Home - Answers Community - Algebra Calculator - Student Deals - MBA Guide - Business Degree Guide - College Algebra - Mathematics - Research Statistics - About
Free Homework Help

Primary Research: Survey, Observation, and Experimental Design


After defining the objective of the research, we first exhaust examining the secondary data available. After this analysis, we determine what type of information needed to meet the objective of the research. The primary/survey research is widely used in the business. This is a very effective tool for business-to-business and business to consumer research.

We perform primary research when the secondary sources do not provide enough information to be analyzed or when we want to find information from our customers or potential customers. Survey is an integral part to perform the primary research.

The survey simply means that we ask questions to our target audience and collect information before analysis. To conduct a survey we need to perform the following steps.


¨ Develop research objective based on the defined problem

¨ Design and develop the survey instrument (e.g. questionnaire)

¨ Develop a representative sample (whom we should talk to)

¨ Develop research methodologies including mode of operation

such as telephone, internet, in-person, mail and a hybrid model.

¨ Develop a test sample to validate the effectiveness of the questionnaire

developed (this step is very important before start implementing the actual project).

¨ Design and develop the analysis tools, most often statistical tools to be used for data analysis

¨ Design tables and format to be used for storing collected data

¨ Determine and plan what will you do for analyzing the comments and answers to the open ended questions

¨ Develop techniques to train thee interviewers

¨ Develop report and presentation formats

¨ Develop a measurement model to ascertain the effectiveness of the research project. This is used after the research is completed and implementation of the marketing program started.

It is very important to develop a representative sample that reflects the characteristics of the actual target audience. There is a misnomer that if you choose a random sample and it is large enough, it will represent the population. It is true that you sample should be random and large enough but it should represent the targeted population to be effective. In business, there are a number of techniques to develop a proper sample that represents the population.

It is very important in designing a proper scale for the questionnaire. The five point or seven point scales are more effective. I like the five-point scale, as it is not large. The scale should be odd as they present a mid-point and helps human to answer properly. I personally do not like even point scales because human mind always look for a mid point. The ten-point scale is very clumsy as it gives too many choices and could easily confuse human mind and may develop response errors. Also, please keep in mind that the subjective scale such as “ do not like, like some extent, like very much etc. etc” should be avoided as we all define them differently.

The observation technique is used mostly to understand human behavior or a situation. While using the observation technique, the marketing researchers record the behavioral, situational occurrences and objects. We do not use any questionnaire or formal analysis. Sometimes, the observation technique is subjective and the result is dependent on individual researcher’s evaluation and judgment. This technique is widely used in psychology, archaeology, TV audience rating analysis and behavior sciences.

There are some ethical issues (borderline in what is right and what is wrong) in this technique. The law should not be broken while using this technique.

In recent years we utilize a hybrid model of direct observation and personal interviewing of the targeted audience.

The observation is a tool available. We could use both human and mechanical devices to achieve our goals. While using this technique we should try to reduce the observation bias.

Experimental Research

This is a technique where we have control over the research situation. This is mostly used for understanding the casual effect on different variables. The assumption is that we know the variables (both independent and dependent) and we have a well-developed hypothesis. We try to:

¨ Confirm or disprove the hypothetical relationship between variables

¨ Estimate the strength of the relationship

We use this technique to prevent preconceived notions from excluding potentially useful results.

The experimental design helps us to achieve the goal as explained above. The stages are:

¨ Manipulation (such as price, advertising, positioning etc. etc. etc.)

¨ Treatment levels/experimental groups (combines levels of manipulation to

which the individual is exposed)

¨ Control Group: Subjects who are not exposed to treatments

¨ Measurement: The recording of a response of the respondent by any

means such as survey and observation.

The variation of the results of an experiment depends on:

¨ Treatment effects

¨ Measurement effects

¨ Interaction effects (between the above two)

¨ Randomness

¨ Other unknown or ignored variables

The success in interpreting the result of an experiment is to separate the above-mentioned effects.

There are various kinds of experiments used in this context. These are:

¨ Laboratory experiments

¨ Field experiments

¨ Natural experiments

comments powered by Disqus
Get Final Exam Answers from ACCNerd