A Theoretical Knowledge of Interpersonal Safety Skills Is Not Related to Children’s Ability to Protect Themselves in Real-Life Situations ()
1. Introduction
Throughout their lives, children may be exposed to a variety of safety threats such as abduction (Miltenberger, Sanchez, & Valbuena, 2015) . AMBER Alert Europe, the European Centre for Missing Children, indicates that every year 300,000 children go missing across Europe (https://www.amberalert.eu/). A national helpline of the organization “The smile of the child” in Greece reports receiving at least 200,000 calls per year concerning children in danger. Data from the official records of the Hellenic Police indicate that almost 300 children are missing per year. However, most of the missing children involve teenagers and have a positive outcome since most of them return safely to their homes (https://www.hamogelo.gr/). The chances of a child being abducted in Greece are low and they are even less when it involves abduction by a stranger, exactly as it is in the United States (for review see Drew et al., 2022 ). Even though a safety threat like abduction is still relatively rare in Greece, many parents and educators have concerns about children’s related behaviors in order to stay safe with strangers or with people they know.
Especially young children have not developed self-protection skills and have difficulty in determining when a behavior can be dangerous or when a reaction is considered acceptable or not (Wurtele, 2009; Zhang et al., 2013) . Research has shown that 75% - 90% of young children follow an unknown person willingly, when she/he appears with an “attractive lure”, such as toy, sweet or presented as a friend with the teacher or parent (Goldfarb, O’Brien, & Krackow, 2008; Marchand-Martella et al., 1996; Poche, Brouwer, & Swearingen, 1981) . This means that it is very easy, if the right way of contact is used, to fool a child and put itself in a situation that poses a danger. Discriminating the presence of a safety threat is necessary for the child to avoid, escape, and report the incident (Miltenberger, 2008) . Most children would reply negatively if they were asked whether they would follow a stranger or not. However, it has been indicated that there is a low correlation between what children answer and what they do in a real situation (Carroll-Rowan & Miltenberger, 1994; Miltenberger et al., 1991) .
Most surveys have evaluated safety threats and skills children need to respond safely, through questionnaires and personal interviews (Miltenberger et al., 2009; Miltenberger & Hanratty, 2013) . However, it has been reported that many children while stating that they would never follow an unknown person, in fact they do so when they are in conditions of a real-world setting (Miltenberger et al., 1991) . Thus, it seems very important to conduct assessments in naturalistic settings (e.g., school, park or home) and without children’s awareness that they were being tested. This form of assessment could allow researchers to determine whether children would behave in a way that is consistent with their verbal statements of what they should do in an abduction situation or not (Carroll-Rowan & Miltenberger, 1994; Olsen-Woods, Miltenberger, & Foreman, 1998) . All this evidence about how children react can stay safe with strangers and people they know comes mainly from studies in USA. Although child safety has been of great concern to the policies of various countries in Europe, to our knowledge there is no relevant scientific study in Greece or in any other European country that assesses the ability of young children to recognize and react properly in hypothetically unsafe situations.
After all, the aim of the present study was to assess preschoolers’ knowledge of interpersonal safety skills and their ability to recognize and react properly in potentially unsafe situations. The evaluation was done in two different contexts. Initially, children’s knowledge to recognize, react and inform about hypothetical dangerous situations was assessed through a personal form of an interview. Then, it was investigated whether participants can be lured to leave away with a stranger in a naturalistic context.
2. Methodology
2.1. Participants
The 555 healthy pre-school aged children (291 boys and 264 girls with mean age = 61.64 ± 6.96 months) who participated in this study were sampled from various kindergarten schools in Serres, a big town of Northern Greece. Schools were selected by convenience sampling according to teachers’ willingness to participate in the present study.
All parents or legal guardians provided written informed consent prior to participation, which was approved by an Institutional Review Board for use of Human subjects, allowing the children’s involvement in the program and access to relevant information. This study was approved by the ministry of education Athens and by the local University Ethics Committee.
2.2. Testing Procedures and Instrumentation
Measurement Process
1) Realistic test based on observed protective behaviours test (OPBT; White et al., 2015 )
The Observed protective behaviours test (OPBT; White et al., 2015 ) is a two-part in vivo behavioural test designed to assess children’s ability to implement protective behaviours knowledge and skills. It is based on the behavioural skills training measures used by Johnson et al. (2005) and Gunby, Carr, & Leblanc (2010) . The test produces measures of a motor response to the lure, a verbal response to the lure, and disclosure. In the present study, a test, which was carried out in realistic conditions similar to those of previous research, was designed. More specifically, it was involved two potential lures, the incentive (e.g. “Come see the games in my car”) and assistance request (e.g. “Help me find my keys”).
The aim of the test was to assess the child’s ability to demonstrate the appropriate self-protection skills (avoid - escape - inform). The realistic test that took place before the personal interview with the children, took the form of a social experiment. An assessment of the children’s willingness to follow an unknown young woman who tried to lure them by candy or a toy. We chose a woman confederate as it has been previously suggested that children are more prone to trust a woman and therefore would be more willing to follow her (Harris et al., 2012; Koenig & Harris, 2005; Li et al., 2020; Pasquini et al., 2007) .
The test was held during school hours, with each student separately. Initially, each student went out to the school yard with his/her teacher. Teacher did not inform the child that an assessment was taking place. Then, a teacher found an excuse in order to leave the child alone asking him/her to wait for a while, without moving away. As soon as the child was left alone, a young, unknown woman entered the school yard and pretended that she needed help to find her keys that she had apparently lost. After the child helped the woman to find the keys, she immediately offered him/her a candy as a reward. Then, she engaged the child in friendly conversation and tried to elicit information about the type of his/her favorite toy. Then, the unknown woman suddenly pretended that she had a toy, exactly as his/her favorite one in her car and asked the child to follow her. She used phrases like these: “You are so kind. You were really helpful. I really like you. I have a gift that I want to give to you” and “Let’s go together in my car and get the present”. If the child seemed reluctant to follow her, she promised that she would give him/her the gift and would return immediately. That way the teacher wouldn’t realize that he or she disobeyed her.
If the child followed her, as soon as they approached the car, she pretended that she had forgotten the keys of the car. However, she promised that she would return the next day to bring him/her the toy. As she was leaving, the woman kindly greeted and asked the child to keep her presence a secret between them, without revealing it to anyone. Then, the child returned to another class where another researcher was waiting in order to interview him/her in What If Situations Test (Wurtele, Hughes, & Owens, 1998) . After the end of the interview, the child would go to another class where another teacher would take care of him/her until all the children would be assessed. This ensured that the children would not discuss anything related to the experiment or interview before all participants finished.
2) What If” Situations Test (Wurtele, Hughes, & Owens, 1998)
The What-If-Situations-Test (WIST) is an instrument that evaluates the reaction of children to hypothetically dangerous situations. It is a personal form of interview that assesses children’s ability to recognize, resist and inform about dangerous situations. It includes 3 scenarios where each one includes 3 different reactions. 1) WIST SAY, such as refusing to go along with the inappropriate requests through a verbal response (ranging from 0 to 3); 2) WIST DO, such as leaving the dangerous situations (ranging from 0 to 3); and 3) WIST TELL, such as telling the CSA occurrence to trustworthy adults (ranging from 0 to 4). In general, the maximum scores of three subscales “SAY”, “DO”, “TELL” were 3, 3, and 4, respectively, and the range of total skills score was from 0 to 10. Scores were evaluated as in a previous study: 0 corresponded to an extremely low level of knowledge, 1 to 5 to knowledge below average, 6 to 9 to knowledge above average, and 10 corresponded to an excellent level of knowledge (Yu et al., 2017) . Only 2 scenarios of the questionnaire were used in the present research. The first scenario is what would you do if a stranger whom you helped in something (e.g. find something, carry something heavy, give him/her a piece of information) and wants to reward or treat you for it. The second scenario is what would you do if a stranger asks you to follow him/her in the car because he/she has something you like there. This evaluation tool has been reported to have sufficient reliability (0.7 - 0.79) and internal consistency (0.8 - 0.89) (Jin et al., 2016; Yu et al., 2017) .
3. Results
3.1. Realistic Test Results
The reactions of the children in the realistic test are presented in Table 1. More specifically, descriptive statistics were used to describe the percentages of children who agreed to take the candy “lure” and follow the unfamiliar young woman in the car. Of the total of 555 children, 474 (85.4%) took the candy, while 81 (14.6%) did not. Regarding their reaction when asked to follow the unknown woman in the car, out of the total of 555 children, 400 (72.1%) did follow, while 155 (27.9%) did not.
3.2. The Children’s Knowledge as Reflected in the Questionnaire
In Table 2 are presented the exact number and percentages of children’s knowledge are presented as assessed through the questionnaire in the two hypothetical scenarios. More specifically, in the first scenario (getting the candy lure as a reward after helping) it is observed that children’s knowledge is almost equally divided between low knowledge (n = 317, 110 extremely low, 207 below average) and high knowledge (n = 238, 83 above-average, 155 excellent). Similarly, based on the second scenario (to follow the young woman in the car), answers are equally divided between low (n = 269, 90 extremely low, 179 below average) and high knowledge (n = 286, 111 extremely low, 175 below average).
3.3. Relationship between Children’s Knowledge and Reaction during Realistic Test
Table 3 shows the relationship between children’s theoretical knowledge about the two scenarios and their reaction to the similar realistic test. The results showed that the children’s level of knowledge was independent of how they reacted to the realistic test. This means that whether the children had a high level of theoretical knowledge or not, this was not related to their reaction to the test in real conditions, since there was no statistically significant relationship, not even
![]()
Table 1. Percentages of children according their reactions to the realistic test.
a. Sample of a Table footnote ( Table footnote is dispensable).
![]()
Table 2. Children’s level of knowledge about the two scenarios.
![]()
Table 3. Relationship between children’s knowledge and reaction during realistic test.
in the candy scenario (χ2 (3) = 8.274, p = 0.080), nor in the scenario of the car (χ2 (3) = 8.563, p = 0.105). For example, as it was seen in Table 3, line 3, from the total number of participants that had theoretical knowledge above average, the percentage of them that took the candy lure (15%), comparing with those that did not (14.8%), was almost the same. Similarly, in the realistic test the percentages of participants that followed the stranger woman in her car or not were similar (18.7 and 23.2, respectively). This means that even if a child understands in theory that a situation is dangerous, it does not mean that children will react properly when they had to deal with it in a realistic setting.
4. Discussion
The aim of the present study was to assess preschoolers’ knowledge of interpersonal safety skills and their ability to recognize and react properly in potentially unsafe situations. Although the risk of a young child to be involved in a dangerous situation for his/her life is one of the biggest concerns of modern societies, to our knowledge, this is the first study applied to such a large sample of 555 preschool children in a European country. Until now, most of the publish safety skills research was carried out mainly in the USA and few in Australia.
According to the results of the realistic test, 85% of the children took the candy “lure” from the unknown young woman, while 72% followed her in the car. However, it has been reported in previous studies that the majority of young children (75% - 90%) follow an unknown person willingly, when he or she appears with an “attractive lure”, such as a toy, sweet or presented as a friend to the teacher or parent (Goldfarb et al., 2008; Marchand-Martella et al., 1996; Poche, Brouwer, & Swearingen, 1981) . It is concerning that across experiments and conditions of the present study, about 3/4 of the children left with the confederate and took the candy from the unknown young woman without feeling exposed to any danger. The fact that children didn’t know they are being assessed reflected their behavior in a realistic setting. Nevertheless, as it was mentioned before, in order to identify and avoid a potential danger, it is necessary to apply an appropriate test, which will seem very real, without however scaring or stressing children. This test could also be used as a tool for the objective assessment of personal safety skills, since there is the possibility of recording the reaction to a realistic condition (Fryer, Kraizer, & Miyoshi, 1987, Walsh et al., 2015) .
The results of the present study indicated that children’s theoretical knowledge of whether a situation is dangerous or not is not related to their reaction in a condition of a real-world setting. These findings are in accordance with the results of earlier studies which showed that there is a low correlation between what children answer and what they do in a real situation (Miltenberger et al., 1991; Carroll-Rowan & Miltenberger, 1994; Olsen-Woods et al., 1998) . Thus, it seems that even when the children knew what the correct answer to give to a possible situation of danger is, it was not always possible to “convey” the specific knowledge into a correct reaction, a finding that has been mentioned in a previous study (Goldfarb et al., 2008) . Additionally, it has been mentioned before, that even when children can describe or exhibit appropriate safety skills in a role-play, they cannot always put those abilities into practice when an adult is absent and children are unaware of the assessment (Himle et al., 2004; Miltenberger et al., 1991) .
Also, it seems that the children in the present study focused on the gift the confederate was offering. This finding is consistent with research using the simulated-risk paradigm, which shows that preschool children can be easily lured away (Goldfarb et al., 2008; Johnson et al., 2005; Li et al., 2020, White et al., 2018) . Moreover, in previous research it has been reported that a surprise toy, a candy or a cute pet, is particularly tempting for young children (Holcombe et al., 1995; Li et al., 2020) .
It is very likely that the large percentage of children in the present study followed “the unknown person” because this role was played by a kind, charming, young woman who did not cause them fear. Usually, in the minds of young children the stranger that could harm them is a man, big, wild, with a beard, dirty, old clothes and not good-looking. However it has been mentioned before that children trust more, unfamiliar adults who look like their familiar persons such as their parents or teachers (Zakharchenko et al., 2015) .
5. Conclusion
The present study revealed that the majority of preschool children reacted wrongly to the realistic test, no matter if they had knowledge about self-protection skills. This means that even though they knew how they should behave when they encounter a safety threat, when encountered a simulated safety threat that appeared real, children did not perform the safety skills. These results clearly indicate that children need to learn interpersonal safety skills. Practically, understanding children’s vulnerability to be lured by a stranger has important implications for parents and educators as they could plan and develop appropriate educational programs, so as to ensure children’s safety, which should be a priority within every modern society. Future research should consider teaching young children that it is not the strangers that are dangerous, the danger is the behavior of anyone, stranger or known adult who has not been authorized to take them and attempts to lure them away.