Mother-infant floor-play interactions were observed in a room approximately 5 m. x 2.5 m. An adult chair and several toys (ball, book, doll, stuffed animal) were placed in the centre of the room. Two videocameras on tripods were situated in opposite corners of the room. A one-way mirror divided the playroom from an adjacent monitoring room, from behind which the experimenter could observe the interaction. Most of the mother-infant behaviour could be monitored from one or other camera.
Prior to their arrival, mothers were informed that their interactions would be videotaped and that the purpose of this procedure would be explained after the experiment was over. When the mothers arrived, they were told: "We would like to observe you and your infant playing together on the floor just as you would at home for about 10 minutes". They were told that the experimenter would inform them of the onset of videotaping. The experimenter entered the room after 5 minutes, and announced that he would start videotaping the session. In actual fact the whole 10-minute session was recorded without the mothers' knowledge.
Following the l0-minute period, the experimenter informed the mothers that the full l0-minute session had been recorded. They were told that they could view the videotape and that if they wished, the section which had been recorded without their knowledge would be erased. None of the mothers requested that any part of the tape be erased, and in fact most actually enjoyed seeing themselves on television.
The videotapes were coded for a variety of mother and infant behaviours, including the mother's proximity to the infant, touching the infant, looking, smiling, vocalizing or laughing at the infant, offering toys, demonstrating the use of toys and engaging in constructive play. For the infant, the following behaviours were coded: touching the mother, looking, smiling or laughing at the mother.
A reliability study was carried out in which an independent observer scored the videotape of one typical mother-infant dyad; his results were correlated with those of the main scorer, and these are shown in Table l:
Table 1: Results of study of inter-observer reliability
| Mother's behaviour | r |
|---|---|
| Mother's proximity to the infant | .70 |
| Touching the infant | .65 |
| Looking | .40 |
| Smiling | .35 |
| Vocalizing or laughing at the infant | .35 |
| Offering toys | .55 |
| Demonstrating the use of toys | .60 |
| Engaging in constructive play | .30 |
| Infant behaviours | r |
| Touching the mother | .70 |
| Looking | .35 |
| Smiling | .30 |
| Laughing at the mother | .35 |
All these correlations were statistically significant, and hence inter-observer reliability was regarded as satisfactory.
The data were then analysed by a series of t-tests comparing behaviour in the condition where the mothers were aware of being videotaped with behaviour in the unaware condition. The results of these analyses are shown in Table 2:
Table 2. Results of t-tests on mother and infant behaviour
( * Results significant at the .05 level; all tests two-tailed)
| Mother's behaviour | t |
|---|---|
| Mother's proximity to the infant | 2.566* |
| Touching the infant | 1.357 |
| Looking | 0.577 |
| Smiling | 0.345 |
| Vocalizing or laughing at the infant | 0.789 |
| Offering toys | 2.978* |
| Demonstrating the use of toys | 2.579* |
| Engaging in constructive play | 0.987 |
| Infant behaviour | t |
| Touching the mother | 1.765 |
| Looking | 0.564 |
| Smiling | 1.113 |
| Laughing at the mother | 2.009 |
The experimenters concluded that: