Chernoff Faces

Visualization Type

Values The main purpose of Chernoff Faces is to display multiple variables at once each dictating a part on the human face (ears, hair, eyes, and nose) based on numbers in a dataset.

Encourages Since we read the human face, we notice the small differences within the data. In addition, we end up walking away with a certain image that remains in the back of our minds.

Discourages Because the data alters the humanistic features, some people misinterpret the faces if they do not have a key.

Pre-processing An algorithm is used to translate the categories of raw data into a specific part on the human face.

Mapping Each category will be converted a range depending on the part of the human face. For example, the mouth can change from a frown (negative) pokerface (zero) or a smile (positive.)Usually the human features with a "scarier" look means a more negative response.

Images Generally, good examples should at least have a key or the information will be misread.