SHARON NELSON’S EXPERIMENT ON YOUNG CHILDREN’S MORAL CRITERIA (1980)
What is morality? It’s a very debatable definition, that’s for sure. For the sake of simplicity, however, we will define morality as the presence of these three features: distinguishing between right and wrong, acting on the distinction between right and wrong, experiencing pride after virtuous acts and shame or guilt after acts that violate one’s moral standards.
Jean Piaget, who is arguably one of the most influential developmental psychologists in history, examined the morality of children. He introduced his own theory of moral development which he based on his own observations of and interviews with children. The basic assumption of his theory is that “moral development is a function of cognitive development”.
Piaget outlined three stages of moral development: (a.) Premoral stage (from birth until four to six years of age), (b.) Heteronomous morality stage (from six to 10 years of age), and (c.) Autonomous morality stage (from 10 to 13 years of age). Below six years of age, Piaget believed that children had very little awareness of rules. In the second stage, Piaget said moral judgements of the children are based on outcomes rather than motives while it is in the third stage that a child learns to formulate their own moral rules and makes judgements based on motives.
Nelson believed that Piaget underestimated the children. She conducted two experiments to test the basis of children’s moral criteria: basically, did children judge someone’s actions based on the reason for their actions or the consequences of their actions?
In the first experiment, the sample consisted of 60 preschool children between three to four years of age and 30 second-grade children between six to eight years of age. Approximately half the children in each grade were female and the other half was male. The children were all from a middle-class, urban area and primarily white. Their parents gave consent for them to be participants.
The experimenters created four versions of the same story. In each version, a boy acting with a good or bad motive threw a ball at his friend and this resulted in a good or bad outcome. This particular storyline was chosen because in a different study, children regarded “throwing a ball” as a neutral act. Motive descriptions came before outcome descriptions, and the actor’s overt behaviour was constant. The levels of motive and outcome were:
- Good motive – Boy was playing with a ball but his friend didn’t have any toy to play with. Boy wanted to throw a ball to his friend so that they could play catch together.
- Bad motive – Boy was playing with a ball and was mad at his friend at the time. He wanted to throw the ball at his friend to hit him on purpose.
- Good outcome – Boy threw the ball to his friend, who caught it and began playing happily.
- Bad outcome – Boy threw the ball to his friend but it hit him on the head and made him cry.
To reduce memory constraints and to examine the effects of motive salience, two sets of black and white drawings were made to accompany the information presented in each version of the story. There were two sets of drawings: the first set conveyed the motive of the actor implicitly by implying it through the actor’s facial expressions while the second set conveyed it explicitly with thought bubbles connected to the actor’s head.
Two sets of faces were used as rating scales. There were three smiling faces that represented the degree of the boy’s goodness from “a little bit good” to “very good” while three frowning faces represented the degree of the boy’s badness from “a little bit bad” to “very bad”. There was also a small neutral face representing a judgement of the boy being “just okay”. This created a seven-point scale for the children to use.
Children of each age were randomly put into one of the three story-presentation conditions. There were 20 children in each group at the 3 year old age and 10 children in each group at the 7 year old level. Every child heard all four stories and the order of presentation to each child was randomly determined. An independent-measures experimental design was used.
Children were interviewed individually by an experimenter. Before the actual study began, children were familiarised with each point on the rating scale and then given two practice stories to define the “very good” and “very bad” extremes on the scale. The children in the picture-motive-explicit condition were also given practice to familiarise them with the cartoon conventions used to illustrate the actor’s motive.
When the experiment began, children were told to listen carefully because later they would be retelling the stories aloud. After each story, children were asked whether the little boy in the story was good, bad, or “just okay”. Then they had to indicate the degree of good or bad by pointing to a face on the rating scale.
In both of the picture-presentation conditions, the drawings were introduced one by one at specific points during the story. They were placed side by side in front of the child. After the child made their judgement, the drawings were removed and children had to retell the same story aloud. If motive or outcome information was omitted when retelling, specific questions were asked to draw out the information, for example: “Why did the boy throw his ball?” and “What happened after the boy threw his ball?”
The results supported the original hypotheses, which were:
- Children as young as three years old can and do use motive information when making moral judgements if the information is explicit, salient and available.
- When motive and outcome have opposite valences, children usually recall the story to make them congruent.
The results can be summarised as:
- When the motive information was pictured explicitly, good and bad outcomes had a greater effect on judgements compared to when motive information was implicitly pictured or not pictured at all.
- Good outcomes were rated more positively than bad outcomes.
- Children showed a significant use of outcome information in the ‘good motive’ stories under all types of presentation, but in the picture presentations they showed a significant use of outcome information when the motive was bad.
- Three year olds made more errors when recalling motives and outcomes, compared to seven year olds.
- Three year olds made relatively more errors when recalling motive valences than outcome valences when the information was conflicting than when it was congruent.
- The congruency or in-congruency of valence information had no effect on the pattern of recall errors made by seven year olds.
A second experiment was conducted to investigate the possibility that the emphasis given to motive by three year olds in the first experiment may reflect a confounding of information about the valence of the cues with the presentation order of the source of the cues within the stories.
The sample of the second experiment was made up of 27 preschool boys and girls. The average age was around three years old.
Children were randomly allocated into one of the three presentation groups. The procedure was identical to the procedure of the first experiment.
In all stories and presentation types, the description of the outcome came before the description of the motive.
The results of the second experiment showed the following:
- Good outcomes were rated more positively than bad outcomes.
- Good motives were rated more positively than bad motives.
- Whenever the motive or outcome was negative, the other cue in the pair had less influence on the child’s judgement.
- Judgements made in the verbal presentation condition were less influenced by motive than those in the picture presentation condition.
- Children made more errors when recalling story information in the ‘good motive’ stories if the outcome was bad.
- In ‘bad motive’ stories, there were more errors made when the outcome was good.
- More errors were made for in-congruent pairings rather than congruent pairings.
Type of research method
This was a laboratory experiment.
An independent-measures experimental design was used.
The three types of story presentation were the independent variables, specifically: verbal presentation, verbal with picture explicit presentation, and verbal with picture implicit presentation.
The dependent variables were the response of the child and the amount of information that was successfully recalled.
- Replicable: The main procedures of the experiments were standardised, meaning they can be replicated and the reliability of the results can be checked.
- Relatively generalisable sample (size and gender): Although the experiment was not very large-scale, it still had a reasonable sample made up of boys and girls of different age groups.
- Good ecological validity: The experimenters used stories that are very realistic and likely to happen in real life. This makes the children’s judgements more natural and improves the ecological validity of the studies.
- Child-friendly experiment: The rating scales and stories were all modified to suit a child. This means that it is less likely that the sample would become confused and distort the results.
- Cultural bias: The experiment did not really acknowledge cultural factors: there is always the possibility of children’s judgement changing if they were born into different cultural norms, different ethnic backgrounds, different socio-economic statuses, etc. All the children in this study were from a similar background, social class and ethnicity.
- Artificial environment: Although the ecological validity of this experiment is not low, the environment it was conducted in was artificial and unrealistic.
- Demand characteristics and experimenter bias: It is possible that, due to the experimenters being present and the artificial surroundings, children (especially the slightly older ones) may have modified their answers without meaning to. Maybe they were influenced by the setting, the experimenters or maybe they wanted to “give the right answer” and gave answers that they thought were expected of them. Also, the experimenters talked to the children (e.g. when telling them to listen carefully) and this could have influenced the child, also resulting in bias.
- Informed consent: This was obtained from the parents of the children involved in this study.
- Deception: There was no deception.
- Confidentiality: No personal details were made public; confidentiality was kept.
- Emotional or physical harm: There was no chance of harm.
- The right to withdraw: Nelson didn’t give much information about this. It is likely that parents could withdraw their child from the study but this isn’t confirmed.
- Debriefing: Since there was no deception and the participants were all children, there was not much need for explicit debriefing.
Reference: Nelson, S. A. (1980). Factors Influencing Young Children’s Use of Motives and Outcomes as Moral Criteria. Child Development. 51: 823-829.