Poster #53 - A New Mother Vocal Affect Scale for 4-month Mother-Infant Face-to-Face Interaction

• In Event: Poster Session 05

Abstract

In prior research on mother-infant face-to-face interaction at 4 months, using second-by-second microanalysis of split-screen video, infant vocal affect has functioned differently from infant facial affect. Infant vocal distress (but not facial distress) was associated with disorganized attachment at 12-months. Moreover, simultaneous positive facial affect and negative vocal affect within the same second was associated with disorganized infant attachment at 12 months (Beebe et al., 2010). Finally, infants of mothers who were pregnant and widowed on 9/11 were more likely than controls to switch from a negative affect in one channel (e.g., facial affect) to a positive affect in another channel (e.g., vocal affect) in the next second (Beebe et al., 2020).

Despite these interesting findings, no complementary video microanalysis mother vocal affect (MVA) code exists in the literature. Reviewing the literature, (a) some researchers assessed direction and function of maternal speech (e.g., “infant directed” vs. “other directed; “naming or imitation”) (Lopez et al., 2020); (b) other researchers utilized a broad affective coding scheme, for instance, labeling speech as “positive” or “negative” (Morelen et al., 2016), or “fearful” “happy” or “neutral'' (Mumme, 1993; Mumme et al., 1996); (c) still others combined MVA with other modalities, such as gaze and facial affect, into a larger “maternal affect” measure (Cohn & Tronick, 1987). For example, Tronick et al.’s (1980,1985) Monadic Phases System and Manual combined complementary modalities for mother and infant (e.g., gaze, head orientation, affect) into “monadic phases,” such as “avert” and “social play.”

Our study aimed to develop an MVA code that would code maternal vocal affect in a more detailed way and complement the Beebe et al. (2008, 2020) second-by-second microanalytic coding of IVA, which has 6 ordinal levels: (1) Cry, (2) Angry/Protest, (3) Fuss/Whimper, (4) No Vocal Affect, (5) Neutral/Slightly Positive, (6) High Positive.

From an existing data set of 4-month mother-infant face-to-face interactions, we (a) reviewed N=13 randomly selected videos to establish a working draft of the MVA scale; (b) made several rounds of revisions to differentiate the categories; (c) coded second-by-second the first 2.5 uninterrupted minutes of another randomly selected N=5 interactions; and (d) consulted with the P.I. for final revisions. To establish inter-rater reliability, we randomly selected and independently coded N=5 additional 2.5 min interactions, achieving K ≥ 0.70 for all 5 interactions.

Our MVA scale has 9 ordinal levels of vocal affect ranging from most negative to most positive:

1. Frustration/Irritation/Disapproval
2. Sadness
3. Anxiety/ Desperation
4. Woe/Concern
5. No Vocal Affect
6. Directing/Not Frustrated
7. Neutrally/Positively Engaging
8. Playing
9. Laughing

We identified two additional “discrepant codes” that lie outside the MVA scale and qualify moments of conflicting affects: (1) content vs. tone discrepancy, and (2) positive/negative switch (within 1 second).

The Mother Vocal Affect code provides a new dimension for the exploration of mother-infant interaction. This code, while complementary to Beebe et al.’s established infant vocal affect scale, offers the potential for unique insight into maternal affective experience as it relates to risk factors such as maternal depression or insecure attachment outcomes.

Authors

• Michaela De Filippis, New York State Psychiatric Institute
  Presenting Author

• Maya Raphaella Shustik, New York State Psychiatric Institute

• Beatrice Beebe, New York State Psychiatric Institute