The aim of the two studies reported was to test two extensions of the dissonance model. The hypothesis was that over longer time periods the congruence between initially dissonant emotional attitudes with regard to a past (not recent) life period will become more pronounced. The main variables were the evaluative attitude to the whole past period in terms of happiness, expressed as a ratio of unpleasant (U) to pleasant (P) checked adjectives, the ratio of U:P memories recalled from that period, U and P being judgements from the viewpoint of the present, and the wish to return to that period.

The sample of Study I included 126 Kibbutz children, who were all the children in three Kibbutzim similar in child-rearing methods in the following three age-groups: 8.1–10, 10.1–12 and 12.1–14. Each child was interviewed individually according to a questionnaire about his childhood (the period 1–6 yr). The results revealed no significant differences between the groups in the U:P memories ratio (average = 2·15) or U:P adjectives ratio (average = 0·38), and no significant correlations between the variables of wish to return to childhood, U:P adjectives and U:P memories, but for the correlation between the latter two in the 8.1–10 yr old group. The results refute the hypothesis in revealing the persistence of dissonances.

In Study II 35 adults in the age range 30–50 yr, chosen randomly from three localities, were interviewed in an identical manner about the period of 20–25 yr as well as about their evaluation of happiness in personal childhood, wish to return to childhood, evaluation of happiness in childhood in general and of humanity in the past. In addition to corroborating fully all the results of Study I in regard to the age of 20–25 yr it was found that evaluation of happiness in personal and general childhood are correlated, as well as the lack of wish to return to one period with that to another. Possible explanations and implications of the persistence of the dissonances are discussed.