How does our ability to detect faces change as we grow up?
As children grow from ages 5 to 10, they develop a selective visual preference for horizontal details when identifying faces, but not when identifying houses.
Source
A face detection bias for horizontal orientations develops in middle childhood
What they did
To study visual development, researchers tested 43 children between 5 and 10 years of age using a rapid touchscreen classification task. Participants viewed grayscale images of faces and houses that were digitally filtered to show only horizontal patterns, only vertical patterns, or both. Each child completed 144 total trials to determine if age relates to how quickly they process horizontal visual cues compared to vertical ones.
What they found
The researchers excluded data from 1 child due to low accuracy, leaving 42 children for the final response time analysis. Although overall accuracy exceeded 90% across all groups, older children showed a faster response to horizontal face features than vertical ones. This was demonstrated by a significant positive correlation of 0.32 between age and horizontal bias for faces, whereas the correlation of 0.096 for houses was not significant.
The limits
What it doesn't show
This study does not establish whether this emerging horizontal bias is driven by biological brain maturation or by the sheer volume of real-world experience children gain with faces. Additionally, the experimental task relied on static, grayscale, front-facing images, meaning the findings might not fully capture how children process moving or color-rich faces in everyday natural settings. Finally, the relatively small sample size of 42 analyzed participants limits the generalizability of these developmental trajectories to broader populations.
Key terms
- information bias
- The selective reliance on a specific subset of visual features, such as certain spatial orientations, to accomplish a recognition or detection task.
- horizontal orientation energy
- Visual information represented by horizontal patterns and structures, which are highly prominent in features like the eyes and mouth.
- ex-Gaussian distribution
- A statistical model combining a normal distribution and an exponential distribution, often used to fit skewed response time data.
- second-order configuration
- The spatial relationships and distances between individual features of a face, such as the space between the eyes.
- holistic processing
- A cognitive mechanism where a face is perceived as a unified whole rather than a collection of independent features.
- biological bar-code
- A computational concept describing how the alternating horizontal bands of light and dark across facial features create a unique signature for face detection.
Flashcards
Want these cards to stick?
Save the deck to NoteFren and study it with spaced repetition.
Quiz yourself
What is the main objective of this study regarding face processing in middle childhood?
Common questions
Why are horizontal lines so important for recognizing faces?
Facial features such as the eyebrows, eyes, and mouth naturally align horizontally. This arrangement creates a distinct, barcode-like pattern of dark and light bands that the human visual system relies on to identify faces quickly.
Why didn't the study find a horizontal bias for houses?
Houses contain a mix of vertical and horizontal structures, like walls and roofs, that do not have the same biological configuration as human faces. The lack of correlation for houses suggests that the developing horizontal bias is specific to social, face-related visual stimuli.
How was response time measured and analyzed?
The researchers used an ex-Gaussian distribution to model response times. This method combines normal and exponential curves, helping to manage delayed responses and outliers without losing valuable behavioral data.
Did the age of the children affect their categorization accuracy?
No, accuracy was exceptionally high (averaging over 90%) across all age groups. The developmental differences were only apparent when looking at how quickly children of different ages responded to different visual filters.
More on Face processing