|Related Online Training modules|
|Setting a Weight in the Data Tab|
|Setting a Weight in the Variables and Questions Tab|
|Creating a Weight with 1 Variable|
|Creating a Weight with 2 Variables|
|Creating a Weight by Wave|
|Generally it is best to access online training from within Q by selecting Help : Online Training|
Weighting, also known as sampling balancing, is used to make adjust the results of studies to make them more representative (e.g., if a study has 20% men, but the population has 50%, weighting can be used to bring the results of the study into line with the population).
Some additional information about weighting is displayed below on this page. Also see:
In addition to the drop-down, the presence of weighting is also reflected in the sample size description at the bottom of the screen.
base n = from 151 to 215; total n = 223; 72 missing; effective sample size = 156 (70%); 80% filtered out
This description indicates that :
- The sample size varies for different cells in the table, the cell with the smallest sample size being 151 while the largest cell has a sample size of 215. (We can see each cell’s sample size by selecting Base n from Statistics – Cells (see Statistics).
- The total sample size of the table is 223 (i.e., if there were no missing data, each cell would be based on a sample size of 223).
- Each cell has up to 72 missing values (i.e., 223 minus 151).
- The Effective Sample Size is 156 (this is computed on the largest base n of 215). In most instances, weighting causes a decrease in the statistical significance of results. The effective sample size is a measure of the precision of the survey (e.g., even if you have a sample of 1,000 people, an effective sample size of 100 would indicate that the weighted sample is no more robust than a well-executed, un-weighted, simple random sample of 100 people).
- The weighted sample size of 223 has the same precision as an un-weighted sample of 70% the size (i.e., 156 / 223) for this table.
- The percentage shown in brackets indicates the ratio of the effective sample size to the total sample size. A score of 100% indicates that the weighting has caused no loss of precision.
- The final statistic (80% filtered out) shows the proportion of the total number of observations in the study that is not available for the analysis due to a filter having been applied to the table.
Weights, sample size and population
In commercial market research, weights generally have two roles: to correct for bias in the sampling (e.g., over-representation of a particular segment) and to gross up the sample size to match known population totals (e.g., the number of people in a country). Q accommodates both of these roles. When a weight is applied, all percentages and averages automatically take the weight into account. Estimates of the population size can be obtained by selecting Population from Statistics – Cells, Column Population in Statistics – Below and Row Population in Statistics – Right.
- Setting Up Your Data in Q for an overview of how to set up data in Q
- File Formats Supported by Q
- Manipulating data files
- Basic Workflow For Checking and Cleaning a Project
- Constructing Variables to Make Analysis Easy
- Setting Statistical Assumptions When Setting Up Projects
- Advanced Data Tidying
- Exporting, Copying and Printing
- Updating Projects with New or Revised Data
- Converting Other Files Types into SPSS or CSV Data Files
Pages in category ‘Weighting’
The following 24 pages are in this category, out of 24 total.
- How To Apply Different Weighting Structures To Different Time Periods
- How To Apply Different Weighting Structures To Sub-Samples
- How to Cap Weights
- How to Combine Categories in a Pick One Question for Weighting
- How to Construct a Weight for a Self-Weighting Sample
- How To Create Expansion Weights
- How To Force Weighting Targets To Add Up to 100%
- How to Make Merged Categories Appear in Targets when Weighting
- How To Set The Weighted Sample Size
- How To Trim A Weight
- How to Weight Only a Part of a Data File