Category:The axis set



The axis set

Axis set is a property attached to the data cache specification; it contains information for automatically generating mathematical expressions when creating diagrams and reports, both through the Quick Report, Cross Diagram, as well as the upcoming Tabulation Module.

An axis set is a set of axes, where each axes represents information related to data in the system (typically a question) and details of how this data should be shown when used in a diagram.

The data cache specification just provides us with information of what columns exist in our data table by one name for each column, but it does not tell us how this information should be interpreted or presented in a diagram context.

The axis set is the place where we can specify how we wish to represent the data from the data table in case a user wants to analyze that data. A typical basic scenario is, that one column in the data cache specification represents a question from a questionnaire, e.g. “Gender”. Let us onwards call the column “Gender”. When extracted to the data table the Gender column will consist of a lot of 1s and 2s. We can then look to the question to inform us about, what 1 and 2 represents and find out, that 1 means “Man” and 2 means “Woman”.

Catglobe does although not look up the information from the question at the immediate time of showing a diagram. There are a number of reasons for that (and thus why a different representation of the data is needed) which we will try to explain below, so that you may understand why we require users to set up axis sets before doing data analysis of the data of their data cache specification.

• We may not wish the diagram title and the column names to be exactly the same as what was stated in the question’s question text and answer option texts. We may e.g. not want the title of a diagram to be “What is your gender? This question is required and you must answer to continue.” This text is too clumsy and would not look good in a diagram, whereas it is very suitable for the text in a questionnaire. We would rather want the diagram title just to be “Gender” and the columns therein to be “Man” and “Woman”. By creating an axis set we enable the user to edit how any question text is represented in diagrams!
• If we only were able to have data represented as they were asked from questions, we would not be able to create different representations based on what we really want to make analysis on. If in the “Gender” question we had an answer option called “I do not wish to answer”, then perhaps we do not want to show this in the diagram we are making, but just those who answered either “Man” or “Woman”. By creating an axis set we enable the user to edit which answer options to include in diagrams!
• We may for many questions wish to represent the data differently than how their answer options looked. If we e.g. had a question “What is your preferred drink?” with answer options 1: Beer, 2: Wine, 3: Soft drinks, 4: Water, we may want to represent it like A: Alcoholic beverage, B: Non-Alcoholic Beverage. By creating an axis set we enable the user to sum up more answer options into one for different representation in diagrams!
• We may also have numeric questions which do not have answer options in the original questionnaire. If e.g. we had a numeric question asking for age, we might want to represent the answers in age groups in our diagram, e.g. “Young people aged 0 to 18” and “Mature people aged 19 and up”. By creating an axis set we enable the user to represent data of all formats in diagrams!
• We often also want to create representations of data which come from more than one question. We might e.g. want to represent Gender and age in 1 diagram based on 4 columns, e.g. “Young men aged 0 to 18”, “Young women aged 0 to 18”, “Mature men aged 19 and up” and “Mature women aged 19 and up”. By creating an axis set we enable the user to merge more question answer options into one diagram!
• Typically, if you have a question like “Which of the following soft drinks do you drink” (let us call the question “DRINK”) with options “Coke” and “Pepsi” as only options, we can calculate percentages who drink Coke as ((No. of coke drinkers)/(No. of people answering the question “DRINK”)*100). But this really only gives us the number of people who answered the question “DRINK” as the base for the percentage calculation. Prior to this question there may have been another question like “Do you drink soft drinks?” (let us call the question “SCREEN”) leading us to this question if the answer was yes. If we instead use this as the base for our calculation of percentage, i.e. percentages who drink Coke is ((No. of coke drinkers)/(No. of people answering the question “SCREEN”)*100) then we get a different percentage. Both calculations are relevant for different scenarios and by having the axis set we can predetermine what diagrams that are automatically generated will use as base.
• We have data in our system which does not come from questions originally. This is data from users, answer sheets, groups, customized columns and more. By having the axis set we can specify how they should be shown when it comes to options, base calculations, and mathematical calculations. By creating an axis set we enable the user to add extra information onto columns which come from other data sources than questions!

In short, the axis set gives us the flexibility to pre-define how data is to be presented in reports and gives us a lot of flexibility to help pre-define this! Let us learn a little more on how the axes of an axis set look!