Quantitative research can be defined as the systematic empirical investigation of social phenomena via statistical, mathematical or numerical data or computational techniques. In contrast to qualitative research, which is typically used to explore a certain phenomenon and to generate ideas or hypotheses, quantitative research is used to test these ideas or hypotheses. A range of statistical methods are used to describe parameters and to test whether hypothesised relations are likely in the population. Another difference between quantitative and qualitative research is that quantitative research collects numerical data like numbers or percentages, whereas qualitative research collects non-numerical data like interview statements or stories. As a consequence, quantitative research can be used to measure the incidence of various views and opinions in a chosen sample.
Underlying quantitative research are two different types of research approaches: descriptive research and experimental research. Descriptive research aims to describe some phenomenon. Experimental research aims to identify cause and effect (causality).
Descriptive research aims to provide an accurate description of a phenomenon or a particular situation. This type of research identifies a given number of variables and describes the relationship between these variables. In contrast to experimental research, descriptive research does not say anything about causality between two or more variables. This is due to the fact that the variables are measured ex post and they are not subject to direct manipulation. Variables that are typically measured in quantitative consumer research are:
The following types of quantitative methods can be labelled as descriptive research: surveys, secondary (market) data and meta-analysis.
Surveys (questionnaires) - Surveys (questionnaires) are a frequently used method of quantitative consumer research. Surveys are a method to collect a wide range of information from a large number of individuals, often referred to as respondents, in a systematic way. Surveys are very well suited to find out from a large number of people how they think, feel and behave with regard to a certain topic. In addition, surveys are often used to get to know more about the demographic and psychographic characteristics of a certain population and are consequently often used as a basis for consumer segmentation.
Important issues in survey research are the sampling methods of individual units from a population in such a way that they are representative for the population as a whole or for a specific target group. In addition, much attention has been paid to the different ways of collecting the data. Traditionally, surveys were paper-and-pencil interviewing, face-to-face interviewing or telephone. Nowadays, these techniques have been almost entirely replaced by web-based, online surveys. Finally, social researchers devote a lot of attention to how questionnaires could be constructed and to methods for improving the number, accuracy and validity of responses to surveys. In social science, in order to increase the validity of surveys, so-called validated measurement scales and items are used.
A distinction can be made between cross-sectional and longitudinal surveys. Cross-sectional studies involve data collection from a population, or a specific target group, at one specific point in time. In contrast, longitudinal studies involve repeated collection of the same variables from the same population, or the same specific target group, over a period of time. Longitudinal studies are often used to study developments or trends. Longitudinal studies make observing changes more accurate than cross-sectional studies and could help to improve the quality of research.
Secondary (market) data - In contrast to surveys, which is data collected directly by the researchers (i.e., primary data), a lot of data is already published or collected in the past by other parties. This is called secondary data.
In consumer and marketing research, secondary data often involves, but is not limited to, market data, like purchase data, financial performance measures and trade statistics. Examples of this type of data are: sales figures, price developments of products, turnover and profit of companies or industries, consumption figures, etc. Useful sources of secondary data are statistical databases, company (annual) reports and scanner data of retail organisations.
Meta-analysis - A meta-analysis is a specific type of secondary data research and refers to methods that focus on combining the results from different, previously published, studies. The aim of a meta-analysis is to identify patterns that may come to light or effects that cancel out each other in the context of multiple studies. A systematic literature review often forms the basis for a meta-analysis.
Experimental research is used to identify a causal relationship between a cause (i.e., independent variable) and an effect (i.e., dependent variable). A cause is an event or treatment that makes something else exist. An effect is the difference between what did happen when the event or treatment occurred and what would have happened when the treatment did not occur. In consumer science, this treatment is often induced by manipulating an independent variable. For example, to test the effect of product colouring (= independent variable) on product liking (= dependent variable), a group of participants which has to evaluate a blue coloured product can be compared with a group of participcants which has to evaluate a product without colouring. In experimental design this implies that you have to use at least two groups: an experimental group (which receives a certain treatment) and a control group (which does not receive that treatment).
An important aspect to infer causality is that nuisance or other variables that could affect the dependent variable are controlled. For example, participants (often called subjects in experimental research) should be randomly assigned to experimental conditions to rule out the possibility that personality characteristics influence the results. In addition, ideally, an experiment has a pre-test (measuring the dependent variable before the treatment) and a post-test (measuring the dependent variable after the treatment), to ensure that only the independent variable (treatment) is accountable for the observed changes in the values of the dependent variable.
In sum, controlled conditions allow researchers to isolate the effects of one or more (manipulated) independent variable(s) on a dependent variable. Dependent variables are typically measured in terms of behaviour: choice behaviour (e.g., product choice, willingness to pay, etc.), sensory (i.e., see, hear, smell, feel, taste) or physiologic behaviour (e.g., blood pressure, brain activity, etc.).
An often heard disadvantage of the experimental method is its artificiality. Experiments are conducted in controlled settings, so the outcomes are often less generalizable to real-life situations. Furthermore, some argue that experimental researchers tend to view human behaviour as mechanistic and manipulable, while it is much more complex.
The following types of quantitative methods can be labelled as experimental research: laboratory experiments (including physiological responses), field experiments and quasi-experiments.
Laboratory experiments - As the name already implies, a laboratory (lab) experiment takes place in a laboratory under controlled conditions. This allows researchers to be better able to manipulate variables (treatments) and to exert more control over the outcomes of the experiment. A major advantage of a lab experiment is therefore its high internal validity (i.e., a causal relation between two variables is properly demonstrated). However, since external factors are eliminated, the external validity (i.e., the extent to which the results can be generalized to other situations and to other people) is often low. In consumer research, lab experiments are often conducted with students and by using computers.
Physiological responses - Relatively new in consumer science is the use of physiological responses (i.e., heart rate, blood pressure, muscle activity, eye movements, and brain activity) as dependent variable in experiments. This allows researchers to study the impact of certain stimuli (e.g., products, new technologies) or emotions on human’s physiological system. Advanced technologies (like fMRI, EGG, skin conductance techniques, etc.) increasingly enable researchers to measure these physiological responses.
Field experiments - Field experiments are experiments conducted in a real-life setting. Specific target groups are studied in their own environment, without knowing that they are subject to an experiment. Field experiments, like laboratory experiments, generally randomize subjects into treatment and control groups and compare outcomes between these groups. In contrast with lab experiments, a big advantage of this method is that external validity of this type of research is relatively high. Consumers’ natural behaviour can be studied in a naturally occurring environment. On the other hand, researchers have less control in field experiments, and it is more difficult to proof a causal relationship between the experimental treatments and the outcome variables.
Quasi experiments - One step further than field experiments is the quasi experiment. In a quasi experimental design also real-life behaviour is studied, but lack the elements of random assignment and the use of a control group. This implies that with quasi-experimental studies, it becomes more difficult to demonstrate a causal link between the independent and dependent variables. However, quasi experiments are useful when randomization is impractical and/or unethical. And the external validity of these experiments is high. A natural experiment is a quasi experiment in which consumers are exposed to treatments that are determined by nature or by other factors outside the control of the investigators.