1.4 Information bias

Lastly, separate from sampling, bias can arise in how we record or measure observations. Some of these biases are types of cognitive biases that and others are statistical biases. Here are a few types of information biases that can occur:

• Response bias/Self-report bias/Social desirability bias: Bias occurs when the recorded response does not accurately represent the true value for the individual due to wording of the question, ordering of the questions, format of response, or to increase social desirability. Most people like to present themselves in a favorable light, so they will be reluctant to admit to unsavory attitudes or characteristics (e.g. weight, income, alcohol consumption, mental illness) or illegal activities in a survey, particularly if the survey is completed in an interview setting and/or the results are not confidential. Additionally, this could arise if the possible responses do not allow for accurate reporting (e.g. gender identity, race).

• Recall bias: People often unintentionally make mistakes in remembering details about the past. If the study design is retrospective in that it requires units to rely on their memory, we may get bias in the information collected.

• Measurement error: Technologies that measure variables of interest may not always be accurate and human calibration of those instruments may be off as well.

There are many other forms of bias, many of which are studied by cognitive and social pscyhologists. When designing questionnaires and surveys, it is important to keep these in mind. A researcher must first decide what specific information they want to collect and then figure out the best way to accurately collect that.