Market researchers are often interested in population estimates. If a survey shows that 10% of respondents fall into a particular category they might be interested in how many people that proportion represents. A simple approximation can be made by multiplying the percentage of people who fell into the category by a target population number that the dataset is said to represent. Crunch lets you set the target population for a dataset, and see variable summaries and calculations in terms of that population.
When a dataset has a target population set, the display controls allow you to select POP. When you select that option all cards, analyses, and multitables will show the population estimate rather than the raw survey numbers or percentages.
Time series considerations
The population is assumed to be constant over time.
In a normal crosstab, the population is assumed to be distributed across all cells, which is not true of a survey wave variable (categorical variable with dates). Instead, each wave is assumed to represent the entire population.
Setting the target population
To set the target population, go to the dataset properties menu:
You can enter a Target population size and an order of magnitude for the population. For example to create a target population of 9 million, enter 9 in the target population, and select Millions from the drop-down. Make sure to save the changes you made to the dataset properties.
