Journal of Scientific Exploration, Vol. I 1, No. 3, pp. 323- 335, 1997
A Psychological Comparison Between Ordinary Children and Those Who Claim Previous-Life Memories
ERLENDUR HARALDSSON
Faculty of Social Science, University of Iceland, 101 Reykjavik, Iceland
Abstract - It has been proposed that some normal-psychological factors may explain why some children speak of having had a previous life. Some of these and other psychological factors, which may further our understanding of children claiming previous-life memories, were the subject of this study. Psychological tests were administered to 30 children in Sri Lanka, aged 7 to 13, who at an earlier age had claimed to remember a previous life, and a control group of equal size. Children claiming previous-life memories show a higher level of cognitive functioning; perform much better in school, have a larger vocabulary, obtain higher scores on the Raven Progressive Matrices (brief test of intelligence), have better memory, and are not more suggestible than their peers. As a group they are gifted children. Parents found them to argue a lot, prefer being alone, be more nervous and stubborn than their peers, more perfectionistic, and more concerned about cleanliness. Teachers found them excellent to be pupils. As research continues the alleged memories of these children are found to be only one part of a pattern of characteristics that so far seem to defy a normal explanation.
Introduction
Investigations of cases of the reincarnation type (CORTs) have primarily been concerned with their verification: Do the statements reported by the children about a previous life correspond to verifiable events in the life of some de- ceased person? There is another potential research approach to these cases, which has until now been given only slight attention. It is concerned with the psychological aspect of the children. Does the psychological make-up of children speaking of previous-life memories contribute to an explanation of their claims? Do these children differ psychologically from children in general?
In critical discussions of child cases of alleged memories of a previous life, it has been argued that some normal psychological or socio-psychological fac- tors may lead a child to make claims of previous-life memories. For example, in a review of Stevenson's Cases of the Reincarnation Type, Vol. 11, Ten Cases in Sri Lanka (1977a), Brody (1979) proposed some potential factors: a rich fantasy life, a need to compensate for social isolation, high suggestibility (in cultures where belief in reincarnation plays a major role), dissociative tendencies, attention seeking, and disturbed relations with parents which may cause the child to claim that it belongs somewhere else. Twenty years ago Stevenson (1977a) pointed out that subjects of these cases who claimed a life in a socio- economic class different from that of their families seemed to remember a life in a higher class more often than one in a lower class. This could support a hypothesis of "escapist fantasies", but it is also open to other interpretations. The aim of this paper is to test some of these normal psychological hypotheses.
There are also claims concerning the psychological characteristics of CORTs that come not from thoughtful critics but from those who live with the young children and know them best. For example, it is sometimes reported to investigators that CORTs are more mature than other children, or that their command over their mother tongue is extraordinary in that they speak more like grown-ups than children. These claims have not yet been formally tested. In 1973, I. Stevenson, L. P. Mehrotra, and the author started a psychological study of children in India claiming previous life memories, but this study came to an abrupt end when one of us met with a serious traffic accident in India.
The main concern of the present study is the cognitive development of children who claim to remember a previous life, and the question of whether they differ from other children on some of the above-mentioned psychological factors, which may dispose children to make claims about a previous life. Do the abilities and personalities of children reporting previous-life memories differ significantly from children in general? Do they show a greater tendency to confabulate than other children? Are they more suggestible? Do they tend to live in social isolation? Are there indications of a greater tendency for dissociative processes in them than in other children?
In the present study a group of Sri Lankan children reporting previous-life memories and an equal number of control children were administered a battery of psychological tests, and their parents and teachers were interviewed and given questionnaires about their behavior; their school records were also obtained.
During the last eight years I have investigated over forty new cases in Sri Lanka, which is one of the countries where some CORTs can be found each year (Stevenson, 1977a). In a few of these cases the child's statements fit facts in the life of some person who lived before the child was born (Haraldsson, 1991; Mills, Haraldsson and Keil, 1994). In other cases - and they are more common in Sri Lanka than in most other countries - no person was found whose life corresponded to the child's statements (see also Cook et al., 1983a, 1983b). In some of these "unsolved" cases, some of the statements were detailed enough to be verifiable or falsifiable, and they were found to be false.
Most children with active previous-life memories are aged 3 to 5 years, and few objective psychological tests exist to assess the above-mentioned factors in children so young. A further complication is that these children are rare and difficult to find; but a meaningful comparison of them with other children requires a sample of adequate size. Therefore, in order to obtain a sufficiently large sample for this study, all available subjects up to age 13 had to be included. Since some of the tests could not be used with children younger than 7, the children in our sample ranged in age from 7 to 13 years. At this age most of the children had stopped talking about their previous-life memories but all of them had at an earlier age consistently talked about these memories over some period of time.
Part of the data of this study was presented in a paper published in the Journal of Nervous and Mental Disease (Haraldsson, 1995). This paper presents an extended sample, and contains almost no missing data.
Method
Subjects
The subjects were 30 children in Sri Lanka who had reported memories of a previous life. There were 12 boys and 18 girls, who ranged in age from 7 years and 1 month to 13 years and 2 months. A control group consisted of 30 children of the same age, the same sex, and from the same neighborhood, who had not spoken of a previous life. The mean age for children with memories was 9 years and 4.7 months, and it was 9 years and 4.1 months for the control group ( t= 1.08, n = 60, n.s.). The children were from a large area of south and central Sri Lanka and were about equally located in towns and in rural areas. Of the 30 cases involving claimed previous-life memories, 22 had previously been investigated by the author, and detailed reports had been published on five of them (Haraldsson, 1991; Mills et al., 1994). Eight cases had been investigated by Stevenson and his associates, but no report has been published on them.
Psychological Tests
The Peabody Picture Vocabulary Test - Revised (Dunn & Dunn, 1981) consists of a list of 175 words arranged in order of increasing difficulty. As each item is read to the child, the child is shown four black-and-white illustrations on a page and asked to select the picture that best illustrates the meaning of the orally presented stimulus word. The test was translated into Sinhalese by P. Vimala. It was administered without being standardized for Sinhalese children. Instead, a control group was used for comparison.
The Coloured Progressive Matrices (Raven, Court, and Raven, 1976, 1984) is described as "a test of clear thinking and observation". It tests the capacity to reason by analogy (Raven, 1963) and is frequently used as a brief test of intellectual efficiency. It was designed for use with young children and anthropological studies with people who do not understand the English language.
The Gudjonsson Suggestibility Scale (GSS)(Gudjonsson, 1984, 1987) was developed "to assess the individual's responses to 'leading questions' and 'negative feedback' instructions when being asked to report a factual event from recall" (Gudjonsson, 1992). A short fictitious story is read to the subject, after which helshe is asked to relate what is remembered of the story. Then the subject is asked 20 questions about the content of the story, 15 of which are suggestive in some way. Finally, the subject is firmly told that helshe has made a number of errors, and that it is therefore necessary to go through the questions once more.
The GSS measures: 1) Free Recall (the number of items remembered from the story); 2) Confabulations (the number of items offered under free recall that are not found in the story); 3) Yield Suggestibility (the number of items yielded to before negative feedback is given); 4) Shift Suggestibility (a distinct change in the nature of the reply to the 15 suggestible and 5 non-suggestible questions; 5 ) Total Suggestibility (the sum of Yield and Shift). There are two forms of the Gudjonsson Suggestibility Scale, the story of Form 2 being more appropriate for children (Gudjonsson, 1987; Danielsdottir et al., 1993). The test was translated into Sinhalese and adapted for Sri Lankan children by two Sri Lankan psychologists (Shanez Fernando and P. Vimala).
Child Behavior Checklist - Parent's Form (Achenbach & Edelbrock, 1983) is designed to record competencies and behavioral problems of children aged 4 to 16, It also surveys social competence and school performance and reveals the kind and number of a broad range of problems (118 items) that children may have. It was administered by P. Vimala to the mother of the child or to another close relative. The CBCL was translated into Sinhalese by P. Vi- mala, and hand-scored. Our testing sessions were quite long, and at the sug- gestion of P. Vimala, we decided not to ask about items I11 (social competence), IV (social activity), and four additional items (no. 59, 60, 73, 96) concerning sexual problems.
Child Behavior Checklist - Teacher's Form (Achenbach & Edelbrock, 1986) was administered to one of the teachers of each child. The CBCL-T is designed to obtain teachers' reports of their pupils' problems, adaptive func- tioning, and school performance. Most of the items are the same as in the Parent's Form. School performance figures (grades for test results) were obtained directly from school principals or from the children's school-performance books.
Statistical Methods
The data for this matched sample of subjects and control children were analyzed by the paired-samples t-test and the Wilcoxon matched-pairs signed- ranks test. The comparisons between "solved" and "unsolved" cases in Table 4 are analyzed by the two-sample t-test.

Psychological Comparison Between Children - Procedure
Each subject was visited unannounced at his or her home or school by the author, an interpreter, and a Sri Lankan psychologist. Some member of our team had in most cases previously interviewed the child and his or her parents about the case. Teachers helped us find a control child in the same class as the subject whose birthday was closest to that of the subject. If the subject was at home, we searched for a control child from the neighborhood, who was as close in age to the subject as possible. We expressed our appreciation by gifts of sweets and ball-point pens to these and to other children of the house, particularly after the session was over. All the families were cooperative and helpful. The CBCL - Teacher's Form was administered almost a year and a half later than the other tests/questionnaires to the teachers of 22 of the 30 pairs.
Results
Children claiming previous-life memories were found to differ in several respects from other children, as can be seen in Table 1. A pair-wise comparison of them with the group of control children revealed that their performance on the Peabody Picture Vocabulary Test (PPVT) was much higher (t = 5.38, n = 60, p < .0001, all tests two-tailed), which indicates that they had a substantially greater knowledge of words and a better understanding of language. This finding is strengthened by their higher average school grades for Sinhalese (their mother tongue) of 69.68, compared to 52.83 for the control group (t = 4.54, n = 60, p < .001).
The results of the Raven's Progressive Matrices are also significantly higher for children claiming previous-life memories (t = 3.23, n= 60, p<.01), indicating their greater capacity to reason by analogy. Furthermore, the results on the Gudjonsson Suggestibility Scale show that they have a better memory for recent events (t = 3.28, n = 60, p<.01) and that they are not more suggestible than other children (t = - 1.83, n = 60, ns). There are no indications that they confabulate more than their peers. If the number of confabulated responses is calculated as a percentage of total responses, children claiming previous-life memories show significantly fewer confabulations than the control children (t= 2.43, n = 60, p <.05)
The school performance of children reporting previous-life memories is better than that of control children, according to their parent's assessment on the Child Behavior Checklist (t = 5.69, n = 60, p < .0001). More importantly, their higher school performance is confirmed by their examination grades obtained from their schools (see Table 2). Their mean grades are much higher than those of the control children (t = 4.13, n = 60, p < .0001). When they are ranked according to their position in class (percentile = 73.8) and tested against the percentile of their controls (percentile = 49.8), their superiority as pupils is also obvious… At the age of 10, children of Sri Lanka can attend competitive national examinations. Those who obtain excellent marks are given admission to highly selective schools. Some of our children were in such schools.
Furthermore, teachers rated the target group significantly higher on adaptive functioning than their peers (t = 3.66, n = 56, p < .01); the children claiming previous-life memories learn more, behave better in school, and work harder than their peers (see Table 1.)

According to their parents, children reporting previous-life memories appear to have some behavioral problems, and the number of problems reported varied widely. The Child Behavior Checklist - Parent's Form reveals for the target group a much higher problem score (t = 4.50, n = 60, p < .0001). Also, according to their parents, the target group is neither more nor less socially active or socially competent than other children. On the other hand, according to their teachers (who observe the children only in school) these children have no more behavioral problems than other children (t = .28, n = 56, n.s.). In fact, they seem better adapted (see Table 1).
Individual items of the CBCL on which the children reporting previous-life memories score significantly higher or lower than control children are listed in Table 3.
According to their teachers, children reporting previous-life memories differ significantly from other children on 10 items on the CBCL, only three of which are also reported by their parents. Teachers and parents agree that they "feel they have to be perfect", that they like to be alone, and that they get less involved with other children.
It is interesting to note that the teachers also reported that children with previous-life memories fear more than other children that they might do something bad, are more unhappy, sad or depressed, get along better with other children, are less disobedient in school, are less impulsive and explosive, and have fewer difficulties in following directions than their peers.
Discussion
There can be no doubt that some very young children develop images and fantasies of past events (that occurred before they were born) which they regard as being memories of events they have experienced. The question is: How do these images or fantasies arise? Are they nothing but fantasies giving somehow a false impression of being memories, such as occurs in some deja vu experiences? Are they fantasies mixed with images of actual events that the child obtains paranormally? Or are they reflections of real events directly observed by the consciousness of the child before it was born? The study reported here has been an attempt to test the appropriateness of some mundane, normal psychological explanations for the origins of these images or fantasies.
The children of this study all reported memories of a previous life during the typical age range of around 3 to 5 years. When they were given psychological tests at the ages of 7 to 13, they were found to differ in many respects from other children. They were different both in cognitive development and in behavior. Their greater verbal skills were particularly distinctive. They had a larger vocabulary and a greater command of language than their matched peers. They performed excellently at school. The evidence for this comes both from their grades and their teachers, who found them to work harder, learn more readily, and behave better than other children. Children claiming memories of a previous life tend more than other children to think they must be perfect; such perfectionism could well contribute to their excellent performance in school.
These findings were somewhat unexpected and away from earlier conjectures about what would predispose a child to claim memories of a previous life. One of these suggests that the children are unusually prone to fantasies. The present study included no entirely satisfactory measure of this trait; nevertheless, the number of items added during free recall of the story given in the Gudjonsson Suggestibility Scale provided an indication of confabulation. The children who claimed past-life memories had a slightly lower score on this measure, although the difference was not statistically significant. We must note, however, that children who have a tendency to fantasize and who also have good memories and reasoning powers may not add details to a verbal test of memory.
One subsidiary finding relevant to the question of suggestibility deserves noting. Our sample included 16 children whose cases are "unsolved", which means that no person has identified the events of whose life corresponded to the child's statements; in the other 14 cases such a person was identified and their cases are considered "solved". The total suggestibility scores of the solved cases were significantly lower than the scores for the unsolved cases (t=2.27, n=30, p= .03). See details in Table 4.
The role of suggestion has especially been favored by persons who point out that the cultures in which these cases are most readily found have strong beliefs in reincarnation. Those who put forward this undoubted fact as a sufficient ex- planation of the cases overlook the evidence that parents often try to suppress the children and that the children for their part may vehemently oppose attempts at suppression. A good example of a child's persistence against parental opposition occurred in the case of Dilupa Nanayakkara, whose Roman Catholic parents tried to suppress her talk of a previous life (Haraldsson, 1991). Stevenson and Chadha (1990) and Mills (1989) have also some- times observed how these children (in India) withstand considerable pressure from their parents to stop talking about their apparent memories. Some of the children are scolded and a few even beaten for talking about a previous life. Many of them tell their mothers that they are not their real mothers, something the mothers do not like to hear and that often makes them afraid they may lose the child. It must, however, be added that in many cases the children's claims receive support and even encouragement from the parents, especially in solved cases. Nevertheless, it seems doubtful that the solved cases at least can be attributed solely to suggestions from the cultures in which they occur.
Social isolation is another factor that might contribute causally to a child's claiming to remember a previous life. The data of the present study do not indi- cate this clearly. On the one hand, the parents of the children claiming past-life memories do not rate them as less active or less competent socially than their peers. Their teachers, moreover, report that they are higher than their peers in adaptive functioning, which does not indicate that they are socially isolated. On the other hand, these children do like to be alone more than other children and they are often withdrawn. Other children sometimes tease them (probably often about their claimed memories) and this may stimulate them to withdraw. They seem to live under a considerable emotional strain. They are also, how- ever, argumentative, stubborn, and garrulous; these are characteristics that, on balance, hardly indicate social isolation. It is also pertinent to note that only 5 of the 30 children had no brother or sister; on average the children had two or three siblings and therefore could not have had a socially isolating environ- ment. The hypothesis that children who live in social isolation are more likely to claim memories of a previous life does not seem to be supported by this study.
As already mentioned, the children claiming past-life memories show "op- positional" characteristics: they are argumentative, stubborn, excessively talkative, and sometimes combative and threatening toward other persons. They score high on the parents' ratings for the Child Behavior Checklist (CBCL). It is difficult to know how to interpret such behavior, especially since the children seem in other respects more mature than other children. Is the abnormal behavior the result of antagonisms they have brought on themselves by their claims to remember a previous life-claims that are often uncongenial to their parents? Also, could their preference for being alone derive from a wish to think quietly about their memories? Or do they feel different and estranged from other persons because they have memories of a previous life and the oth-ers do not? We must be cautious in interpreting these differences.
What conclusions may be drawn from this study? Brody's (1979) hypotheses of social isolation and high suggestibility as possible causative factors for the reporting of memories of a previous life in young children were not confirmed in the present study. The same is true for a rich fantasy life as far as it is tested by a tendency to confabulate. We can say that so far as they have been tested, the normal psychological hypotheses have not been found adequate to explain why some children claim to remember a previous life. It is still possible that unusual tendencies to dissociation may cause children to talk about a previous life; this hypothesis will be tested in a further study.
It clearly emerges from this study that children reporting previous-life memories are a group of unusually gifted children. Experts on gifted and talented children often speak of "giftedness" to describe such children (Heller, Monks & Passow, 1993). Furthermore, an expert on gifted children has made the comment that the pattern of the problem items of the children who claim to re- member previous lives in this study is similar to that encountered with children of high ability (F. J. Monks, personal communication).
This brings us to the question of why children in this study who claim past- life memories show such a high level of cognitive development compared to ordinary children. Are children of high ability more likely than other children to create stories of a previous life, or, are children who remember a previous life more gifted?
The fact that a few of the cases investigated by the author (Haraldsson, 1991; Mills, Haraldsson & Keil, 1994) give rather striking evidence of para- normally acquired knowledge or memories of past events which the child is not found to have experienced in its life span, does not indicate that all these children are telling stories of fantasies. It is also the impression of Keil (1991), Mills (1989) and Pasricha (1990) for some of their cases, and, of course, of Stevenson, that some paranormality is involved. It is hence more likely that the giftedness of children who claim to remember past lives is one of a pattern of characteristics found in these children, rather than that gifted children are more likely than other children to imagine stories of a previous life.
Of course, our finding of giftedness needs a replication in another country before we can say that giftedness is a general characteristic of children who claim to remember past lives. If giftedness is found to be a general characteristic of the cases, and informal impressions make it seem rather likely, we have found one more characteristic of children who claim to remember past lives. If giftedness is found to be a general characteristic
of the cases, and informal impressions make it seem rather likely, we have found one more characteristic of children who claim to remember previous lives, for which, along with memories, phobias. predispositions (to certain
behavior patterns unusual or not found in the child's environment), birthmarks and deformities, we find in a fair number of cases no easy normal explanation. Furthermore, Stevenson (1987) has reported that some children reveal untaught or unlearned skills, some of which may require a paranormal interpretation.
These above mentioned characteristics fall increasingly into a pattern which must be viewed as a whole. The alleged memories have gradually become only one part of the phenomena that defy any easy normal explanation, and hence make these characteristics open to a paranormal interpretation. All of those (Keil, Mills, Pasricha and the present author) who in recent years have made a consistent effort to independently investigate these cases, have come to the conclusion that some of them do require a paranormal interpretation.
The crucial and most difficult question is, which paranormal interpretation? Mills and Pasricha have tended to support the reincarnation hypothesis, whereas Keil has argued that the hypothesis of extrasensory perception (super- psi) can sufficiently explain the phenomena (Keil deals mostly with claimed memories). I have remained silent on this issue. I wanted to wait until I had personally gathered more data on which to base my evaluation.
This area of research has many uncertainties and potential pitfalls which make solid data often difficult to obtain. After investigating more than 50 cases in Sri Lanka during a period of eight years, I have gradually found my- self increasingly critical of the super-psi hypothesis. It may suffice to explain the claimed memories, but I find it unconvincing (intuitively I must admit) as an explanation of deep-seated phobias and perplexing behavior patterns, not to mention birthmarks and malformations, for which, I must add, I have so far found rather little evidence in Sri Lanka. Birthmarks and malformations must begin to form during the development of the embryo and long before the child is born, which makes the super-psi hypothesis look even less satisfactory.
It may be premature to ask the question, but, tentatively accepting the reincarnation hypothesis, is the early maturity and giftedness of these children also something that they, in some form or sense, bring with them into their childhood in the same way as some of their other characteristics?
References
Achenbach, T. M. & Edelbrock, C. (1983). Manual for the Child Behavior Checklist and Revised Child Behavior Profile. Burlington, Vermont: University of Vermont Department of Psychiatry.
Achenbach, T. M. & Edelbrock, C. (1986). Manual for the Teacher's Report Form and Teacher Version of the Child Behavior Profile. Burlington, Vermont: University of Vermont Department of Psychiatry.
Brody, E. B. (1979). Review of cases of the reincarnation type. Ten cases in Sri Lanka by Ian Stevenson. Journal of Nervous and Mental Disease, 2, 167,769.
Cook, E., Pasricha, S., Samararatne, G., Maung, W., and Stevenson, I. (1983a). A review and analysis of "unsolved" cases of the reincarnation type. Part I, Introduction and illustrative case reports. Journal of the American Society for Psychical Research, 77,45.
Cook, E., Pasricha, S., Samararatne, G., Maung, W., and Stevenson, I. (1983b). A review and analysis of "unsolved" cases of the reincarnation type. Part 11, Comparison of features of solved and unsolved cases. Journal of the American Society for Psychical Research, 77, 115.
Danielsdottir, G., Sigurgeirsdottir, S., Einarsdottir, H. R., and Haraldsson, E. (1993). Interrogative suggestibility in children and its relationship with memory and vocabulary. Personality and Individual Differences, 14,499.
Dunn, L. M. & Dunn, L. M. (1981). PPVT Vocabulary Test - Revised. Manual for Forms M and L. Circle Pines, Minnesota: American Guidance Service.
Gudjonsson, G. H. (1984). A new scale of interrogative suggestibility. Personality and Individual Differences,5,3,303.
Gudjonsson, G. H. (1987). A parallel form of the Gudjonsson Suggestibility Scale. British Journal of Clinical Psychology, 26, 215.
Gudjonsson, G. H. (1992). The Psychology of interrogations, Confessions and Testimony. Chichester: John Wiley & Sons.
Haraldsson, E. (1991). Children claiming past-life memories: Four cases in Sri Lanka. Journal of Scientific Exploration, 5, 2, 233.
Haraldsson, E. (1995). Personality and abilities of children claiming past-life memories. Journal of Nervous and Mental disease, 183,7,445.
Heller, K. A., Monks, F. J. and Passow, A. H. (1993). International Handbook of Research and Development of Giftedness and Talent. London: Pergamon.
Keil, J. (1991). New cases in Burma, Thailand, and Turkey: A limited field study replication of some aspects of Ian Stevenson's research. Journal of Scientific Exploration, 5,27.
Mills, A. ( 1989). A replication study: Three cases of children in Northern India who are said to remember a previous life. Journal of Scientific Exploration, 3 , 2 , 133.
Mills, A., Haraldsson, E., and Keil, J. (1994). Replication studies of cases suggestive of reincarnation by three different investigators. Journal of the American Society for Psychical Research, 88,207.
Pasricha, S. (1990). Claims of Reincarnation. An Empirical Study of Cases in India. New Delhi: Harman Publishing House.
Raven, J. C. (1963). Guide to Using The Coloured Progressive Matrices, Sets A, Ab, B. Dumfries, Scotland: Director of Psychological Research, The Crichton Royal.
Raven, J. C., Court, J.H. and Raven,J. (1978). Manual for Raven's Progressive Matrices and Vocabulary Scales. General overview. London: H. K. Lewis.
Raven, J. C., Court, J. H. and Raven, J. (1984). Manual for Raven's Progressive Matrices and Vocabulary Scales. Coloured Progressive Matrices. London: H. K. Lewis, p. 3.
Stevenson, I. (1977a). Cases of the Reincarnation Type. Ten Cases in Sri Lanka, Vol. 2. Charlottesville: University Press of Virginia.
Stevenson, I. (1977b). The explanatory value of the idea of reincarnation. Journal cfNervous and Mental Disease, 164,305.
Stevenson, I. (1987). Children Who Remember Previous Lives: A Question of Reincarnation. Charlottesville: University Press of Virginia.
Stevenson, 1. (1990). Phobias in children who claim to remember previous lives. Journal of Scientific Exploration, 4, 243.
Stevenson, I. & Chadha, N. K. (1990). Can children be stopped from speaking about previous lives? Some further analyses of features in cases of the reincarnation type. Journal of the Soci- ety for Psychical Research, 56,818,82.

