🧪 This platform is in early beta. Features may change and you might encounter bugs. We appreciate your patience!
non_response_bias
Non-response bias occurs when individuals who do not participate in a survey or study differ systematically from those who do. This can distort results because the collected data reflects only a self-selected subset, not the full target population. The bias is especially problematic when the reason for non-response is related to the variable being studied.
A workplace satisfaction survey has a 40% response rate. Dissatisfied employees who have already mentally checked out are less likely to respond, making the workplace appear more satisfying than it actually is.
A pharmaceutical company surveys patients who completed their 12-week drug trial to assess satisfaction with the medication. Patients who dropped out early due to side effects are not included, making the drug's tolerability appear far better than it was across the full trial population.
An online news outlet runs a poll asking readers whether they trust mainstream media. Because the outlet's regular audience skews toward media sceptics who are motivated to participate, 74% respond 'No' — a result the outlet then cites as evidence of a broad public trust crisis, ignoring that casual or satisfied readers rarely bother to vote.
Binary (yes/no) questions an LLM must answer to identify this aspect:
Is there a significant portion of the target sample that did not respond or participate?
Type: binaryCould the non-responders systematically differ from responders on key variables?
Type: binaryWas any analysis conducted to compare responders and non-responders?
Type: binaryAre the findings generalized to the full population without accounting for non-response?
Type: binaryNon-response bias occurs when individuals who do not participate in a survey or study differ systematically from those who do. This can distort results because the collected data reflects only a self-selected subset, not the full target population. The bias is especially problematic when the reason for non-response is related to the variable being studied.
People assume that those who responded are representative of the whole group. The reasons for non-participation are often invisible to the researcher, and low response rates are frequently downplayed or ignored in reporting.
Report response rates transparently. Conduct non-response analysis by comparing known characteristics of responders and non-responders. Use follow-up contacts, incentives, or statistical weighting to reduce and adjust for non-response.
Political polling frequently suffers from non-response bias. Certain demographics are harder to reach by phone, leading to skewed predictions. The 2016 and 2020 US presidential polls underestimated support for certain candidates partly due to differential non-response.
Participants who choose to join a study differ systematically from those who do not.
The statistical error of drawing conclusions from a dataset that has been filtered by a survival or success criterion, without accounting for the filtered-out cases. The surviving sample is systematically different from the full population, and conclusions drawn from it are biased.
Systematic exclusion of certain participants from a study distorts results.
Studies with statistically significant or positive results are more likely to be published, while null results remain unpublished. This distorts the published literature and inflates apparent effect sizes in meta-analyses.
Differential accuracy in remembering past events between study groups.
Multiple filtering or inclusion steps systematically alter the composition of a sample.
Use these tools to detect, analyze, or train this aspect.