H567/02 Section A - Lee et al. (1997): lying and truth-telling

Culture, sample and design

Earlier work on children's truth and lie judgements had relied heavily on Western samples. The researchers considered whether social norms mattered alongside factual accuracy. In particular, a norm of modesty may make claiming credit for a good deed appear less desirable, and denying that deed appear acceptable. This predicts a contextual evaluation, not a rule that a whole nation approves every lie.

The sample contained 120 Chinese children from elementary schools in Hangzhou: 40 each aged 7, 9 and 11, with 20 boys and 20 girls in each age group. There were 108 Canadian children from Fredericton: 36 aged 7, 40 aged 9 and 32 aged 11. They included boys and girls and were mainly middle class. The original report did not have equivalent socioeconomic information for the Chinese sample. There were 228 children altogether; unequal background information complicates attributing all differences to culture.

Age and cultural group were naturally occurring variables. This was a cross-sectional comparison, not a longitudinal follow-up of the same children. Each child was randomly assigned to either social stories, concerning direct effects on another child, or physical stories, concerning objects or the environment. Within that set, the child heard all four combinations of good/bad deed and truth/lie. Thus the design combined independent story-type groups with repeated judgements within each child, alongside naturally occurring age and cultural groups. Random story assignment did not randomly assign culture or age.

Stories, ratings and procedure

Each child heard four illustrated stories. Social stories included helping or harming another child; physical stories included improving or damaging things in the school environment. The deed and the later statement were distinct parts of the task:

A child first evaluated the deed, then heard the character's truthful admission or false denial and evaluated the statement separately. This separation mattered: approving the helpful action did not require approving an admission that could be seen as boasting.

Children used a seven-point chart: +3 very, very good; +2 very good; +1 good; 0 neither good nor naughty; −1 naughty; −2 very naughty; −3 very, very naughty. Stars represented positive responses, crosses negative responses and a circle the middle. Children were trained in the scale and reminded of its meaning. They could answer verbally or point to symbols. These ordered judgements supplied quantitative rating codes, but equal psychological distances between adjacent labels were not established.

Two story orders were distributed using a randomisation table, with approximately half in each order; the phrasing of good/naughty questions was alternated. These controls reduced systematic order and wording effects. Discussion after the task explored explanations, offering qualitative information about norms. Preliminary analyses found no significant gender or order differences, so the principal results combined these groups. That finding does not prove gender and order can never matter in another sample.

Results and the cultural interpretation

Children in both samples generally approved good deeds and disapproved bad deeds. They also generally evaluated truth after a bad deed positively and a lie after a bad deed negatively. The main cultural contrast concerned statements about good deeds.

The following are means of the coded statement ratings, combining social and physical story conditions from the original study. Positive means fall on the good side; negative means on the naughty side. They are not percentages of children who lie.

With age, the Chinese groups evaluated good-deed admissions less positively and false denials more positively. The mean denial rating changed from −0.60 at seven to 0.98 at eleven, crossing the neutral point. Canadian good-deed admissions remained positive, and denials remained negative. The Canadian values did not change monotonically with age, so do not describe both series as steady increases.