Statistical Weighting

When you want to be certain that you’re sample is representative for the population you’ve studied you can use a technique or procedure called statistical weighting. If you’re looking for a representative sample, it has to be of the same composition as the population that you’re studying.

Example: imagine a population where 40% of the people are above 40 years of age and 60% are below 40 years of age. If your sample has a composition of 60% above 40 years and 40% below, it isn’t representative for the population.
However, age might not be all that relevant as a variable, when you’re looking for accurate results. Variables like having experienced the product or service have a much larger influence on the attitudes and behaviour that your population displays towards said product or survey.
Standard demographics like age and gender might very well not be all that interesting. Still, many researchers default to statistics like that regularly and far too easily when weighting a certain data set or applying statistical weights.

Think carefully when you’re considering which variables you’ll use when applying statistical weighting. Picking the ‘wrong’ variable could give you results that aren’t representative for your project at all.

Also look at

Statistical Weighting Methods

When you’re looking for a statistical weighting procedure or application, it is very important to remember that the key to a balanced and representative sample is controlling for known biases in comparison with the target population.

Lees verder >

How to Assign Weights to Variables

Has anyone ever asked you about sampling weights, what they are or how they are calculated? I’m here to tell you what sampling weights is and how it’s done. With sampling for surveys, you can always end up with a sample that is not perfectly representative of your known population. For instance, the age or gender might be off or skewed a bit, and you may have to calculate the weights that fix this bias in your sample.

Lees verder >