We are searching data for your request:
Upon completion, a link will appear to access the found materials.
Ordinal: aims to create as its name says, an order of value, according to the preference of the respondent. For example, on a scale, A is preferred to B, but it is not identified how much A is less than B.
Nominal: are the most common scales in marketing research. Their numbers serve to identify the respondent's choice and not to determine order or even if A is better than B. Numbers are associated with response points to create an organization on the scales. It is a classic example of nominal scale, gender and dichotomy (yes; no) and semantic differential (pure impure).
Interval: These are questions that aim to compare ranges and measure how far one preference is from another. Currently they are the subject of endless discussions between statisticians and marketing academics when applying statistical tests, after all they are considered discrete but can go through a process of approximation and become continuous.
A similar process is described by Cunha (1997), when the author approaches the Correspondence Analysis (CA) technique and comments that the best employment variables for this technique are the qualitative ones or those that have been categorized. Example of interval scales: 1, 2, 3, 4, 5, very dissatisfied; dissatisfied; indifferent; pleased; very satisfied.
Reason: are the continuous variables. Weight, age, income, are examples of questions of reason.
Below are the possible statistical models for the types of scales addressed.
Scale Type | Possible statistics |
Ordinal | All of central tendency |
Nominal | Fashion and Chi square |
Interval | Means, standard and mean deviation, amplitudes, variance, z and t test, correlation and regression. |
Reason | All of the previous |