OBSERVED EMOTION FREQUENCY
s1
Supplementary Material
Expanded Method: Positive and Anger Emotion Frequency and Intensity
Positive and anger emotion frequency and intensity. Observers scored children’s
positive and anger emotions in classes, recess, and lunch in the fall semester, two to three times a
week (approximately 9-12 weeks). Each child was observed by two or three different coders.
Observers had a list and corresponding picture collage of participants for each class and rated (0
= no evidence; 3 = strong evidence) children’s positive (e.g., expressions of happiness, joy,
excitement, amusement, pride, positive anticipation, awe) and anger (e.g., expressions of anger,
frustration, annoyance, irritation) emotion after observing for 30 s (generally, children were not
coded again until the entire list of present children was coded; Mtime-coded = 64 min, range: 16-133
min). There were eight cases that had a number of observations on the higher end (above 117
min); however, results in subsequent analyses remained the same with or without these eight
cases. Ratings were coded on a scale from 0 to 3 (0 = no evidence of emotion; 1 = emotion seen
once, small in intensity and brief (< 3 seconds); 2 = two indicators of emotion that were small in
intensity and brief; one indicator of emotion that was small in intensity but lasting 4-9 seconds;
or one indicator that was medium in intensity and lasting less than 5 seconds; 3 = three or more
indicators that were small in intensity and brief; two or more indicators medium in intensity; one
or more indicators, small in intensity and lasting more than 10 seconds; one or more emotional
displays medium in intensity lasting more than 5 seconds or any high intensity indicator).
Positive emotion and anger/frustration were coded based on the frequency, intensity, and
duration of multiple behavioral domains. The facial criteria for positive emotion included the
following: smiles, grins, cheeks raised, and eyes squinted (in an intense smile) or wide and bright
(when excited or joyful). Behavioral and vocal criteria for positive emotion included jumping up
OBSERVED EMOTION FREQUENCY
s2
and down, chest and head up, upbeat tone, laughter, and giggling or excited squeals. Facial
criteria for anger included pursed or pressed lips, jutted jaw, wide eyes with ‘hard’ stare,
furrowed brows and/or forehead. Behavior criteria for anger included slamming hand/fist on
something or roughly in the air, knocking things over, stomping, shoulders and chest square and
tall, arms stiff and in low V, fists curled, as well as vocal content, tone (e.g., loud, harsh/cutting,
short/sharp intonation), and vocalizations (e.g., screaming in anger, growling menacingly,
grunting or roughly sighing).
Prior to observing child interactions in participating schools, observers received several
weeks of training, which included rating child interactions in pre-coded videos and/or in pilot
preschool settings. Bi-weekly checks were made for agreement with the coding supervisor.
Reliability ratings were obtained from a set of pre-coded videos (which were used for reliability
purposes starting in the second year of the study) and randomly selected live scans,
simultaneously rated by a second observer (Totaltime = 1,907 min) in the fall (ICC = .96
[positive], .88 [anger]).
For each child, emotional frequency was operationalized as the number of instances each
emotion occurred regardless of its intensity (a score of at least 1 [minimal evidence]) divided by
the total number of scans per child for the given emotion (Mpositive-frequency = .41, range: .12-.78;
Manger-frequency = .02, range: .00-.15) (similar to Fabes et al., 2002). To assess emotion intensity,
observers’ codes for a given child were averaged across all observations for each emotion with a
score ≥ 1 —when at least minimal emotion was observed (Mpositive-intensity = 2.25; Manger-intensity =
1.64). That is, observations scored a value of 1, 2, or 3 were averaged for intensity.