SFU's Introduction to Vividata
UFV's Using Vividata
Use this template and replace the name (and year, if applicable) of the data set you are looking at:
Vividata. (2019). Diet/Health - Views - I indulge in my cravings for sweets. Retrieved from https://vividata.ca/
This page walks you through how to create and read a data table in Vividata using the following sample question.
Split = Who do you want to learn about (target population)? This is displayed in the columns (vertical) in blue. The term is called split because as you add variables they split the entire population of Canada up so you are only seeing data about your target population.
Question = What do you want to know about the target population? This is displayed in the rows (horizontal) in orange.
Step 1: Choose variables
Explore the variables in the left menu by using the keyword search bar or by clicking on an item to expand the category. In the example below, Travel has been expanded.
Step 2: Add variables as split or question
Hover over a variable in the left drop-down menu to:
The View answers option allows you to see all of the response choices, which can help you choose the most accurate variables.
You can also use filters to narrow your target population. Filters remove survey respondents who are not qualified or experienced enough to answer a subsequent question. For example:
Subsequent questions will only be asked of those people who have answered "yes":
When you are satisfied with the variables you've added as splits and questions, press Generate.
The following data table is from our sample topic and associated variables.
In the count column, read all numbers (except for unweighted base) with 000 at the end.
-- Vacation Trips (all Canadians 14+) 15,915 = 15,915,000
-- Vacation Trips (British Columbians 14+) 2,577 = 2,577,000
Use "Of [split], x have [question variable]" to frame how you describe the table data.
-- "Of [British Columbians aged 14+], 2,577,000 have [taken an overnight vacation trip in the past 12 months]."
The unweighted base is the total number of people who identify in your split (target audience). The unweighted base is the only number in the count column you to not add 000 to the end.
The weighted base adjusts the number of respondents (unweighted base) so they are proportional to the actual base (in this case, all Canadians 14+).
Practice: Interpret the numbers for British Columbians using the chart above and the text template below:
Of the [unweighted base] respondents, who represent [weighted based] British Columbians who are 14+, [percentage] have taken an overnight vacation trip in the past 12 months. This [percentage] represents [count] British Columbians 14+.
Tip: Just because you can add it as a split or question, it doesn't mean that there is enough data available for statistically sound results. For example, if your unweighted base is very small (15), it's not reasonable to extrapolate that those 15 people are representative of 30.3 million Canadians (if that's what your weighted base is). If you try to generate a table where there is insufficient data, Vividata will warn you by highlighting the problematic cells in the table.