropriate once the research study is completed.]

An inductive research methods approach begins with specific observations made by a market researcher who begins a study with an idea or a topic of interest, just as in a deductive approach to research. However, in an inductive approach, the researcher does not consider related theories until much further along into the research. From the observations or measures that the market researcher conducts – generally in the field and not in a laboratory setting – clusters of data or patterns begin to emerge. From these regularities or patterns, the market researcher generates themes that come analysis of the data.

Inductive Research Steps

 

• SPECIFIC – Observations & Measures
• Topic of Interest
• Data Collection
• Data Clusters or Patterns
• Data Analysis
• Themes Emerge
• Generalizations
• Disseminate Findings

 

Quantitative Research and the Hypothesis

If the market researcher is conducting quantitative research, at this point, theories can be considered. However, if the market researcher is conducting qualitative research, then the formal hypothesis testing does not take place. Rather, the market researcher may formgeneralizations based on the strength of the data and themes that have emerged.

Data collection and data analysis in qualitative research is iterative. That is to say, it data collection doesn’t happen all at once and then — as though the market researcher has thrown a switch — data analysis begins. Rather, some data is collected, which is considered by the researcher, and then some more data is collected and considered, and so on. At a certain point, when sufficient data clusters or patterns have emerged, the market researcher will decide that thedata collection can slow, stop, or change direction.

Data collection and data analysis in quantitative research are distinct stages. To mingle data collection and data analysis in the manner of qualitative research would compromise the integrity of the data. Some scientist would say that a lack of boundaries in the data collection and data analysis processes causes the data to become contaminated and the research to lack rigor. Findings from such compromised research would not be viewed as robust.

Causal Inquiry, Exploratory Inquiry, and Everything In-Between

Bottom-up research methods feel more unstructured, but they are no less scientific than structured top-down research methods. Because each type of research approach has its own advantages and disadvantages, it is not uncommon for a study to employ mixed methods. A market researcher who uses mixed methods applies a deductive research approach to the components of the study that shows strong theoretical ties. Alternately, an inductive research approach is applied to the components of the study that seem to require a more exploratory inquiry.

Its a misrepresentation to form a mental picture of deductive approaches and inductive approaches as two sides of the same coin. In practice, they are two ends of a continuum. Deductive research is associated with linearity and a search for causal relationships. Inductive research is associated with depth of inquiry and descriptions about phenomena. Mixed methods can be placed at about mid-point on that continuum with an emphasis on research breadth.

This article contains a much simplified explanation about the different types of deduction and inquiry. There are many layers to market research. The content in this article just begins to scratch the surface. For instance, if we consider the philosophical grounding of deductive and inductive reasoning, we might refer to the approaches as positivistic and naturalistic.

Link to original article: http://marketresearch.about.com/od/market-research-quantitative/a/Market-Research-Deductive-Versus-Inductive.htm