Marketing Research For Non-Research Professionals: Inquiring Minds Want to Know

Do you ever feel that your research department or supplier is speaking a foreign language? Well you are not alone. Many marketers are not familiar with the terms use in marketing research. By popular request we compiled some of the most frequently used terms for our non-research clients who have inquiring minds.

Types of Research

The following describes the most common types of research:

Primary Research: Research that is developed and collected for a specific end use, usually generated to help solve a specific problem (observation, surveys, etc.)

Secondary Research: Research that is designed and collected for multiple end users, developed for some other purpose than the problem at hand (syndicated services like Nielsen, Hoovers, almanacs, etc.)

Qualitative Research: The collection of in-depth, non-numeric data, traditionally collected face-to-face in an unstructured manner, concerned primarily with understanding consumer attitudes and motivations

Types of qualitative research frequently conducted include: focus groups, one-on-one interviews, telephone in-depth interviews.

Quantitative Research: The systematic collection of numeric data, usually involves sampling techniques and sizes which enable the analyst to make projectable conclusions

Types of quantitative research frequently conducted include: customer satisfaction surveys, product preference tests, tracking studies, awareness and usage studies.

Terms Commonly Used in Market Research

The following terms are often used when designing a study or in presenting the findings:

Respondent: A person participating in a research study

Population: A finite or infinite collection of individual things – people, objects or events (i.e., all people in Dallas who own a Honda)

Sample: A portion of a population (percent of people in Dallas owning a Honda)

Random Sample: Each possible sample from a population has an equal and known chance of being selected (can be very difficult to achieve)

Quota Sampling: Interviewers asked to return with a quota of interviews from certain kinds of respondents (e.g., 20% with women who own Hondas)

Proportion Sampling: A sample in which the number of elements drawn is proportionate, relative to the number of elements of the population

Client-Supplied Sample: A list of employees, customers, vendors, etc., supplied by the client commissioning the research

Screening: Procedure often used when the population to be sampled is a small part of the general population. It consists of selecting a sample of the general population and then qualifying (screening) the sample of the general population for the specific population to be interviewed

Gross Incidence: The percent of the entire population who are product-category users; for instance, of the entire female heads-of-households in Dallas (population), the percent that have purchased Kodak film in the past six months; the client usually supplies this proportion

Net Incidence: The net incidence after all qualifiers have been factored in (i.e., in the example above, qualifiers might be that all women must be age 18+ and own a home)

Benchmark Study: The initial measurement against which all subsequent measurements are compared in a tracking study

Tracking Study: The purpose of a tracking study is to study attitudes, awareness, or buying habits over a period of time in a consistent fashion. The sample, as well as the interviewing techniques, must be consistent throughout the tracking study, since the tabulations from one study to another will be analyzed for changes or shifts in awareness, attitudes or buying habits. Tracking studies are often sold in waves for use over a specified length of time.

Monadic: Test design where a respondent evaluates only one product (or idea), having no other product for comparison

Paired Comparison Design: Respondent evaluates two objects at a time, he or she selects one of the two according to some criteria (i.e., overall quality, performance)

Completion Rate: The percentage of qualified respondents from whom a completed interview is obtained; the expected completion rate for any given study depends on the persuasive skill of the interviewer, the length of the interview, the number of call-backs, the sensitivity of the interview and how long the study will be in the field

Tables: The cross tabulations from the quantitative questionnaire results

Banner: One set of banner points that fit on one page; usually up to 21 points or column headings

Banner Point: The horizontal column headings in cross tabs (tables), usually represents the sub-groups being used in the analysis (i.e., male/female, users/non-users, income, age, etc.)

Stub: The vertical row of responses in cross tabs which represent the potential answers for each question (yes/no, definitely would/probably would, etc.)

Hypothesis: An assumption or guess about some characteristic of the population being sampled (i.e., hypothesis = all females in Dallas have blonde hair)

Descriptive Statistics: Used to summarize and describe data rather than generalize from a small body of data to a larger system of similar data (inductive statistics); commonly used descriptive statistics include Chi-square, T-tests, and A-tests

Chi-square: Measure of the goodness of the fit between two numbers observed in the sample and the numbers we should have seen in the sample given the null hypothesis is true (that is, there is no difference between the two groups)

T-test: Used to determine if the means from two independent samples differ enough to conclude that the populations are statistically different with respect to their means, used on sample sizes less than 30 (Z-test is the same calculation; used with sample sizes over 30 or with known variance)

SignificanceTesting: Statistics that help determine the probability of obtaining the observed results (differences, etc.) by chance, if this probability is very small, smaller than some pre-specified value called the significance level, we say the results are significant at that pre-set level

Confidence Intervals: A range around the sample estimate in which the population estimate is expected to fall with a specified degree of confidence, usually 90 to 95 percent

About Kathleen Turner

Kathleen Turner (Kathy Yeaton), MBA, President of i2s Advantage (formerly Marketing & Research Partners, based in Dallas Texas), helps Fortune 500 and start-up organizations create and sustain a competitive advantage and build value. Kathleen is a frequent writer and speaker in numerous associations. Contact Kathleen for other articles or more information at 214-632-0183, via email KatTurn@gmail.com, or visit i2S Advantage online at www.i2sadv.com.
No comments yet.

Leave a Reply