Abstract:

People recognize familiar faces more quickly and accurately than unfamiliar faces. Previous studies found that both holistic processing and part-based processing play a role in recognizing familiar faces efficiently. However, familiarization with faces may not happen and be accomplished in a short period of time for both types of processing. Few research has focused on the progress of familiarization with individuals' faces in real life: Did it reach a ceiling-level of recognition in a short period of time (sudden change) or over a long period of time (gradual change)? And what role did part-based processing and holistic processing play in the recognition improvement during face familiarization?

In the current study, we recruited 57 undergraduate students, 30 from one class (Group 1: 20 females, M _age = 18.4, SD = 0.8) and the remaining 27 from another class (Group 2: 18 females, M _age = 19.2, SD = 0.8). They got along with their classmates for 1 month (Group 1) and 13 months (Group 2) of daily interactions, respectively. Three delayed-matching tasks (Whole-face Learning, Regional-part Learning, and Regional-periphery Learning) were used to test the participants' face recognition performance in part-based processing and holistic processing. The “Regional-part” was an eye or the mouth, respectively. The “Regional-periphery” was a face with an eye masked or the mouth masked. In each trial, a fixation point (800 ms) was presented at the center of the screen, then the learning stimulus (500 ms) was presented with the specific part (eye or mouth) at the fixation point location. After that, four whole faces (5000 ms) were displayed on the screen after a blank (1000 ms). The participants were asked to choose the face with the same identity as the learning stimuli.

Results showed: (1) The recognition performance of familiar faces improved with the increase in daily-interaction time. The familiarity effect (the difference between recognizing the classmates and strangers) of participants with 13 months of daily interaction (M = 7.9%) was higher than that of participants with 1 month (M = 3.2%; F(1,55) = 7.98, p= 0.007, η_p² = 0.13). (2) The recognition performance of familiar faces (Regional-part: M = 80.1%; Regional-periphery: M = 92.4%) was higher than that of unfamiliar faces (Regional-part: M = 75.9%; Regional-periphery: M = 85.5%) in the Regional-part Learning task (F(1,55) = 22.94, p< 0.001, η_p² = 0.29) and in the Regional-periphery Learning task (F(1,55) = 61.90, p< 0.001, η_p² = 0.53), which indicated that both the part-based processing and holistic processing were enhanced by familiarity. (3) With the increase in face familiarity, the performance of the eye region and mouth region showed regional asynchronization. Participants with 1 month of daily-interaction time showed no familiarity effect in recognizing the region of the eye (familiar: M = 75.7%; unfamiliar: M = 72.4%; F(1,55)= 3.78, p= 0.057, η_p² = 0.06), but a significant familiarity effect in recognizing the peripheral region of the eye(familiar: M = 91.7%; unfamiliar: M = 88.4%; F(1,55) = 7.57, p= 0.008, η_p² = 0.12); and they showed significant familiarity effects both in recognizing the region of the mouth (familiar: M = 84.7%; unfamiliar: M = 78.5%; F(1,55) = 12.39, p= 0.001, η_p² = 0.18) and the peripheral region of the mouth (familiar: M = 92.0%; unfamiliar: M = 83.9%; F(1,55) = 20.40, p< 0.001, η_p² = 0.27). Participants with 13 months showed a familiarity effect in recognizing the region of the eye (familiar: M = 78.3%; unfamiliar: M = 71.1%; F(1,55) = 16.17, p< 0.001, η_p² = 0.23), and a stronger familiarity effect in recognizing the peripheral region of the eye (familiar: M = 93.5%; unfamiliar: M = 86.4%; F(1,55) = 29.76, p< 0.001, η_p² = 0.35); and they showed no familiarity effect in recognizing the region of the mouth (familiar: M = 81.8%; unfamiliar: M = 81.4%; F(1,55) = 0.04, p= 0.837) but a familiarity effect in recognizing the peripheral region of the mouth (familiar: M = 92.4%; unfamiliar: M = 83.3%; F(1,55) = 23.05, p< 0.001, η_p² = 0.30). (4) For unfamiliar faces (r = 0.412, p = 0.024; r = 0.439, p = 0.022) and familiar faces with 1 month (r = 0.428, p = 0.018) and 13 months daily interaction (r = 0.609, p = 0.001), the performance of recognizing the peripheral region of the eye was positively correlated with the performance of recognizing the whole faces. For familiar faces with 13 months of daily interaction, the performance of recognizing the region of the eye was positively correlated with the performance of recognizing the whole face (r = 0.562, p = 0.002).

The results implied that familiarization of face identity was gradual. Familiarization promoted part-based processing and holistic processing. The improvement of the holistic processing preceded that of the part-based processing. In addition, the improvement in recognizing the eye region and the mouth region showed regional asynchrony during familiarization. Familiar face recognition has eye region specificity. Holistic processing in the eye region plays a key role in the establishment of memory representations of familiar faces. These results suggest that Regional Asynchrony and Eye Region-specificity emerged during the familiarization. New knowledge for understanding the mechanism of face memory representation formation was provided.

Key words: familiar face, daily interaction time, eye region, holistic processing, part-based processing