Issues in the assessment of social competence in children.

Foster et al. (1979) wrote a narrative review. They compared two ways to measure how well children get along with peers.

The first way is sociometrics: kids vote for classmates they like or dislike. The second way is direct observation: adults watch and tally social acts like sharing or joining a game.

The authors listed strengths and weak spots of each tool. They did not run new kids or collect new numbers; they summed up the field at that time.

The review found that sociometrics give a quick class-wide picture but can be swayed by one popular child.

Direct observation shows real behavior but takes lots of time and may miss rare yet important moments.

Neither method alone was judged good enough for clear, fair decisions about a single child’s social skill level.

Cameron et al. (1996) later tested the sociometric claim. They tracked the same kids across weeks and saw that peer votes jumped around. Their data back up the 1979 warning: sociometrics are shaky for single-case decisions.

Cohen et al. (1993) updated the story fourteen years on. They added that most tools still lack “ecological validity,” meaning scores do not predict how kids act on the real playground. The 1993 paper acts as a successor, keeping the 1979 worries alive while adding new ones.

Kalyva (2010) showed low agreement among parents, teachers, and children with Asperger syndrome. This extends the 1979 point: mixing views from many adults is needed because one rater can miss key deficits.

When you write a social-skills goal, pick at least two sources of data. Pair brief peer nominations with short live observations. Repeat both at different times before you decide if an intervention worked. This guards against flaky scores and gives families clearer feedback.