PARENTAL WEBSITE-DESCRIPTIONS OF CHILDREN’S IMAGINARY COMPANIONS

Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: http://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.


Introduction
Imagine a child talking to her mother clearly jealous of a toy that another girl has, pointing at a space where in fact no one is standing!This might seem like odd behaviour.However, the child may be talking about an 'imaginary friend' (defined as make-belief friends or playmates).In this case, the behaviour is actually quite normal.Having an imaginary or pretend companion, is a common experience for schoolchildren.About 50 to 65% of the children have one or more imaginary companions at a certain point of childhood (Hoff, 2005;Pearson, Rouse, Doswell, Ainsworth, Dawson, Simms et al., 2001;Taylor, Carlson, Maring, Gerow, & Charley, 2004).In addition, several studies have been carried out investigating differences between children with and without experiences of imaginary companions.None of these studies indicate that imaginary companions are a signal for psychopathology (Gleason, Jarudi, & Cheeks, 2003;Taylor, 1999;Taylor et al., 2004).Yet, scientific knowledge may or may not correspond to the knowledge of parents.In other words: parents might worry about whether it is normal for their child to have a pretend friend even though pretend friends are a common, normal phenom-enon in childhood.This brief report deals with the question of parental concerns about pretend friendships in children as reported on internet-forums.
Beliefs about imaginary companions have changed over the years (Klausen & Passman, 2007).A common belief used to be that imaginary friends would be created by children in order to compensate for poor social relationships (e.g., Harvey, 1918;Bender & Vogel, 1941).However, even though imaginary friends can serve similar functions as real friends (Gleason, 2002), they do not indicate a lack of real friendships (Gleason, 2004a).In fact, they are even thought to stimulate cooperativeness (Gleason et al., 2003).This might be due to the increased possibility of practice with perspective taking of children with pretend friends compared to children without pretend friends (Roby & Kidd, 2008).With increasing knowledge derived from research, imaginary friends in childhood have therefore come to be considered by scientists as a normal part of children's pretend play (Klausen & Passman, 2007).
Simultaneously, imaginary companions also "speak to the imagination" of many people and the concept is frequently used for fantastic, scary storylines.For example, in horror movies imaginary friends turn out to be ghosts (e.g., The Amityville Horror) and personified objects (where a child uses a stuffed animal or doll as an imaginary companion) turn out to be monstrous (e.g., Child's Play).Although these movies are clearly fictive, they may give adults negative associations to the idea of imaginary companions.In addition, parents may have concerns about imaginary companions for religious reasons.For example, some fundamentalist Christians might feel there is an association with the devil (Taylor & Carlson, 2000).Furthermore, whereas imaginary companions correspond with the use of imagination and play during childhood, the onset of similar, but more persistent symptoms in adolescence/adulthood in the absence of recognition of the imaginary component can indicate psychopathology (e.g., risk for schizophrenia; Gupta & Desai, 2006).The similarity between imaginary companions and hallucinations or a lack of reality-awareness may be confusing for parents.In short, parents may become concerned when their child has a imaginary companion.
This concern might be related to the function the imaginary companions have, or at least are perceived to have according to the parents.As described above, a common thought used to be that children with imaginary companions lack social skills ('deficit hypothesis').Others argued against this with the 'gifted hypothesis' that these children are bright and creative (Seiffge-Krenke, 1997).Research has invalidated the deficit hypothesis.Children do use imaginary companions for companionship (Seiffge-Krenke, 1997), but rather in a positive, adaptive way.Furthermore, researchers have also described imaginary companions as serving the following ego-functions: understanding and agreeing, supporting, helping, listening to, projecting the ego-ideal and soothing loneliness (Inuzuka, Satoh, & Wada, 1991).Finally, children can use their imaginary companions to accuse their own misbehaviour (Gleason, Sebanc, & Hartup, 2000).
In addition, the functions described above might be seen as more adaptive for younger children than for older children who, for example are expected to have certain more advanced emotion regulation skills (Holodynski & Friedlmeier, 2006).Perhaps, therefore, parents are more likely to be concerned about imaginary companions in older children.In addition, in the scientific world, it is well-known that many young (preschool)children have imaginary companions (Taylor et al., 2004).Older children, in contrast, tend to less overtly act out their fantasy, which has lead to more unawareness of the prevalence of imaginary companions in middle childhood (Pearson et al., 2001;Taylor et al., 2004).
Besides function and age, gender might be related to parental concerns.After all, girls show different pretend-play than boys.Girls' playing behaviour focuses more on relationships than that of boys (e.g., play house), whereas boys' pretend play focuses more on action and fantasy (e.g., beating a monster ;Fein, 1981).It is imaginable that parents may feel less concerned when their child's imaginary companion fits this gender-role behaviour (i.e., girls with an imaginary friend, boys with an impersonated object; Taylor et al., 2004) rather than contradicts it.

The current study
The aim of this study was to find out whether parents show concern about imaginary companions.This would indicate a need for information for parents in order to reassure them.Parental descriptions of children's imaginary companions have been addressed in the past (Gleason, 2004b), but this study addresses parental concerns about imaginary companions in contemporary society (i.e., 2003-2008).Internet-forums were searched as the source of information about parental attitudes.Internet-forums have the disadvantage of attracting people who for some reason want to express their feelings about their children's imaginary companions, which may have caused a bias in the results.This study should therefore be considered as a first step of achieving insight into parental concerns about imaginary companions.There are nevertheless several advantages of using the internet as the information-resource over (randomly) asking people to participate.First, the information is freely available without having to give a prompt.Therefore, the message of parents is uninfluenced by research purposes or assumptions of the parents about research purposes.Second, it has been found that people frequently use internet-forums for sharing information and feelings (Hura, 1998;Kummervold, Gammon, Bergvik, Johnsen, Hasvold, & Rosenvinge, 2002;Wesemann & Grunwald, 2008).Many people state that they find it easier to discuss personal information with others on the internet (often using a pseudonym) than face-to-face (Kummervold et al., 2002).It can therefore be expected that the messages on the internet or less influenced by social norms.
Several characteristics were examined as potential influences on parental attitudes.It was hypothesised that the perception of the functionality of imaginary companions could be related to parental concerns.Given the lack of more specific information, this hypothesis was approached exploratory.Further, it was hypothesised that parents would be less concerned about imaginary companions in younger children and this could be associated with the function.Finally, it was expected that parents would be less concerned about imaginary companions in girls than in boys.

Method
The internet was searched in English, Dutch, and German.Synonyms of imaginary companions were used (imaginary friend/companion, pretend friend/companion).There were 35 websites found: 11 from Germany, 9 from the USA, 7 from the UK, 6 from the Netherlands, 1 from Austria and 1 from Switzerland.These websites contained 89 posts by parents that were about present imaginary companions, in their own children.Posts about other children or passed occurrences were excluded.Most often, the parent who had posted the message was the mother (90 mothers, 4 fathers, 6 unknown).Of the messages, 80.9% were original posts and 19.1% were reactions to previous posts.The website forums dated from 2003-2008.The function and parental concern were double coded, with the following intraclass correlation coefficients: companionship function (ICC = .77),compensation/projection function (ICC = .90),emotional support function (ICC = .77),misbehaviour function (ICC = .83),parental concern (ICC = .82).Items on which the coders disagreed were discussed until a code was agreed upon.

Descriptives
All but two parents mentioned the gender of their child: 66.7% were girls, 31.1% were boys.The age ranged between 1 and 8 years old, SD = 3.7.The functions that were described are summarised in Figure 1.Many parents did not explicitly state a function.Of those who did, the companionship function was described most often.Gender of the imaginary companion was mentioned in 69.7% of the cases.Many children had a imaginary companion of the same gender (41.9%) or multiple imaginary companions at least one of which was of the same gender (35.5%).This result was similar for boys and girls, χ(2) = 0.69, p = .71.About half of the girls had more than one imaginary companion (54.2%), whereas only 28.6% of the boys had multiple imaginary companions, χ(1) = 5.04, p = .03.There were even a few descriptions of invented families (5 girls and 1 boy).The majority of the imaginary companions (80.7%) were human, independent of child-gender, χ(2) = 1.29, p = .52(8% of the children had both human and animal imaginary companions).Only 5.6% of the parental descriptions were about personified objects.Finally, 78.7% of the parents mentioned the age of the companion (relative to the child), which was the same as the child in 72.9% of the cases and was older in 12.9%, independent of child-gender, χ(3) = 0.16, p = .98.

Parental concerns
Of all parents, 55.1% expressed concerns.As expected, children's age was associated with parental concern.On average, parents who expressed concern about imaginary companions had older children than non-concerned parents, (M = 3.36, SD = 0.79 versus M = 4.00, SD = 1.54), t(61) = -2.25,p = .03(equal variances not assumed).Figure 2 shows that parents of children older than 5 tended to be concerned whereas there was more variation in parental concern for younger children.Parental views on the functions of the imaginary companion were unrelated to parental concerns.Given the small number of personified objects, the association with gender and parental concerns could not be tested.There was no main gender effect on parental concerns We explored whether parents who responded were less likely to express concern, but this was not the case, χ(1) = 1.92, p = .17.

Figure 1 Functions of imaginary companions mentioned by parents on internet-forums
Misbehaviour e.g., invisible friend is blamed when something is gets broken (11%)

Functions of pretend friends
Emotional support e.g., is there for the child when (s)he is nervous (9%) Companionship e.g., is a playmate for the child (29%) Compensation/projection e.g., has the toys that the child would like to have (9%) As discussed in the introduction, posts on internet-forums will have selfselection bias: all messages were from people who wanted to express their feelings about their children's imaginary companions.Our study can therefore be seen as a preliminary step in investigating parental concerns about imaginary friends.
The finding that there was variation in parental concern for the younger children, but that after age 5 parents expressed concern was expected.Probably, not all parents know that also in middle childhood imaginary companions are common and harmless.An additional explanation of parental concerns about imaginary companions in older children might be that these children have imaginary companions for a longer time.Some of the descriptions indicated that parents can feel imaginary companions belong to a phase (e.g., "my 3 year old daughter has a friend called Molly who goes everywhere with us I don't worry about it as I'm sure Molly will go when she is ready").
In conclusion, the results of this study indicate that there is a need for more information about pretend friends for parents.This could, for example, be achieved through easy to find, clear Websites on the topic, and promoting teachers to discuss the topic at parent meetings.We would like to particularly stress the need of this increase in information for parents of older children (aged 5 and above).