Vol 16. pp. 865-876. 1992 Primed in the [ I.S.A. All rightsreserved
0145-213-1/92$ 500 ~ .(X) ('op)mght 'c:,1992 PergamonPressLld
( "htM It,u~e & .Vegh'ct.
SEXUAL ABUSE PREVENTION EDUCATION FOR YOUNG CHILDREN: A COMPARISON OF TEACHERS AND PARENTS AS INSTRUCTORS SANDY K. WURTELE, LAURA C. KAST, AND ANASTASIA M. MELZER University of Colorado, Colorado Springs, CO
A b s t r a c t - - T h e purpose of this study was to compare teachers and parents as instructors of a personal safety program. O n e hundred seventy-two Head Start preschoolers were randomly assigned to a personal safety program taught by their teachers, parents, both teachers and parents, or to a general safety control program. Following program participation, children taught the personal safety program by their teachers, parents, or both, demonstrated greater knowledge about sexual abuse and higher levels o f personal safety skills compared with those in the control group. Gains in knowledge and skills were maintained at the 5-month follow up. Children taught by their parents showed greater improvements in recognizing inappropriate-touch requests and in their personal safety skills compared with children taught by their teachers, and children who received the program both at h o m e and .school were better able to recognize appropriate-touch requests and to demonstrate higher levels of personal safety skills compared with children taught only at school. The emotional costs associated with participating in the program were minimal, and both parents and children rated the program positively. The advantages of home-based instruction tbr young children are discussed and suggestions for future research are offered. Ke.l" l!'ords--Child sexual abuse. Prevention, Personal Satiety.
INTRODUCTION IN RESPONSE TO the widespread problem of child sexual abuse, a number of school-based prevention programs have been developed and implemented, primarily for children in the elementary grades. However, as a significant portion of sexual abuse victims are under the age of seven (Finkelhor, 1984), attention has now turned to prevention for a younger age group. Although preliminary research suggested that preschool-aged children are unable to learn sexual abuse prevention concepts (e.g., Borkin & Frank, 1986: Gilbert, Daro, Duerr, LeProhn, & Nyman, 1988), recent studies have shown that when taught using a behavioral approach (including modeling, rehearsal, and social reinforcement), young children in preschools can learn personal safety skills (Harvey, Forehand, Brown, & Holmes, 1988; Kraizer, Witte, & Fryer, 1989: Stilwell, Lutzker, & Greene, 1988; Wurtele, 1990: Wurtele, Kast, Miller-Perrin, & Kondrick, 1989). Given that not all young children are enrolled in preschools, yet remain vulnerable to sexual abuse, home-based prevention programs also need to be developed and evaluated. A limited nttmber of studies have attempted to determine the efficacy of parents as prevenThis project was funded by a FIRST Award (MI! 42795) from the National Institute of Mental Health. Received for publication July 3, 1991, final revision received September 30. 1991, accepted December 31. 1991. Reprint requests should be addressed to Sandy K. Wunelc, Department of Psychology, P.O. Box 7150, University of Colorado, Colorado Springs, CO 80933-7150. The authors wish to thank Judy Gibian-Mennenga. Joy Hines. Deborah Stinson. Heather Moxley. the intem'iewcrs, and the parents and teachers from ltead Start. 865
86~
s. Wurtele. I.. Kast. and A. Melzer
lion instructors. In two studies, Miltenberger and colleagues (Miltenberger & Thiesse-Dufl~,, 1988, Miltenberger. Thiesse-Duffy, Suda, Kozak, & Bruellman, 1990) found lhat 4- to 7-yearold children's performance did not improve following training by their parents with the Red Flag, Green Flag Prevention Book (a commercially available prevention training book: Rape & Abuse Crisis Center, 1986), but following behavioral skills training by experts, all the children reported the correct prevention responses. Although the authors suggested that parents are ineffective instructors of personal safely skills, this conclusion is premature given the studies" methodological problems. For example, the small sample size limits the generalizabilitv o f l h e results. Additionally. no check was made on whether parents followed the instruction to include rehearsal of ,safely skills. Furthermore, the expert-taught behavioral skills training phase always followed the parent-training phase, and thus its effectiveness may have depended on the prior presentation of the parent-taught portion of the program. Recenl work refutes lhe assumption that parents are ineffective instructors (Wurtele, Currier, Gillispie, & Franklin, 199 I, Wurtele, Gillispie, Currier, & Franklin. 1992). For example, in Wurtele et al. (1991), middle-class parents taught their children a personal safety program that incorporated behavioral rehearsal and self-reinforcement tbr mastery of knowledge and skills. Results suggested that when using a developmentally appropriate program that employs a behavioral approach, parents are indeed eflizctive instruclors. Children (aged 3 1/2 to 5 1/2 years) significantly improved their knowledge abut sexual abuse, and also improved their abilities to recognize inappropriate-touch requests, and to respond appropriately in these situations. Similar results were obtained in Wurlele el al. (1992), in which parents of children enrolled in a Head Start program were asked to leach the personal safety program at home. Preschoolers participated in either a home-based program, a school-based program, or an attention-control program. At posttesting, no significant ditt~rences were found on knowledge and skill tesls between groups of children taught by parents or teachers, and children in both of these groups demonstrated greater knowledge about sexual abuse and higher levels of personal safety skills compared with those in the control groups. These results suggest that parenls are as effective as teachers at teaching personal safety skills to preschool-aged children. Although recent studies have shown that as a group, preschool-aged children instructed in personal safety skills perform better than control children, few program participants, whether taught by teachers or parents, achieve 100% accuracy in recognizing and responding to potential abuse. Some program participants were unable to generalize their newly learned abilities to recognize and respond to the novel situations included in the assessment devices. Therefore, research eflbrts must be aimed at enhancing generalization of their knowledge and skills. In order for behavioral change to generalize, the training should occur over time, persons. and settings (Stokes & Baer, 1977). For example, Stokes, Baer, and Jackson (1974) found that when a greeting response was taught to retarded children by one trainer, the response did not generalize across experimenters. However, when a second experimenter trained the response, it generalized to over 20 members of the stall'. Griffiths and Craighead (1972) similarly programmed generalization across settings. A 30-year-old retarded woman exhibited correct speech articulation in a nontraining setting only after she was trained in two diflizrent seuings. These sludies suggest that by increasing the number of trainers and settings used when leaching skills, generalization will increase. The purpose of this study was to compare knowledge and skills of preschool children who had participated in a control program (C) or in a personal safely program taught by their teacher (T), or parent (P), or both leacher and parent (T+ P). We predicted that at postlesting, children in groups T, P, and T + P would have significantly more knowledge and skills than children in the C group. In addition, we predicted that the children in the T + P group would perform significantly better than children in the T and P groups, as these children would be taught by two dill~rent trainers in two difl~zrent settings (home and school), and thus should
Personal .safety
867
be better able to generalize their newly learned knowledge and skills to novel situations. Furthermore, we predicted that enhanced knowledge and skills would be maintained over time, and would be achieved with minimal or no negative emotional or behavioral effects.
METHOD
SII[~]t'cL~ One hundred and seventy-two Head Start preschoolers (73 boys and 99 girls) in a Colorado community of 290,000 served as subjects (mean age = 55.4 months). Forty-eight percent of the children came from two-parent families, 49% from single-parent homes headed by the mothers, 2% from father-headed homes, and 1% from families headed by "'others" (e.g., grandparents, aunts, uncles). Fifty-three percent of the children were White, 23% were Black, 23% were Hispanic, and 1% were Asian. Fifty-nine percent of the children's families had an income level of $9,999 or less per year, and the remainder were reported between $10,000 and $30,000. Parental consent and child assent were obtained before the initial assessment. Informed consent forms were sent to all parents: 91% of parents gave permission for their children to participate in the program and to be interviewed.
Instructor.s, Interviewers, and Caseworkers Treatment programs were taught by either parents or classroom teachers, and control programs were taught by teaching assistants. All classroom instructors were females and had substantial teaching experience. The interview team consisted of 25 female undergraduate. one male undergraduate, and two female graduate students who successfully completed eight hours of interviewer training, during which time they learned to administer the child dependent measures in a standard fashion. Interviewers were blinded to the hypothesis and group assignment of the children. They interviewed different children at pretest, posttest, and follow-up testing, and were supervised throughout the project. Fourteen female undergraduate students were also trained as caseworkers and assigned to families that taught the program at home. They contacted the parents by telephone during program presentation to ensure compliance, enhance motivation, and answer questions.
Child Dependent Measures Two measures were employed with the child subjects; the "What If" Situations Test (WIST), and the Personal Safety Questionnaire (PSQ). The WIST measures children's abilities to recognize and respond in hypothetical abusive situations (but does not measure children's abilities to apply their personal safety skills in real-life situations); the PSQ measures children's knowledge about sexual abuse. Both measures have been used in previous studies (see Wurtele, 1990; Wurtele et al., 1991 ; Wurtele et al., 1992). The measures and procedures received approval from the University of Colorado Institutional Review Board. "What If" Situations Test (WIST). The WIST consists of six brief vignettes: three describe appropriate requests to touch or look at the child's genitals (e.g., a physician wanting to touch the child's injured private parts), and three describe inappropriate requests (e.g., a neighbor wanting to take pictures of the child's private parts). (None of these specific situations was taught in the personal safety program). In order to determine children's ability to recognize abuse, after each vignette children were asked, "Would it be okay for [person in vignette] to [do activity requested]?" Correct responses receive one point. Responses to this question yield two Recognition scores: (a) an Appropriate Request Recognition Score, which measures a
868
s. Wurtele, L. Kast, and A. Melzer
child's ability to recognize the three appropriate-touch requests (range = 0-3); and (b) an Inappropriate Request Recognition Score, which measures a child's ability to recognize the three inappropriate-touch requests (range = 0-3). (For Appropriate and Inappropriate Request Recognition Scores, o n e - m o n t h test-retest reliabilities, based on Controls' pre and posttest scores, were .63 and .70, respectively, both ps < .01 ). For inappropriate request vignettes correctly recognized, children were asked: (a) " W h a t would you say to [person in vignette]'?", (b) " W h a t would you do?", (c) " W o u l d you tell anyone about [the situation]?", if so, "'Who would you tell?", and (d) " W h a t would you say to [person n a m e d in c]?.'" Responses to each of these questions receive 0 to 2 points, depending on the effectiveness of the responses. For example, a tentative verbal refusal (e.g., "'I d o n ' t want to") receives 1 point, whereas a definite refusal ( e.g., "No! D o n ' t do that") receives 2 points. Children's responses to question c (i.e., " W h o would you tell?") were also categorized according to type of resource person mentioned (i.e., parents, teachers, police, relatives, professionals). Children's scores on questions "'a'" through "d'" ( m a x i m u m = 8 points per vignette) were s u m m e d across the three inappropriaterequest vignettes for a Total Skill Score of 24 (one-month test-retest reliability, based on Controls" pre and posttest scores, was .86, p < .01). Children's responses were scored by two independent raters, and interrater reliability (Kappa coefficient) was .93.
Personal Sali'ty Questionnaire (PSQ,). The PSQ consists of 10 questions designed to assess children's knowledge about sexual abuse (e.g., "'Is it okay for kids to touch a bigger person's private parts?"). Children responded by saying "Yes," "'No" or "I don't know." Correct answers receive I point (scores range from 0 to 10). One m o n t h test-retest reliability (Pearson r based on Controls" pre and posttest scores) was .70, p < .01. To assess attitudes toward sexuality, children's responses to two questions were analyzed separately (i.e., "Is it okay for kids to touch their own private parts?": "'I)o you like your private parts?"). For these two items, a positive attitude toward sexuality was reflected in a "yes" response. At posttesting. children were also asked if they liked the program. Teacher and Parent D(Twndent Mea,sures. l'eacher Perception Questionnaire ('IT()). The T P Q consists of a checklist of 17 behaviors: eight positivc behaviors (e.g., "'scems self-confident") and nine negative behaviors (e.g. "'cries easily"). For each child, the teacher was asked to indicate if any of the behaviors had increased since the child had participated in the program (yes or no), and if yes. was the increase a problem. Parent Perc~7~tion Questionnaire (PPQfl. The PPQ consists of the same checklist of 17 behaviors as the T P Q with three additional items ("expresses need for privacy": "'has difficulty separating from parent": "has sleeping problems"). Parents were asked to indicate if any of the behaviors increased after receiving the program (yes or no), and if yes, was the increase a problem. Parents also rated the global effect of the program on their children (good, bad, or no effect), indicated whether they would allow their children to participate again in a similar program on body safety, and those parents who taught the program evaluated it on a n u m b e r of dimensions. Procedure All the children were pretested (using the W I S T and the PSQ) one week before program presentation. They were then randomly assigned, within a classroom, to one of four personal safety training groups [Teacher (T), n = 41; Parent (P), n = 44: Teacher 1- Parent (T ~-P). n 43: or Control (C). n - 44]. Within a classroom, half the children were taught the personal safety program by their leachers (groups T and T + P ) , and the other halt" were taught the
Personal ml+ety
869
general safety control program (groups P and C). Children in the T group received the personal safety program only at school, children in the P group were taught the program only at home, and children in the T + P group received the program both at school and home (after receiving each day of the BST at school, they were taught the parent version at home). Parents o f subjects assigned to groups P and T + P were contacted by telephone, informed of the nature of the study, and asked if they would teach the program at home. All parents initially agreed to teach the program: however, one parent later chose not to teach it, one parent was hospitalized and could not teach it, and 11 parents were excluded as subjects due to failure to meet the criteria set for program completion (following Kazdin's 1986 recommendation to ensure treatment integrity), leaving 87 parents in the P and the T + P groups. At school, the programs were presented in 15-minute sessions on four consecutive school days, and were led by classroom teachers and assistants who read from scripts and used pictures to depict stories. Controls were taught various safety skills related to fire prevention, pedestrian safety, poison safety, etc., with no mention of sexual abuse. At school, the Behavioral Skills Training Program (BST) (Teacher version: Wurtele, 1986) was presented to children in groups T and T+P. The BST is a 4-day program that teaches children personal safety skills from a behavioral perspective. Working with small groups (ranging in size from 4 to 10), teachers instructed children: (a) that they are the bosses of their bodies: (b) to identify the location of their "private parts": (c) that it is appropriate for doctors, nurses, or parents to touch children's private parts for health or hygiene reasons: (d) that otherwise, it is not okay to have their private parts touched or looked at by a bigger person. especially if the person wants them to keep it a secret: (e) that it is wrong to be forced to touch a bigger person's private parts: and (f) that an adult's inappropriate touching of the child's private parts is never the child's fault. Children were taught the rule that, "It's not okay for a bigger person to touch or look at my private parts" (unless they need help, for example, if their private parts get hurt or are sick). The program included a number of stories abut children in both innocuous and potentially dangerous situations with various people (e.g., babysitters, relatives). Children practiced discriminating between appropriate and inappropriate touch requests, and were taught (via modeling, rehearsal, praise, and feedback) the appropriate verbal (e.g., say "No?") and motoric responses (e.g., get away, tell someone) to make in the inappropriate situations. Children received stickers after completing each day of the program. Like the Teacher version, the BST Parent version included a script and accompanying pictures (to enhance children's attention to the program, different stories were used in the T + P parent version). The Parent version also included a "'Token T i m e " packet, stickers, and crayons for children to use to color the pictures. The " T o k e n T i m e " packet contained 29 objectives of the program. As the concept or skill was mastered, the children placed a sticker of choice next to the description of the objective. We included this device to help parents track their children's progress, to maintain their children's attention, and to provide self-reinforcement for the children (Brigham & Stoerzinger, 1974). Parents were also encouraged to praise their children when they acquired the new skills, Parents (77% of whom were mothers, 3% fathers, 11% mothers and fathers, and 9% mothers and other t~amily members) taught the BST at home, and caseworkers maintained telephone contact with them to answer questions, enhance motivation, and ensure compliance. We attempted to measure parental compliance in teaching the program in several ways: (a) blank spaces were provided intermittently throughout the script for parents to fill in their children's answers, allowing us to check for completion and rehearsal: (b) parents were asked to record the starting and ending times for each day, revealing the total time spent teaching the program (which averaged 2-1/2 hours): (c) "'Token T i m e " packets were checked for completion: and (d) caseworkers monitored parent's progress. On the two consecutive school days after the program had been taught, all children were
S. Wurtele. 1.. Kast, and A. Melzer
870
Table I. Means (and Standard I)eviations) of Four Dependent Measures by Group at Pre-, Post-, and Follo~-up Testing Group
Pretesl
Posttest
Follow-up
Inappropriate Request Recognition (range 0 - 3 ) r P T ~P ("
1.37 1.09 1.47 1.41
(I.39) (1.27) (I.29) (I.40)
2.29 (1.15~ 2.71 (0.67) 2.65 (0.84) 1.77(I.27)
2.61 (0.89) 2.78 (0.57) 2.85 (0.53) 1.95(I.26)
WIS'I Total Skill (range
0145-213-1/92$ 500 ~ .(X) ('op)mght 'c:,1992 PergamonPressLld
( "htM It,u~e & .Vegh'ct.
SEXUAL ABUSE PREVENTION EDUCATION FOR YOUNG CHILDREN: A COMPARISON OF TEACHERS AND PARENTS AS INSTRUCTORS SANDY K. WURTELE, LAURA C. KAST, AND ANASTASIA M. MELZER University of Colorado, Colorado Springs, CO
A b s t r a c t - - T h e purpose of this study was to compare teachers and parents as instructors of a personal safety program. O n e hundred seventy-two Head Start preschoolers were randomly assigned to a personal safety program taught by their teachers, parents, both teachers and parents, or to a general safety control program. Following program participation, children taught the personal safety program by their teachers, parents, or both, demonstrated greater knowledge about sexual abuse and higher levels o f personal safety skills compared with those in the control group. Gains in knowledge and skills were maintained at the 5-month follow up. Children taught by their parents showed greater improvements in recognizing inappropriate-touch requests and in their personal safety skills compared with children taught by their teachers, and children who received the program both at h o m e and .school were better able to recognize appropriate-touch requests and to demonstrate higher levels of personal safety skills compared with children taught only at school. The emotional costs associated with participating in the program were minimal, and both parents and children rated the program positively. The advantages of home-based instruction tbr young children are discussed and suggestions for future research are offered. Ke.l" l!'ords--Child sexual abuse. Prevention, Personal Satiety.
INTRODUCTION IN RESPONSE TO the widespread problem of child sexual abuse, a number of school-based prevention programs have been developed and implemented, primarily for children in the elementary grades. However, as a significant portion of sexual abuse victims are under the age of seven (Finkelhor, 1984), attention has now turned to prevention for a younger age group. Although preliminary research suggested that preschool-aged children are unable to learn sexual abuse prevention concepts (e.g., Borkin & Frank, 1986: Gilbert, Daro, Duerr, LeProhn, & Nyman, 1988), recent studies have shown that when taught using a behavioral approach (including modeling, rehearsal, and social reinforcement), young children in preschools can learn personal safety skills (Harvey, Forehand, Brown, & Holmes, 1988; Kraizer, Witte, & Fryer, 1989: Stilwell, Lutzker, & Greene, 1988; Wurtele, 1990: Wurtele, Kast, Miller-Perrin, & Kondrick, 1989). Given that not all young children are enrolled in preschools, yet remain vulnerable to sexual abuse, home-based prevention programs also need to be developed and evaluated. A limited nttmber of studies have attempted to determine the efficacy of parents as prevenThis project was funded by a FIRST Award (MI! 42795) from the National Institute of Mental Health. Received for publication July 3, 1991, final revision received September 30. 1991, accepted December 31. 1991. Reprint requests should be addressed to Sandy K. Wunelc, Department of Psychology, P.O. Box 7150, University of Colorado, Colorado Springs, CO 80933-7150. The authors wish to thank Judy Gibian-Mennenga. Joy Hines. Deborah Stinson. Heather Moxley. the intem'iewcrs, and the parents and teachers from ltead Start. 865
86~
s. Wurtele. I.. Kast. and A. Melzer
lion instructors. In two studies, Miltenberger and colleagues (Miltenberger & Thiesse-Dufl~,, 1988, Miltenberger. Thiesse-Duffy, Suda, Kozak, & Bruellman, 1990) found lhat 4- to 7-yearold children's performance did not improve following training by their parents with the Red Flag, Green Flag Prevention Book (a commercially available prevention training book: Rape & Abuse Crisis Center, 1986), but following behavioral skills training by experts, all the children reported the correct prevention responses. Although the authors suggested that parents are ineffective instructors of personal safely skills, this conclusion is premature given the studies" methodological problems. For example, the small sample size limits the generalizabilitv o f l h e results. Additionally. no check was made on whether parents followed the instruction to include rehearsal of ,safely skills. Furthermore, the expert-taught behavioral skills training phase always followed the parent-training phase, and thus its effectiveness may have depended on the prior presentation of the parent-taught portion of the program. Recenl work refutes lhe assumption that parents are ineffective instructors (Wurtele, Currier, Gillispie, & Franklin, 199 I, Wurtele, Gillispie, Currier, & Franklin. 1992). For example, in Wurtele et al. (1991), middle-class parents taught their children a personal safety program that incorporated behavioral rehearsal and self-reinforcement tbr mastery of knowledge and skills. Results suggested that when using a developmentally appropriate program that employs a behavioral approach, parents are indeed eflizctive instruclors. Children (aged 3 1/2 to 5 1/2 years) significantly improved their knowledge abut sexual abuse, and also improved their abilities to recognize inappropriate-touch requests, and to respond appropriately in these situations. Similar results were obtained in Wurlele el al. (1992), in which parents of children enrolled in a Head Start program were asked to leach the personal safety program at home. Preschoolers participated in either a home-based program, a school-based program, or an attention-control program. At posttesting, no significant ditt~rences were found on knowledge and skill tesls between groups of children taught by parents or teachers, and children in both of these groups demonstrated greater knowledge about sexual abuse and higher levels of personal safety skills compared with those in the control groups. These results suggest that parenls are as effective as teachers at teaching personal safety skills to preschool-aged children. Although recent studies have shown that as a group, preschool-aged children instructed in personal safety skills perform better than control children, few program participants, whether taught by teachers or parents, achieve 100% accuracy in recognizing and responding to potential abuse. Some program participants were unable to generalize their newly learned abilities to recognize and respond to the novel situations included in the assessment devices. Therefore, research eflbrts must be aimed at enhancing generalization of their knowledge and skills. In order for behavioral change to generalize, the training should occur over time, persons. and settings (Stokes & Baer, 1977). For example, Stokes, Baer, and Jackson (1974) found that when a greeting response was taught to retarded children by one trainer, the response did not generalize across experimenters. However, when a second experimenter trained the response, it generalized to over 20 members of the stall'. Griffiths and Craighead (1972) similarly programmed generalization across settings. A 30-year-old retarded woman exhibited correct speech articulation in a nontraining setting only after she was trained in two diflizrent seuings. These sludies suggest that by increasing the number of trainers and settings used when leaching skills, generalization will increase. The purpose of this study was to compare knowledge and skills of preschool children who had participated in a control program (C) or in a personal safely program taught by their teacher (T), or parent (P), or both leacher and parent (T+ P). We predicted that at postlesting, children in groups T, P, and T + P would have significantly more knowledge and skills than children in the C group. In addition, we predicted that the children in the T + P group would perform significantly better than children in the T and P groups, as these children would be taught by two dill~rent trainers in two difl~zrent settings (home and school), and thus should
Personal .safety
867
be better able to generalize their newly learned knowledge and skills to novel situations. Furthermore, we predicted that enhanced knowledge and skills would be maintained over time, and would be achieved with minimal or no negative emotional or behavioral effects.
METHOD
SII[~]t'cL~ One hundred and seventy-two Head Start preschoolers (73 boys and 99 girls) in a Colorado community of 290,000 served as subjects (mean age = 55.4 months). Forty-eight percent of the children came from two-parent families, 49% from single-parent homes headed by the mothers, 2% from father-headed homes, and 1% from families headed by "'others" (e.g., grandparents, aunts, uncles). Fifty-three percent of the children were White, 23% were Black, 23% were Hispanic, and 1% were Asian. Fifty-nine percent of the children's families had an income level of $9,999 or less per year, and the remainder were reported between $10,000 and $30,000. Parental consent and child assent were obtained before the initial assessment. Informed consent forms were sent to all parents: 91% of parents gave permission for their children to participate in the program and to be interviewed.
Instructor.s, Interviewers, and Caseworkers Treatment programs were taught by either parents or classroom teachers, and control programs were taught by teaching assistants. All classroom instructors were females and had substantial teaching experience. The interview team consisted of 25 female undergraduate. one male undergraduate, and two female graduate students who successfully completed eight hours of interviewer training, during which time they learned to administer the child dependent measures in a standard fashion. Interviewers were blinded to the hypothesis and group assignment of the children. They interviewed different children at pretest, posttest, and follow-up testing, and were supervised throughout the project. Fourteen female undergraduate students were also trained as caseworkers and assigned to families that taught the program at home. They contacted the parents by telephone during program presentation to ensure compliance, enhance motivation, and answer questions.
Child Dependent Measures Two measures were employed with the child subjects; the "What If" Situations Test (WIST), and the Personal Safety Questionnaire (PSQ). The WIST measures children's abilities to recognize and respond in hypothetical abusive situations (but does not measure children's abilities to apply their personal safety skills in real-life situations); the PSQ measures children's knowledge about sexual abuse. Both measures have been used in previous studies (see Wurtele, 1990; Wurtele et al., 1991 ; Wurtele et al., 1992). The measures and procedures received approval from the University of Colorado Institutional Review Board. "What If" Situations Test (WIST). The WIST consists of six brief vignettes: three describe appropriate requests to touch or look at the child's genitals (e.g., a physician wanting to touch the child's injured private parts), and three describe inappropriate requests (e.g., a neighbor wanting to take pictures of the child's private parts). (None of these specific situations was taught in the personal safety program). In order to determine children's ability to recognize abuse, after each vignette children were asked, "Would it be okay for [person in vignette] to [do activity requested]?" Correct responses receive one point. Responses to this question yield two Recognition scores: (a) an Appropriate Request Recognition Score, which measures a
868
s. Wurtele, L. Kast, and A. Melzer
child's ability to recognize the three appropriate-touch requests (range = 0-3); and (b) an Inappropriate Request Recognition Score, which measures a child's ability to recognize the three inappropriate-touch requests (range = 0-3). (For Appropriate and Inappropriate Request Recognition Scores, o n e - m o n t h test-retest reliabilities, based on Controls' pre and posttest scores, were .63 and .70, respectively, both ps < .01 ). For inappropriate request vignettes correctly recognized, children were asked: (a) " W h a t would you say to [person in vignette]'?", (b) " W h a t would you do?", (c) " W o u l d you tell anyone about [the situation]?", if so, "'Who would you tell?", and (d) " W h a t would you say to [person n a m e d in c]?.'" Responses to each of these questions receive 0 to 2 points, depending on the effectiveness of the responses. For example, a tentative verbal refusal (e.g., "'I d o n ' t want to") receives 1 point, whereas a definite refusal ( e.g., "No! D o n ' t do that") receives 2 points. Children's responses to question c (i.e., " W h o would you tell?") were also categorized according to type of resource person mentioned (i.e., parents, teachers, police, relatives, professionals). Children's scores on questions "'a'" through "d'" ( m a x i m u m = 8 points per vignette) were s u m m e d across the three inappropriaterequest vignettes for a Total Skill Score of 24 (one-month test-retest reliability, based on Controls" pre and posttest scores, was .86, p < .01). Children's responses were scored by two independent raters, and interrater reliability (Kappa coefficient) was .93.
Personal Sali'ty Questionnaire (PSQ,). The PSQ consists of 10 questions designed to assess children's knowledge about sexual abuse (e.g., "'Is it okay for kids to touch a bigger person's private parts?"). Children responded by saying "Yes," "'No" or "I don't know." Correct answers receive I point (scores range from 0 to 10). One m o n t h test-retest reliability (Pearson r based on Controls" pre and posttest scores) was .70, p < .01. To assess attitudes toward sexuality, children's responses to two questions were analyzed separately (i.e., "Is it okay for kids to touch their own private parts?": "'I)o you like your private parts?"). For these two items, a positive attitude toward sexuality was reflected in a "yes" response. At posttesting. children were also asked if they liked the program. Teacher and Parent D(Twndent Mea,sures. l'eacher Perception Questionnaire ('IT()). The T P Q consists of a checklist of 17 behaviors: eight positivc behaviors (e.g., "'scems self-confident") and nine negative behaviors (e.g. "'cries easily"). For each child, the teacher was asked to indicate if any of the behaviors had increased since the child had participated in the program (yes or no), and if yes. was the increase a problem. Parent Perc~7~tion Questionnaire (PPQfl. The PPQ consists of the same checklist of 17 behaviors as the T P Q with three additional items ("expresses need for privacy": "'has difficulty separating from parent": "has sleeping problems"). Parents were asked to indicate if any of the behaviors increased after receiving the program (yes or no), and if yes, was the increase a problem. Parents also rated the global effect of the program on their children (good, bad, or no effect), indicated whether they would allow their children to participate again in a similar program on body safety, and those parents who taught the program evaluated it on a n u m b e r of dimensions. Procedure All the children were pretested (using the W I S T and the PSQ) one week before program presentation. They were then randomly assigned, within a classroom, to one of four personal safety training groups [Teacher (T), n = 41; Parent (P), n = 44: Teacher 1- Parent (T ~-P). n 43: or Control (C). n - 44]. Within a classroom, half the children were taught the personal safety program by their leachers (groups T and T + P ) , and the other halt" were taught the
Personal ml+ety
869
general safety control program (groups P and C). Children in the T group received the personal safety program only at school, children in the P group were taught the program only at home, and children in the T + P group received the program both at school and home (after receiving each day of the BST at school, they were taught the parent version at home). Parents o f subjects assigned to groups P and T + P were contacted by telephone, informed of the nature of the study, and asked if they would teach the program at home. All parents initially agreed to teach the program: however, one parent later chose not to teach it, one parent was hospitalized and could not teach it, and 11 parents were excluded as subjects due to failure to meet the criteria set for program completion (following Kazdin's 1986 recommendation to ensure treatment integrity), leaving 87 parents in the P and the T + P groups. At school, the programs were presented in 15-minute sessions on four consecutive school days, and were led by classroom teachers and assistants who read from scripts and used pictures to depict stories. Controls were taught various safety skills related to fire prevention, pedestrian safety, poison safety, etc., with no mention of sexual abuse. At school, the Behavioral Skills Training Program (BST) (Teacher version: Wurtele, 1986) was presented to children in groups T and T+P. The BST is a 4-day program that teaches children personal safety skills from a behavioral perspective. Working with small groups (ranging in size from 4 to 10), teachers instructed children: (a) that they are the bosses of their bodies: (b) to identify the location of their "private parts": (c) that it is appropriate for doctors, nurses, or parents to touch children's private parts for health or hygiene reasons: (d) that otherwise, it is not okay to have their private parts touched or looked at by a bigger person. especially if the person wants them to keep it a secret: (e) that it is wrong to be forced to touch a bigger person's private parts: and (f) that an adult's inappropriate touching of the child's private parts is never the child's fault. Children were taught the rule that, "It's not okay for a bigger person to touch or look at my private parts" (unless they need help, for example, if their private parts get hurt or are sick). The program included a number of stories abut children in both innocuous and potentially dangerous situations with various people (e.g., babysitters, relatives). Children practiced discriminating between appropriate and inappropriate touch requests, and were taught (via modeling, rehearsal, praise, and feedback) the appropriate verbal (e.g., say "No?") and motoric responses (e.g., get away, tell someone) to make in the inappropriate situations. Children received stickers after completing each day of the program. Like the Teacher version, the BST Parent version included a script and accompanying pictures (to enhance children's attention to the program, different stories were used in the T + P parent version). The Parent version also included a "'Token T i m e " packet, stickers, and crayons for children to use to color the pictures. The " T o k e n T i m e " packet contained 29 objectives of the program. As the concept or skill was mastered, the children placed a sticker of choice next to the description of the objective. We included this device to help parents track their children's progress, to maintain their children's attention, and to provide self-reinforcement for the children (Brigham & Stoerzinger, 1974). Parents were also encouraged to praise their children when they acquired the new skills, Parents (77% of whom were mothers, 3% fathers, 11% mothers and fathers, and 9% mothers and other t~amily members) taught the BST at home, and caseworkers maintained telephone contact with them to answer questions, enhance motivation, and ensure compliance. We attempted to measure parental compliance in teaching the program in several ways: (a) blank spaces were provided intermittently throughout the script for parents to fill in their children's answers, allowing us to check for completion and rehearsal: (b) parents were asked to record the starting and ending times for each day, revealing the total time spent teaching the program (which averaged 2-1/2 hours): (c) "'Token T i m e " packets were checked for completion: and (d) caseworkers monitored parent's progress. On the two consecutive school days after the program had been taught, all children were
S. Wurtele. 1.. Kast, and A. Melzer
870
Table I. Means (and Standard I)eviations) of Four Dependent Measures by Group at Pre-, Post-, and Follo~-up Testing Group
Pretesl
Posttest
Follow-up
Inappropriate Request Recognition (range 0 - 3 ) r P T ~P ("
1.37 1.09 1.47 1.41
(I.39) (1.27) (I.29) (I.40)
2.29 (1.15~ 2.71 (0.67) 2.65 (0.84) 1.77(I.27)
2.61 (0.89) 2.78 (0.57) 2.85 (0.53) 1.95(I.26)
WIS'I Total Skill (range
