Overgeneralization
In the scientific field, overgeneralization refers to the process of applying conclusions about one population in a specific set of circumstances to another population in a different set of circumstances (4, 5). Essentially, observations and what is assumed to be true for some cases are concluded to be true for all cases (6). For example, below is one way we can easily visualize overgeneralization.
_edited.png)
Overgeneralization can be a serious problem for the scientific community as it is often employed by media sources through usage of different wording to catch people's attention which may lead to distortion and misunderstanding the scope of a scientific result (4). It is always important to refer back to the original study on certified scientific sites and check the population scope of the results instead of automatically assuming that the results are representative of different groups and apply to the population as a whole (4, 6).
Remember! Some generalization is often appropriate (e.g., sampling a small area for weeds, finding it has 40% weeds, and deciding most but not all of the field must be covered by 40% weeds), but overgeneralization is often not.
Let's Look at a Case Study Example from : "Does Genetic Signature Exist for Each Type of Cancer?" (4)
Hover over each of outlined boxes to go through the case study!
In 2005, researchers conducted a study which discovered a set of genes linking the spread of breast cancer to the lungs. As the findings were preliminary, researchers insisted that more research needed to be conducted before the link could be considered empirical evidence for the presence of genetic signature in the spread of breast cancer to lungs (Brechman, 2009). What was the issue? *Hover to find out*
Unfortunately the media reported the findings of the study to the public in a completely different light. The media declared the findings to be proof of a genetic signature for “each type of cancer and the organ it spreads to.”
This single sentence represents overgeneralization in the media in various ways. Continue below to find out how!
Issue #1
"The findings from the researchers did not prove the existence of genetic signature in metastasis. The media exaggerated the significance of the findings, thereby distorting the public’s perception of the finding" (4).
"The study only involved research on breast cancer, not every type of cancer as the media has generalized" (4).
Issue #2
"The study did not look into breast cancer metastasis to any organs other than the lungs. The media overgeneralized the results by equating breast cancer to cancer in general and to all body organs, widening the scope of the study to completely inapplicable fields" (4).
Issue #3
As seen in the above example, overgeneralization can exist in science too, where scientists unintentionally and unknowingly generalize their conclusions without sufficient evidence such as assuming results from smaller sample sizes are applicable to broader populations and that what is true in one circumstance will also be true or in other similar circumstances (5, 6, 7). This is also known as induction. As seen most commonly in behavioural science, researchers will take samples from the Western, Educated, Industrialized, Rich, and Democratic (WEIRD) populations and assume results found are representative of the whole population world wide (5).
Additionally, overgeneralization can be found within science when there is a need for results to conform with hypotheses (5). Also known as a failure of modeling, when a mechanism or phenomenon is not fully understood or is unclear, scientists will often propose an inferred model based on their results, which may be heavily biased due to the limited experimental information found (5, 7). Due to unknown underlying confounding variables, overgeneralization can lead to failures of the experiment when other scientists try and replicate it to see if it is sound research (5).
To combat overgeneralization it is important for scientists to have a systematic procedure for selecting representative individuals or groups that accurately define a population (6). Single studies should be conducted in multiple contexts and multiple methods from different fields so that we can better assess a given finding (5). In addition, journal author guidelines can be improved on where a “constraint to generalizability” statement is included in papers where possible (6). This helps raise awareness of overgeneralization and brings more consideration to potential limitations of a study (5, 6).
Let's Check Our Understanding!
Can you think of any other examples of overgeneralization? Comment below!
About Us
Science for Everyone is a Canadian Nonprofit Organization that provides educational resources to help raise the level of scientific literacy in the general population.