Categorical variable
A categorical variable labels or categorizes certain characteristics or traits without establishing a numerical value or order. Categorical variables are often used to represent qualities or attributes that are not numerical. The two main types of categorical variables are:
Nominal Variables: These are used to label or name categories without any quantitative value or inherent order. Examples include gender (male, female, non-binary), types of products (beverages, electronics, clothing), or brand names. Each category is distinct, with no implied order or numeric equivalence among them.
Ordinal Variables: These have a set order or ranking, but the ranks' differences are not necessarily equal. For example, customer satisfaction ratings (such as satisfied, neutral, dissatisfied), levels of education (high school, bachelor's, master's, doctorate), or socio-economic status (low, middle, high) are ordinal variables. They indicate a hierarchy or order, but the intervals between the categories are not quantifiable.
Numerical variable
A numerical or quantitative variable represents data that can be measured and expressed numerically. There are two types:
Discrete Variables: These are countable numbers, typically integers. Examples include the number of products sold or the number of employees.
Continuous Variables: These can take any value within a range and are often the result of measurement. Examples include height, weight, or temperature.
Users can create ad hoc breaks or intervals from numeric variables. Read more about Segmentation of Numeric variables.
Note! DataTile classifies variables as numerical if no value labels are defined in the SPSS file, disregarding the formally declared type.
Date variable
Date variable is a specific variable used to represent dates and times. DataTile converts all dates into internal timestamp representation that allows further processing and handy use such as segmentation.
You can segment date variables when working in a crosstab, in this case, you can create temporary custom intervals using special expressions.
Alternatively, you can derive a new date variable in Meta-Editor, creating time series or seasonal intervals.
Linear Time Series
It is derived from a Date variable and is a categorized sequence of fixed intervals, such as weeks, months, quarters, etc.
Seasonal Time Segments
It is derived from a Date variable. Seasonal categorization involves structuring date into regular intervals to examine patterns within a periodic cycle, like months or quarters of a year. To capture seasonality, DataTile creates categorical variables with selected standard periods.
Time series and Seasonal intervals become permanent variables, saved in a codebook.
Text variable
A text or string variable stores textual data of any kind. DataTile supports any encoding. If encoding is not declared in the original dataset, UTF-8 is assumed.
Weight variable
Weight variables are supposed to be provided as numeric variables in datasets loaded into DataTile.
Such variables should be explicitly declared as weights to enable DataTile to use them in statistical calculations.
A weight variable is used to adjust the survey results so that they more accurately reflect the overall population. This weighting compensates for differences in the probability of selection among respondents or to correct for imbalances in the sample relative to the population.
Click here for more information on how to set a weight variable.
Probabilistic variable
Probabilistic variables play a crucial role in the Reach & Frequency algorithm, which is fundamental to media planning. They are key in quantifying and representing statistical probabilities, especially in the context of media and publication analysis. Typically, these variables assess the likelihood of a publication's reach, often quantified using metrics like Average Issue Readership (AIR).
Probabilities are provided as a numeric variable in data, and they should be marked as such explicitly to enable DataTile to treat them accordingly in statistical calculations.
Multi-response variable
A multi-response set, or MR set for short, is a construct defined on top of a set of several other numerical or categorical variables to represent questions with multiple answers.
DataTile supports well-known dichotomous and categorical MR sets as well as other specific quantitative questions.
Read more about types of MR-sets in DataTile.
Logical variable
Users can create variables with any logical categories, defined from other variables. It is possible to create logical codes of any complexity using AND, OR, ANY operators, groups, and negation.
Click here for more information on how to create logical variables.