Index R contains definitions for terms such as return-on-investment, or ROI, random sampling, representative samples, regression analysis, research design, and reality augmentation.
Return-on-investment, also commonly known as ROI, is one of several financial ratios that show how well an effort is doing in terms of the profit or revenue gained for an investment of funds. Return-on-investment is shown in terms of a percentage or ratio that measures the performance of an investment of a company’s own capital. Companies are focused on what they get for the money they spend. In other words, when capital is invested in research, especially market research as we would have it, what “value” is obtained in return.
In marketing, ROI generally is an expression of the gross profit derived from specific marketing expenses. ROI is all about what invested capital “buys” that is of value to the company spending the money. When a company pays for market research—either by paying the salaries of internal market research staff or by contracting with external market research providers—that company wants to “own” new consumer insights as a result of spending that money. In addition, the quality of those consumer insights.is important in these transactions. Generally speaking, consumer insights must be actionable to be of value.
And Now, A Few Words About Our Samples
Sample selection is important in both quantitative research and qualitative research. A sample is a slice of a larger population, but it is generally a slice that is based on something known or suspected about the larger population. Samples of the larger population are selected in order to be able to say something general about the larger population— target universe—from what is learned about the sample. A research sample is constructed by using techniques and strategies that contribute to reliability and validity. Market research is based on this idea that a representative group of people can be identified and accessed in ways that permit generalization to the target population—which is too impossibly large or too expensive to include in any one research study.
In market research, a representative sample refers to a group of consumers who are characterized by the same sort of traits and preferences as the larger target universe of consumers. The match between the sample and the universe must be strong for all attributes anticipated to be influential to the research outcomes.
Samples that are selected through a randomizing process are referred to as random samples, and they are the gold standard in research. It is not always possible to select samples on a random basis, and it may not even be preferable for some types of research. For instance, qualitative research is generally not concerned with random sampling techniques because the data collected in qualitative research is not often subjected to statistical procedure—for which random sampling has a distinct advantage. Samples may be constructed by techniques that assign members to a particular sample. While this is rational from one perspective, it is complicated by the fact that many research conclusions are made in error because of the impact of chance (probability) in sample selection.
Other Market Research Terms
Reliability is a term used to talk about the "test-retest" quality of a measurement. In the simplest terms, reliability refers the capacity of a measurement (perhaps a test) to obtain the same results over repetitions when all the underlying conditions remain the same. A test would not be considered reliable if it produced different results every time it was used to measure something. Think about what it would be like to use a ruler that was made out of some pliable material that changes shape when it was handled. That ruler would not be at all reliable and it would be very frustrating to use. Moreover, the people using the ruler would not be able to talk meaningfully about measurements taken with that ruler. In research, a number of variables can change over time that impact reliability, and there are statistical procedures to correct for those changes in a manner that still permits data to be reliably collected.