Despite not being able to use any of the published interaction scales, we found two items from the Sibling Interaction Scale (Caro & Derevensky, 1997) to be conceptually useful in developing the IIIS: 1) “sibling body position,” which assessed the distance of the interacting sibling from the targeted sibling; and 2) “intensity of involvement,” which assessed the amount of engagement or attention the targeted child directed toward the interacting child. Clearly, the distance between an infant and the person interacting with the infant is an important variable to assess due to the immaturity of the infants’ visual system, particularly during the first few months (Courage & Adams, 1990; Slater, 2001). If the infant is unable to clearly see an individual’s face, that experience is unlikely to contribute to the development of face processing expertise. Distance is an important consideration when determining how close infants should be to facial stimuli when designing empirical studies— younger infants sit much closer to the facial stimuli than older infants (e.g., compare Blass & Camp, 2004 to Rochat & Striano, 2002). Additionally, for our research questions, it was important to know the amount of time that the individual interacted with the infant as well as the infant’s attention toward that individual’s face to get a sense of how much real-time experience infants had with that person’s face. Attention toward stimuli as measured by amount of looking is a commonly used measure in infant face perception studies (e.g., Kelly et al., 2005, 2007; Langlois et al., 1987; Quinn et al., 2002). Therefore, the IIIS contained three items, each with four options, for the caregiver to note: 1) the individual’s estimated distance from the infant; 2) the amount of time that the individual interacted with the infant; and 3) the infant’s attention or interest toward the individual. It also contained places to note the approximate age, sex, and race of the individual interacting with the infant (Figure 1).

To determine discrepancies in infants’ exposure to same- and other-race faces, mother and father’s face, female and male faces other than mother and father, and same- and other-age individuals, we determined infants’ average PTFE for each of these categories of faces for each infant age group (see Table 3). Based on these percentages, we calculated discrepancy ratios that represented how much more times experience infants had with the more familiar types of faces (i.e., same-race, mother, other female, and same-age faces) relative to the less familiar types of faces (i.e., other-race, father, other male, and other-age faces). The discrepancy ratios are plotted in Figure 3.

* Interest is the percentage of time the infant attended to the individual's face (1 = less than 25%; 2 = 25-50%; 3 = 50-75%; 4 = over 75%)

Upon completion of the development of the IIIS, we developed a more condensed version of the scale (the ICFMIS) that would simplify the recording of data for individuals who frequently interacted with the infant (e.g., parents, siblings, daycare workers). The ICFMIS made the task of recording frequent interactions more manageable. For the ICFMIS, we combined the corresponding points on each IIIS scale item to develop four categories: fleeting involvement; brief involvement; moderate involvement; and high involvement (Figure 2). Caregivers simply noted the frequency of those interaction types. Although this collapsing of the three items into one category resulted in some information loss, the purpose was to make the scale user-friendly for caregivers and avoid a potentially higher participant dropout rate.

* Time refers to the length of the individual's interaction with the infant (1 = less than 10 seconds; 2 = 10 seconds to 1 minute; 3 = 1 to 5 minutes; 4 = over 5 minutes)