“ ——:————————'—*—————— W M mumv veer. “1'7 "“M'“""‘“E“’““*”‘
xsnarv-‘aa‘u ‘ ‘ .t‘fw’fi‘l‘m " - . .

‘ ‘
.mauwm I}. r ous‘us- I tsssouhk-
Lug spa M , '
3‘22““ N l§~ q“...- .. ‘.~ -‘ .\ \‘M M- a. _

 

 
  

  

. an.
rs. :J." .5; “\. u‘. -
‘ . _ ,‘ ,3,- 3. £.;.-. .11.; ‘ L :3..-5‘.¢~t~ »§I~
. _.-, 4- , . .- ruby. -
....-. .-.s - .\-~. . .. .2 . --. -.."'.. - ‘ . . .' .~. .- -- -- < 3'3 '«I "iii...” J. V J .‘N "n."- 4-.
&.~o-‘JL-:o-\Ivs‘ ~‘;:.‘-C ..._“ -.‘~ _.‘u,.l -gu . .g.:.. :KO -I.. a, ‘ ,.. f j .r . —‘ C_-5-‘o 0‘ GO-‘-:“l.d
's . I“... .qI-L..¢a J.‘ ..9. O.—l. _ .‘v .' 0-1 'I -‘ - I .' ‘-' .‘0 -o‘&<""‘ ”WV-‘5“ "I. "“9 .5
'J."At 31%|: u..--\ _\‘ .‘. _ ---L"J‘.t»-. ._ i-‘;‘--:".::.:' . .“ _ g; 1.9. ‘.'-.- .. ' . \' - - . . 'aAN 5,-‘Ls’0. 5“ l L. no.4 6'
V\.- .g,..g..4.', - - . .L; ‘ 2 .‘TIIV‘ fi--. -_. I-LI‘"»I s". LII-:i‘figm ‘s" --~h :8
‘ ‘ 1V»'. “¥.~, ‘.' o_, . ‘ - - .M '-. . . . 99 '. ..'. O..&.\ I a “My. \h 5‘5”“...
, . _ - ' - ..~I’I- MO‘JIN“ ‘."-" w

‘0 -.._ t 155:. IE..- fii’fi'ss
I

    

~. -...§'--.-.. 2P1...“ $.33";-
\ 0 . o- 'I. :...1x':.. .. Lk“ll“...~‘~l
~~-‘...A.‘-.l.u c..-a .. ~‘ ‘
- . . -' .‘c ,._‘ “. ‘-‘\,..,..§,.o.v ..- ‘-,
" :"‘:. w-s -"'~"~1‘~ ' - “MM
0‘ ' .3 o . Q -. - . 03‘~010~.~\§‘ u &‘ b-I-K
‘- y s-‘. 0.4 ..L y“.|< .?\. 9 ' 4‘ 0 fi" ‘-"-‘l
‘ I... I'- I ~ ' L :‘h‘ ..:9‘:A§V tkuI.-.h‘
-. V\---.:§\.I' 3M}; ‘1‘. yq‘i 4“.
- . '.1.I - . \? ‘ “(Ci-5;“:t ~ ~_I-\ 0"." -
.- ,' n~“]x.:. 00“". ”t .\
--’- -.._°oIY. g. :o.p§-r .u3~\..§-

    

 

  

  
   
  
    
  
   
 
      

 

- ‘-
.o::u.f%-Z:h,s,:. it I. L‘qud‘.&*\.\-a.l~\-5“ . ‘15“.
’- ." JQI '4‘..-.:~.i.m... u... -
06 t 7 -~ - . t. . . J-
. . .Q«.-.. .. .1... ~ .1“ Q .21“ '1‘:-‘~:Qt~~i
~ - s o . s .s- - . _
. -- - . ‘ . ’ - IN ‘.I ‘ Ian“ t41“. 5‘ u .9 I ~od-u.‘ . . ;~ .“_
4 3 -; . 00...; . o «I. . h I .5. ‘:..u0..'.i‘ .n.
.. - ~ - . . ~ ‘ 7-5 ‘ ‘ I s".u
r . J I .I I. ' h - i {"9" 2‘9 «$321.
9. - . .- . . . - “‘0 u ' . .. --
' . _ 3._ ~. j». ._. _° .. ‘m‘. 5"; (8-- g: 9 Nu!
.. j -' L 1' L . L5 .5 - i _ . d ' skigu :I't'b‘
L. , l 9 ' h I t I t “ '. . "~¢.‘y...-‘~
\ . , . . , I n u \ ‘.
'I . ..J K .0 I. . O o ' ‘ 9 LL- “u‘ :-\‘3‘\'é
.. J ' - ‘ I . ‘ 3 '- ' M. i " ' 5~Qfi~-"o
- . _—. .v I O .‘ ~ - “a. ' Q‘s‘g. Jay. ”Q.
_ s

\ '.. icfinkb‘ Uk“~~ 'uku‘...f.;
. f¢s~ms UD‘IoQ-o

 

«- .35; ‘ac.~:~£M~-—h §
.: 'k‘ ~:. _;:£Jc- “ ‘.-$Lq . s ' ~03.‘.3L
51...”.1. u: .~. ‘1'“ "‘.”""“‘L‘.’.¢..¢ J m“ “'1‘"

. " * ....,
"~5 ‘ :0“, ‘fi~“5~:::~“'9“ .t: -.

‘5»..-
.. ~.\..t§-.,'Im

.-
43.613182”..x“‘~::-~'~- “aviary. 1‘51““ “

 
  
  

   

      
  
  
    
    
     
          
        
 
  
        
 
 
  
 
  
 
   
   
      
 
 
    
   
 
   
   
 
 

   

  

  
  
 
  

 

    
  

 

 

       
  
 
  

      

       
  

     
    

  

. :2:
.- ... ....'
N‘ 0—. .0- ~Cq - a. 5
":31“. "2; a" .- - ‘JIV‘§‘.A Va} {‘1‘ “M, .
l..._. ‘6 . 41‘ . 5‘1
fr’hflflh.’ .3'2I1: 1": ‘,L‘§" “t“: ‘1": a .‘1- :. ‘3 .5. 0.? ‘3‘.
- -fi. ‘ . I' (‘4 - -. ‘5...QO ' a
k .. '1‘»! I 1£uh-£"xl1 .h‘fl.." hh~umai
5!." Fa.» ~??, ‘ Lu IA'.‘ 13:3.0‘.“ J. -.
I' '» - . '~t.‘ I
cad. I“ .3 :9 . vii}? -/ ’0 up 1%.? fiz‘r'f‘iioi—
. ~L: -.n’o.-r-' _ 34
I «v «w. - "-va.. .9. “‘3. L .
V .' n- ”7.? A :g: ‘5‘“‘: Ifgihkxk: gut: .';‘-.‘H‘:. ‘v‘
' ‘ n O ‘ 0‘ o‘ I ~“‘ O ““1“I.‘
-._..Is an .. . ..I..? 1.1. or;\— "g‘ht -3“.g" dig... .h-a... ukI-r 3‘4. .‘L (. 3.. 3‘
O '45... ‘.;H' ‘ .;‘~3;,:'m3-:‘I ~18? “g” I 1' ‘I-“fluta
. s _a c . "- I. ~ . ‘ v ‘ I d
- .3“ O.‘~'~\-'b¢e.5“.: “:11?! as.)”.- Q " fl 1‘ ~ .' id.” 3.1’u’.‘£iq:;:‘1;;“1..§.‘.~‘
"39‘ If“), . d ‘ 9.”- ‘3... 2': 3;:g? ?.:;.‘. ~;-" 2* . , n H‘ n;‘ . . “‘
~ -~- 0- . v -‘ ' .. - - ‘n 9-.- ‘ < ' -
.ui‘; U; 9.- ’8 07?, l-‘:.‘ P -.’.J {“‘3‘.’ ',.' 2; s; ‘uscat-F'ngu. "w: ‘1
I‘ '- L . .. I. q . . ', s-‘h‘L a- . . ,o - g. 7..
:;:-::\:.. “it". .114?!» Efii'i“ Adi-la ' i. , .p_ I .f..1§ O’N": ‘3." 4"
. . 353:2." "44»: a; "‘1" 53")»; " 1;," -.. “It‘- rear": ““3
.. _. I ‘. ._ I. . " .
‘ . ‘A , .- . '_ !' . '1'4'40 .n: .0. ..'.. ' '
.9, “’2': '42:, "ILL. “.'b-.I — 4-. ’94... ; ‘a'._ l h *1:
w ' ‘.’ ' . - ~ ' . uv:.=-O- ,‘ .‘
J)“: L I ' . k - I?” 'I .";' '.'-.'3—‘o
{luffica ,. .9 .!.§. "' ' tin :
6mg .v 3'. - .' ..-; .3}: y
".1. c -x-y -53“: .N - d-...
'2‘ ‘ ’ -~...-5 v 4 Jun: .13. .1 --
' - <- .‘ r ' ' ~ ' .
v '.‘u.‘~ (3., 5v -._,-'.. .L; ‘. .‘\._"‘IQ..0y-.a (’5‘
.I_- . n v Y-f it,“ "- e‘l‘.‘ ‘9‘ «Eva-xv ‘EA I- . .
l :l a ’1. g d
. L .

  
   

    

v _ a... I
7' 32‘£ff">l.;?’< 1 ... -‘ ='=“““~
m} ,ngfir atria? 1
4‘! In“... P.- ‘a -
«3!» {qx' .‘ ‘ -
[0:34 1.: :fl::“‘. -"I" 1 ‘
.. W . 1 - . ~~
x» In“ 5W5 J'ké .1”, -="’X".y".'.;v‘-u 4v“.'r’1;'.-.‘.~-~-*a“:..4: 4---’
-. . 0., .._..-._‘.. . '-‘ .V w ‘4’"9-1‘ ,-'.; -.-o
......:'=-;.z::x:...r.? *‘ «- F .. um" , - - ' ...
‘- 'r-I

«Fr. .. part; -. H.331“ - n; * -. .. war-”mm «
" ' .,. .- —- ' 0‘
3 a. aha Adar". ““1“.
( 4;“?'."..ch
‘I-u-wn

 
 

. r.
. 3'“:
0‘

  

> 'v.‘

: Ila. - . -‘ a", 5
_ 0‘ y ;;r .'A' “.4 J‘: .I
m, . .’:+a::.p:=':yrts.w3~ '-

. d.: .6 4 H - .»’. - ‘ ~ '7.- II
I- .32‘ L “*‘u‘wx-finnv. _
r .-' ‘. . .‘rre. , ,_ .

mw- 4%" "'
Egan; “3%; 3 .f'“3ff
a. O

v.3

O
.-

    
        
     
   
   
      
  
    
      
  
  

 

l‘u.‘
"I '19.
r 7 0"14’l\o o “’54
£4...“ -.. .33 :5
V. fi"
’33. ~}.

cD-trég.

0 ‘Q
(465.14“ 4!.

 
   

 

 

 

  
   
  
    

   
     
 
    

      
    
   
 
 
    

  

  

  
    
 

9%
-”.q. icy: 3H ’If:k}‘fi ~£z 1 ."-‘r{‘ J" . . '.:fi
-. 3,4" 831%... ,. -'- Lama. "u' ‘2‘; ‘L‘
3 .6 Ar -I .v -- sfi‘.

. g. N: ' '1‘ ‘ «21“ If“ "9"-
I-Q'I ._ Iv!ll'.”_‘“l‘r°.u II}.- . ‘ ‘1‘ 9: .“..'."'Y"
-' 's'y' ‘L4..‘ ' ‘ ’ ‘l 0'... ' £10, ‘ ' y. '.. .‘E’JI‘; 3.“
.' ’. .9 ‘ '1.‘¢: n'o‘ [3“5"Q.I:
.t'JV-t‘; “'f"! ‘gri‘
‘0”; 5 .‘J '23:
g '2»! .- ".2 nan?
' (if ~v'c g. “5’
. 9 ‘. r4 . [‘19.
m1 . . ffiro'W/J'VQI
""9 ""-’{' ‘1."l‘n're1
q r‘ ’I_-" d" “ o‘ ' ’l c) ‘1‘CI".
2' . 'f’ "~f_rb 0’. unov o ‘ 4’- E , (I.

;.'O‘f\? l'.‘.:':.I *‘II-C“.' 3' -§:o.. 1,.
I («NJ 2'. ‘Ilfiw‘l ‘ o .0 3-0 , '- " "‘:’
var! rfh- t"'~-- - e " -.\:-r--.-

. O. " V l'.' tor: .

_‘ , .. ,. .. .
:9!" '4VI szI I. I §!""‘ mf¢-€:’_£'

if. "9 "W”. “yr ‘g‘rJifia'. '9' '_0 o .1. .
. u. n Isl-rt}! I 4.. - I "f‘. ‘ " 3’:
. .7 ”I! "pm .
; ' r;.:.:’ ”mg ”9-1. -0, . J' 'O? : v a
. ,'~. . _ ,I: up”. Qty-- 1.. g > ’ '
'.- ,,_ u ., .I In I
- uou . u I0 .0 0033‘“?!
o:.::.. .4 h:‘-'EI'! -:!Il-.r' .‘3.l’ "
I

37'; I? 2...;- ,1; ”3’3; - ‘
'MifaéwMH-wr m ' °

   
 
 
   

   

 
  
 

  
  
  
  

       
   
   

 

 

. vr: . - -u” A". ‘—
3 . :. .rnrr. .rL-u ~ whoa-*3- 4 :- «Mu—o.
01M. R‘It' anfl‘o‘ ' o o - .‘pO-II- ”0-90,: (0'3” n, )4. . '6 U-..
.o ".539 I.mfi{-.~3!~- ' Ian' v '6'" !“rr.rr‘- tiff-“.331 £0
. “a... o. l‘”' ‘1’ ’W .90 my I" 4‘ ' ‘l '3 “4'0.“ Z'uP‘I'. K Y‘F‘.“"'~"-"J'
'. . . u", I . ‘ ‘ .. ’ . u .0. '. . . fif- yI-g‘PE‘ P‘" '3'“! ['0 I‘I‘O " a. - o — wv—w-nvfwu- D‘.‘.\ 9 -I".'(J
.. .. J . . .--- ,t .. M .. , , .., . $155". {.3 +31: ...';,t.'".";.:!.'a 7733:}.
a O . 0" ".tl‘.’ 'fifé'mrfzté 19“ .y-f‘nr}u;vr:~1 ' ' " .. 3‘
, ' - :- 4r" «tawny: ma IL ~;p

‘v 0 I . ' . '

., .. ' : "w "-“ I “’
I no I "Snub. '

'nl' . U"

   

 

  
      
 
 
   
 
   
      
 

 
 

"O I ”r. «'99 t.‘ 'l ’ ‘ ‘ ”3 '.. b“ "’,‘E‘,.,“,f,., 33", "l
'I) “3- -rL’~l' r‘d‘o no. "In.
‘ L“,

.,.:

    

  

    

,.
H . .
_,'oo~-'.,.,J"l. , ;.,. 3, I"! . g,“
.' f ’ “ - --2 , ILg-w'i-IHIIIL II‘J":FJP’"""'L";!"
' I
l .

COO-J" IOOOO.\

 

  
 

 

”rig-P "‘39:?! If
~va
"I (‘1

’W
O.

 
 
     
  
  
  
 
  
 
   
  
 
 
  
  

 
 

   
 
   

g --
hm, ..:. 'i“ .. 3’,

:5 gnaw“), i3 "Io-(Lpu; .,... '
"f( JHZ'E vfi?rfi'¥‘.’\f.l‘”" ‘W O"“"I'HJ:‘:'

IroIIch-l-fio“ W’I' root, w-v~p( 'fn.....rw I PI-oq. .:.‘
awn-W I'm a r.::;a—’4,.II'IM. “. . , '31“ r, I’w‘." _ _ . .
t » x0 C’O','O"”9 :vI'o-an¥( 0'90.” "W " if? ’ ' J n- r I 7"

’f 'P‘ .0 a - ,Iu-u fi.‘”¢'.

    

’2‘... pm'l ‘U-J-fi‘l ”-9". no

.m'osz:l. ’,.._,c I"-"'-- 3‘» ‘ x;' .-

“3.37.3.8.» . “C... 9”.” g. ,‘
QWLfi-q’ 'l' 1"..0'P’ 00,- Mi,

- nw-v'nlo... ~..: +,'.';

  

. I ”My,

   
  
 
     
  
 
  

 
 
      
  
  
 

 

_ y"o‘«~
o I“. .1 in: I" 3 a» -,
I , or». erfitp'vr‘r.“ Pom, -. 9 «I»!
r- N ’q‘r; .NIg-L. 0W “V VI- w
ido-ad‘o".’9|ol Q,” :'§:3’ . w

”'60:: u- u- m .

..p,-I.IJD - 2.4,III,;...I, ”_,'l “at, H.” h’u.J,-',‘ D,

I p ' v 7303'. Odl‘pzy-a' .- ;.:'~- [9' ”wrurrny-pv-o
wa~ou~ Ava?“ ago 00' o-

,.
¢.v.~430'vo- FF! ”0’ «r ”1' “‘3'; .”‘f’ ~ ~ “
9. 4...-«gr~--v . :.';':::!:5. “4:35. ?3'. -.:':"'~ 3" W“

   
     
       

    
  
  
  

   
          
  
  
   

    
   
   
 

99:94.. 0.}.

  

  

'
.. ' «Ibo-rao .
'- 'r-Odq'lfifio In.” oovno'yv vo- 0’9‘5'0‘
‘I if.» »o£:a-v'vrvwv{;~”f’ .9 ‘33:?! ""l'!;{ ’*' '0' '3»!
0 ¢ -‘ u. v..;0:.~'.r} 0-.
8 . "JLfiZJE'J‘. '
‘I_.. on

ugo§-
0":"I0~

1;. :’:.: I f 9 o , ‘I,;';o"9
9"

 

 

II. 0.... lo -
' II' I’
-

   

     
  

on. 0.-

         

 

 

V '1’..H.'-.“..-‘
(-r' ’C-_ .2“ 9"“. n
o.ov-l. 'I'..""‘ I..~‘)l'00".90" "‘I‘o'o
"OI .. .- tho

2 ‘

    

   

Dim" o 9“

 

  
   

.-
I... ,I

I
010'V‘MMIICO

   
 
  
        
  
    
    
  

 

    

 
   
 
  
 

  

m
I J- I '"o-‘w6p¢'
. .. .~ ~.... .. r'vrp'tan . d""o"-Op~"l"hvm '59..
0" (an 9|.-.4 I. . ' .- I I Up... omrtrr‘.: 3:55:57” py." qqnovl‘v.'0' ‘I 'I.¥Ifr‘:‘ ”on ~o ' .
.o.. .u. (I... .7» out a.“ Iotrnrnocprl-‘D'n'r' ‘105' "-M'V'NII-n-Imv I
-.O"D C" co ..v ,v, ppm-don’t .I l’ "C"."‘.& a“ '4- my :pvuJLatg‘v I :0"er Ivrwpfl I Ono’o' ....
,nmv 4»; 0.9-. m»- Olr' I." o n- 4.... ago’ surgin- “"20..." on w - 5-40 '--~”!'rn ,I- r...
. - 'r“". :.‘|h “"‘. torn“ Dam- i 0.“ [MICL':IO:O‘ '1‘ W" ‘fi‘f‘r’o
. |.-II ru' ~ .0. I’ O 0 I 6'... ‘N I'J'" '0. ’ fi-mnw-rr- "V'" W’W‘SI "‘""" ‘9-
.906-LIlo-II'I'II-I-Ohlv‘90" """'-’" "‘ ' n" ’«P
m... .,,.- .. . ..... .-.q.n-yuco l 33' r1--0p~.. .u-u
OOJIIO'I olI'I-vh'fi'"" "N‘F

  
 
   

  
        
     
         
     
  
  
   
   
  
  
 
  

  

’II a q. 4 . . ." ’:‘ ."‘ I'W "'~
" ‘ ' 9o 4 - astut-oi coo-IA vow-I w.“ an 9.3... :I.. on
9610'.” I I h 0. . . '. '.'.',2,_;' ' . ..--_..-p-.u r9.” "up.vahcp'wzqu.-n.ro.3153’1. w W
' o.vo-'.,(Iu:Iolno'-w Hu‘c.‘ ,...I." I '
I I ‘0." '11 I - O I I
l...‘,'.lb'I-.""u"., . ..".,..:':.
I: pi . I". 0 ”1"(01 H..'.:
II ‘ '- '0 I I . ‘ O .
muffle... 4'“... ......... a.

O~ .v.l . I
' "wa‘. 0‘: w “.0. .olf‘o

 

  
     
  
 
  

   

'4 ".Ollt _.._'.,‘ I“
‘ | "":’.IO'I"'V‘I.'II’O in. .f-..
‘5‘ .,.. . t . . , . ‘l .. It ..1.'.". . O I l‘» 0.0 a. .g'-' 59?;
a - .-. u ,- . - L" - -. "v . Io IO. .- ..I.,-- -. P' - H"
r " "' " . I I at. ' IQI 'r. . 0 4" “l'l v.1 .. '
'l n I 0‘ Q q . . p‘ I I, 1,. .I ‘ ..... v ..,|'~<.-or- , .
' ' '1 ." .L‘ " ':.V' ' ..9 a. I '.‘ .I.I' -II.' . ‘n L p ’:|.O "‘ 'p Q.
I5. . . ll .9 . II 7". I .. W..‘..,IJ A‘ HM" ‘ '
S: I. I I . y' .01 ".‘ .0 * . ...Jl.'..'.‘ ‘.. ‘,‘.';‘ ’1 l ”'|. ' r"‘:-'l" . .. ‘4:.,: 290-". '~ .’
I. '- . . II. I .I . I. o - rv L" I "EI LI“. If? no, I. ....‘, . :,r..‘.. that, I r ‘n”, ..
‘3 I . .. . . .I-;I.IIII...A... . 01- HDIII "\"'.'-| . I “m. l’ “I .4. “VJ?
. , , .. . . . . . ., I . .q. - _ g d. . . .+ _ . . I "::
‘0 I ' ‘ R ‘ ‘ .‘.'.' O. I, ll I .0 I to, I 1" 0.... . 'o 0"... I.¢ - .' I '4‘. “.95.. ”.21... ' ...r_ 00' . 'Or. .':.:.:-. :E:. ‘ ’;1: g", . 'q 4;..fi'! '1." fl .,
.,.. . p- - I .. y .‘ v '1. 5 “.."‘9".'0-...... \y ‘ O u I d .H._.' ' 'r".J""."‘:‘ ~..:..;: .3" ’13..“fi-mx..-“
go I u . ‘ ‘O ' o y 10 I...‘ v . ”a". . . “a I", .Q‘Z-v’--u . '2." .1. ”-9. .u . ..:.. f...--.l§l,.-I "3. I 0'.":..:5 .;.-.I_ -‘AiL :u a...” thm « .‘huv
" ‘ O t o- - ' . . - . .-- . Jo 0......“W d-u‘u‘snn
‘ I . . .. . - ‘ v \.' .. . ..I- -o
-e

 

 

L1 |BRAR| |ES
/ 9‘ moi-«Gm STATE UN‘ ‘5”

\\\\\\\\\ \\\ ji\\\\\\\2\\\\\

\\\\\\\\

“”433 \\\\\\\\\\\\\1\\\\\\\\\\\\\oo
TH 5

This is to certify that the

thesis entitled

Pretesting Effects in the Evaluation of a
Sexual Abuse Education Program for Pre-
school Children

presented by

Grace Christine Gibson

has been accepted towards fulfillment
of the requirements for

M. A. Psychology

 

 

degree in

[inwmwr

Major professor

Date ’0/4944/

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

 

 

 

 

1 LIBRARY
Mlchlgan State
Unlverslty

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

A
DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Lil—+1?!

MSU Is An Affirmative Action/Equal Opportunity Institution
cmmmn

 

=__—__‘____ 7 _.____.___

 

PRETESTING EFFECTS IN THE EVALUATION OF A SEXUAL
ABUSE EDUCATION PROGRAM FOR PRESCHOOL CHILDREN

BY

Grace Christine Gibson

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF ARTS

Department of Psychology

1991

ABSTRACT

PRETESTING EFFECTS IN THE EVALUATION OF A PRESCHOOL CURRICULUM
DESIGNED TO PREVENT THE SEXUAL ABUSE OF CHILDREN

By

Grace Christine Gibson

The sexual abuse of children is a growing problem. One method of
combating this problem is to teach children skills to protect themselves
and prevent the problem from occurring. Many prevention curricula have
been implemented, but few have been adequately evaluated. This study
used a Solomon four-group design to assess the effects of pretesting on
an evaluation of a prevention curriculum.

Subjects were 121 preschoolers, randomly assigned to one of four
conditions: control with pre/posttest, control with posttest only,
experimental with pre/posttest and experimental with posttest only.
Participants were assessed using two measures: (1) the Personal Safety
Questionnaire, a true/false, paper and pencil knowledge test of sexual
abuse prevention concepts, and (2) the 'What If' Situation Test, a
measure which uses short vignettes with accompanying pictures to assess
children's recognition of appropriate and inappropriate touch and the
ability to say no to a perpetrator, to leave a potentially abusive
situation, and to disclose the abuse to an adult.

Analyses of eight component variables showed mixed results for
pretest sensitization. When pretested, preschoolers have difficulty
learning the difference between appropriate and inappropriate touch.
Analysis of individual knowledge components allowed the subtle

pretesting effects to be uncovered and better understood.

To my parents.

iii

ACKNOWLEDGEMENTS

Special thanks to the members of my committee, Anne Bogat,
Marianne McGrath and Ray Frankmann, for their support and guidance in
the completion of this project. Thanks to the SAC research group
(Jackie, Belle, Karen and Tim), to our undergraduate research assistants
(especially Dawn Keish) and to the children, teachers and preschool
directors who participated in this study. Finally, thanks to my 'buddy'
Paul for being willing to take many emergency coffee breaks on very

short notice.

iv

TABLE OF CONTENTS

Page

LIST OF TABLES ...................................................... v
LIST OF FIGURES ..................................................... vi
INTRODUCTION ........................................................ l
Rationale ..................................................... 19
Hypotheses .................................................... 20
METHOD .............................................................. 21
Subjects ...................................................... 21
Materials ..................................................... 21
Design and Procedure .......................................... 25
RESULTS ............................................................. 27
DISCUSSION .......................................................... 43
APPENDICES ......................................................... 50
A: Personal Safety Questionnaire ........................... 50

B: PSQ Subscale ............................................ 52

C: "What If" Situation Test ................................ 53

D: Consent Form ............................................ 66
REFERENCES .......................................................... 68

LIST OF TABLES

Table Page
1. Selected Information About Sexual Abuse Prevention

Program Evaluations ............................................... 10
2. Multivariate Analyses of Variance for Gender

on the PSQ and WIST ............................................... 28
3. Multivariate Analyses of Variance for Age

on the PSQ and WIST ............................................... 30
4. Multivariate Analyses of Variance for Center

on the PSQ and WIST ............................................... 31
5. WIST Pretest Center Means ......................................... 32
6. Analysis of Variance of PSQ Variables for Posttest Scores ......... 34
7. Means for the Posttest Scores on Each of the PSQ Variables ........ 35
8. Analysis of Variance of WIST Variables for Posttest Scores ........ 36
9. Means for the Posttest Scores on Each of the WIST Variables ....... 37
10. Means and Standard Deviations for the Pre and Posttested

Groups on the PSQ and WIST ....................................... 39

11. Repeated Measures Analysis of Variance for PSQ Variables ......... 40
12. Repeated Measure Analysis of Variance for WIST Variables ......... 41

vi

LIST OF FIGURES
Figure Page

1. The Solomon Four-Group Design .................................... 14

vii

INTRODUCTION

The sexual abuse of children has substantially and significantly
increased in the United States since 1980 (Sedlak, 1989). The Study of
National Incidence and Prevalence of Child Abuse and Neglectzl988 (Study
FindingszNIS) cited child sexual abuse as the least common type of child
abuse, (physical and emotional abuse were more common); nevertheless,
24% of all child abuse is sexual abuse. Annually, 155,900 children are
sexually abused. This figure represents triple that of the first study
conducted by NIS in 1980.

It is estimated that as many as l of every 4 women (Russell, 1984)
and l of every 6 men (Finkelhor, Hotaling, Lewis & Smith, 1990) will
experience some form of sexual abuse during their childhood, and an
estimated one third of these children are 6 years old or younger
(Finkelhor & Hotaling, 1984). Overall, more girls are sexually abused
than boys but 32% of sexually abused boys are under the age of six, only
18% of abused girls are that young (Green, 1988).

These figures, alarming as they are, may underestimate the true
incidence (the number of new cases within a specified time period) of
child sexual abuse. There are many reasons why these statistics may be
inaccurate: unknown numbers of children do not report these events
(Wyatt & Powell, 1988), some of the children who do disclose are met

with skepticism and the abuse is not reported (Wyatt & Powell, 1988), or

2
the abuse may be reported but not to the appropriate agency (Zigler &
Hall, 1989). Abuse may also not be reported because of the belief that
parents have the right to treat their children as they see fit (Zigler &
Hall, 1989). Further, there may be an age bias in reporting. Older
children may have higher rates of child sexual abuse because most
studies record the age of the child at the time of the reporting, which
may actually be many years after the onset of the abuse (Finkelhor &
Baron, 1986). Older children are more likely to disclose abuse than
younger children (Finkelhor & Baron, 1986), and this could also increase
older childrens' incidence rates.

Reported prevalence rates of the number of children who are abused
in childhood may also be higher than studies indicate, for a variety of
reasons. Adults may not remember being abused (Peters, Wyatt, &
Finkelhor, 1986). Partially forgotten experiences may not be retrieved
due to inadequate questioning or questions that define abuse differently
from the memories of the abuse (Peters et al., 1986). Adults who were
abused as children often feel shame or embarrassment about the event(s)
(Jacobson & Herald, 1990), so they consciously withhold disclosing the
abuse (Peters et al., 1986).

Whether sexual abuse is disclosed or not, it has serious
consequences. Initially a child may experience school problems,
excessive fears, and severe depression (Saslawsky & Wurtele, 1986).
Browne and Finkelhor (1986) cite reactions of anger, hostility and
anxiety, as well as inappropriate sexual behavior. Immediately after
the abuse, one-fifth to two-fifths of children manifest clinically

pathological disturbances (Browne & Finkelhor, 1986). Adults who were

3
abused as children report marital, familial, and interpersonal
relationship problems (Saslawsky & Wurtele, 1986). They also have
difficulties parenting their own children (Cole & Woolger, 1989).
Browne and Finkelhor (1986) also include depression, self-destructive
behavior, anxiety, feelings of isolation and stigma, poor self-esteem, a
tendency toward revictimization, and substance abuse as other possible
long-term effects. When compared to their nonvictimized counterparts,
less than one-fifth of adult survivors manifest clinically pathological
disturbances. So, extreme long-term effects are not inevitable, but the
potential for initial and/or permanent mental health problems for the
victims of childhood sexual abuse is very real and "should be taken very
seriously" (Browne & Finkelhor, 1986, p. 72).

In response to the increasing incidence and serious aftereffects
of child sexual abuse, many school-based prevention programs have been
implemented. Haugaard and Reppucci (1988) note that "prevention
programs are clearly among the fastest growing components of the
movement to cope with child sexual abuse" (p. 313). Based on adult
victims accounts of childhood abuse, Finkelhor (1984) states that
"children properly educated about the problem of sexual abuse might be
substantially less vulnerable to victimization" (p. 135). Finkelhor
(1984) notes that enough children are at risk "that we need to put more
effort into prevention to forestall abuse before it occurs" (p. 234).

It seems reasonable to teach children prevention skills because
perpetrators are not easily detected (Barnard, Fuller, Robbins, & Shaw,
1989; Bierker, 1989; Kobinsky & Behana, 1984) and once identified,

treating perpetrators and preventing them from reabusing children has

4
not proven successful (Barnard et al., 1989; Becker & Kaplan, 1988;
Crawford, 1981; Finkelhor, 1986; Salter, 1988). Furthermore, primary
prevention programs direct resources to potential victims, not abusers.
The programs are a cost efficient method for reaching large numbers of
children and teaching them prevention skills (Haugaard & Reppucci,
1988). Yet, many school boards and parents are still reluctant to allow
the implementation of prevention programs in their schools
(Anderson-Varney, 1988; Wurtele & Miller-Perrin, 1987), especially when
these programs are directed toward preschoolers.

There are many reasons why adults are reluctant to implement such
programs for preschoolers. Parents and educators often mistakenly
believe that preschool children are too young to be at risk for abuse.
Parents often fear that sexual abuse prevention programs are actually
sex education programs, and they do not want their preschoolers learning
about sex (Kobinsky & Behana, 1984). They worry that the information
taught in the programs will frighten the children (Garbarino, 1987;
Plummer, 1984; Wurtele & Miller-Perrin, 1987) or make them needlessly
wary of affection from all adults (Conte, 1986; Finkelhor, 1984;
Gilbert, Berrick, LeProhn & Nyman, 1989; Kobinsky & Behana, 1984;
Reppucci & Haugaard, 1989; Stone, 1986). Finally, by putting the
responsibility to prevent sexual abuse onto children, some persons are
concerned that adults will become less vigilant and feel secure because
their children are protected (e.g. Reppucci & Haugaard, 1989).

Preschoolers and older children are taught personal safety skills
through a variety of methods (e.g. movies, plays, comic books, and

lectures). Often programs begin with a definition or labeling of

5

genitals. Some programs use the anatomically correct terms (breasts,
penis, and vagina), but most programs use euphemisms such as "private
parts, the parts of your body covered by your bathing suit or underwear"
(Gilbert et al., 1989; Wurtele, Saslawsky, Miller, Marrs, & Britcher,
1986). Most programs teach children the touch continuum: the difference
between good, bad, and confusing touch (Conte, Rosen, & Saperstein,
1986; Conte, Rosen, Saperstein, & Shermack, 1985; Gilbert et al., 1989;
Nelson & Clark, 1986). Good touch is touching that is nurturing and
makes a child feel good. Bad touch is touching that is disturbing and
makes a child feel bad. All programs teach that the inappropriate
touching of a child's "private parts" is bad touch (Gilbert et al.,
1989). Confusing touch may have started out feeling good to a child, or
the child may have liked some parts of the touch or the attention from
it, but then something changes (the touch or the experience of the
touch) and the child begins to feel mixed-up and confused about the
touch (Gilbert et al., 1989). Confusing touch can be a transition
between good and bad touch (Nelson & Clark, 1986). The programs teach
children to refuse bad and confusing touch (Conte et al., 1988) and to
get away from people who make such overtures (Finkelhor & Strapko, in
press). The programs also teach children to tell their parents or other
trusted adults when such touching or attempted touching occurs and not
to keep it a secret (Conte et al., 1986; Finkelhor & Strapko, in press;
Gilbert et al., 1989).

Although researchers have been able to measure change in a child's
understanding of various sex abuse prevention concepts, there are

problems and limitations with the current instruments. First, most

6

evaluations use a composite or mean score to demonstrate that children
have learned the prevention concepts discussed above (Anderson-Varney,
1988; Binder & McNiel, 1987; Blumberg, Chadwick, Fogarty, Speth, &
Chadwick, 1991; Borkin & Frank, 1986, Conte et al., 1985; Downer, 1984;
Garbarino, 1987; Gilbert et al., 1989; Harvey, Forehand, Brown, &
Holmes, 1988; Kenning, Gallmeier, Jackson, & Plemons, 1987; Kolko,
Moser, Litz, & Hughes, 1987; Kraizer, Fryer, & Miller, 1988; Sigurdson,
Strang, & Doig, 1987; Wall, 1983; Wolfe, MacPherson, Blount, & Wolfe,
1986; Wurtele et al., 1988; Wurtele, Marrs & Miller-Perrin, 1987;
Wurtele & Miller-Perrin, 1987). Sometimes participants' responses to a
specific question are analyzed and reported as an example of a programs'
effectiveness in teaching this particular skill or concept (Hill &
Jason, 1987; Nelson, 1981/85; Plummer, 1984; Ratto, 1988; Ray & Dietzel,
1984; Saslawsky & Wurtele, 1986; Swan, Press, & Briggs, 1985). Both
approaches to data analysis are problematic. The use of composite
scores obscures which specific concepts and skills are or are not being
mastered. Using one question dependent measures is not a sufficient or
reliable indication of a child's mastery. Only recently have
evaluations begun to use subscales of particular skills to evaluate
children's knowledge gain (i.e. Liang & McGrath, 1991; Wurtele, 1990b;
Wurtele et al., 1991).

Second, very few evaluations use dependent measures that have been
used before (some exceptions are Anderson-Varney, 1988; Garbarino, 1987;
Ratto & Bogat, 1990; Wurtele et al., 1987; Wurtele et al., 1986); this
prevents researchers from comparing results across studies. Most of the

dependent measures that are used have inadequate or undocumented

7
psychometric properties (Garbarino, 1987; Hill & Jason, 1987; Swan,
Press & Briggs, 1985; Wurtele & Miller-Perrin, 1987). Third, the
questionnaires may be highly reactive.

By definition, when subjects know their behavior is being as-
sessed, the dependent measure is reactive (Kazdin, 1980). "Whenever the
measurement process is not a part of the normal environment it is
probably reactive" (Campbell, 1957, p.299). The very act of measurement
may change subjects' behavior; they are cognizant of being observed and
may modify their responses because of this (Campbell, 1957). Some
measurement techniques are more reactive than others (Kazdin, 1980).
For example, administering the same instrument for pre and posttest
purposes (a common procedure in child sex abuse prevention evaluations)
is a problem (Campbell & Stanley, 1963). Because subjects have already
been exposed to the dependent measure at pretest, they might decide to
alter their responses at posttest, because they know what information
they are being "tested" on (cf. Kazdin, 1980). Some instruments are
more reactive than others (Kazdin, 1980). Instruments that measure
learning and memory change are highly reactive (Campbell, 1957).
Campbell & Stanley (1963) note that the more novel and motivating the
measure, the more reactive one can expect it to be.

Both forced-choice and Open-ended questionnaires are used to
evaluate childrens' knowledge of sexual abuse prevention concepts. The
forced-choice questionnaires are usually very short (7 to 24 questions)
and are administered in a group format, although with younger children
the directions and questions are often read aloud or the questionnaires

are administered individually (e.g. Anderson-Varney, 1988; Binder &

8
McNiel, 1987; Conte et al., 1985; Hill & Jason, 1987; Ratto & Bogat,
1990). The Personal Safety Questionnaire (Saslawsky & Wurtele, 1986) is
one example of a forced choice questionnaire. Some of its questions in-
clude: "Do you have to let grown-ups touch you on your private parts?,"
"If someone touched a kid's private parts and promised never to do it
again, should the kid tell someone?," and "Is it right for grown-ups to
make you touch or look at their private parts?"

Unlike the forced-choice questionnaires, the more open-ended
measures use role plays, puppets, videos or vignettes (imaginary
situations) to assess what children have learned as a result of
participation (e.g. Anderson-Varney, 1988; Borkin & Frank, 1986;
Gilbert et al., 1989; Hill & Jason, 1987; Ratto, 1988; Saslawsky &
Wurtele, 1986; Swan et al., 1985; Wurtele et al., 1987; Wurtele &
Miller-Perrin, 1987; Wurtele et al., 1986). These measures are also
short, usually 4 or 5 vignettes, but they are always administered
individually. Open-ended questions supply a frame of reference for the
participants' answers but, unlike the fixed-alternative questions, they
put a minimum of restraint on the type and expression of answers. For
example, a vignette from the "What-If" Situation Test (Wurtele, Kast, &
Kondrick, 1988) asks children how they would respond if they were
playing outside and a man that they know offered to buy them an ice
cream cone if they let him touch their private parts. Following the
reading of this vignette, five open-ended questions are asked to
determine if the child thought the request was appropriate, what the

child would say and do in the situation, if the child would tell anyone,

9
who the child would tell, and what the child would say to that
person(s).

The open-ended questionnaires used in child sex abuse program
evaluations are both novel and potentially motivating. The use of
puppets, videos, etc. may alter childrens' responses; children may be
responding to the unusual testing experience in a way that is very
different from their responses on common paper and pencil tests or real
life situations. For preschoolers, any type of testing may be reactive
because it is not a normal part of their daily experience. For older
children who are more familiar with testing, the manner in which the
dependent measure is administered (group or individual) could also
contribute to the novelty of the situation and make the measure more
reactive. Because all of the open-ended (reactive) measures are in-
dividually administered, this unusual testing situation may also serve
to change the way children respond to these dependent measures.

In addition to the measurement issues just discussed, sexual abuse
prevention program evaluations are fraught with methodological problems.
,In a review of 31 evaluation studies (see Table l), 18 did not randomly
assign the participants, 16 did not use a control group, 6 had no
pretest, and only 4 attempted to control for pretest sensitization
effects. Different experimental designs have been used in these
evaluations. The most common design is the pretest/posttest design.
Twenty-five of the 31 studies reviewed used a variant of this design,
but 13 of the 25 did not use a control group. Of the 12 studies that
did use a control group, only six of these randomly assigned children to

experimental or control conditions. The remaining six of the 28 studies

10

 

Table 1

Selected Information About Sexual Abuse Prevention Program Evaluations

Study N Age/ Random Control Design
Grade Assignment Group

into Groups

 

Anderson-Varney,
1988 25 4 yrs by center yes pre-post

Binder & McNiel,
1987 83 5-12 yrs no no pre-post

Blumberg et al.,
1991 264 K-3 gr. by room yes pre-post

Borkin & Frank,
1986 83 3-5 yrs no no post

Christian et al.,
1988 131 preschool no no pre-post

Conte et al.,

1985 40 4-10 yrs by child yes pre-post
Downer,

1984 85 9-10 yrs. no yes pre-post
Garbarino,

1987 73 2,4,6 gr. no no post

Gilbert et al.,
1989 118 3-5 yrs ? no Grp 1-pre-post
Grp 2=post only

Harvey et al.,
1988 71 K by child yes pre-post

Hill & Jason,
1987 43 3-5 yrs by child no pre-post

11

Table 1 (cont'd)

 

Study N Age/ Random Control Design
Grade Assignment Group
into Groups

 

Kenning et al.,

1987 (study 1) 72 l & 2 gr. no no pre-post
(study 2) 44 1 & 2 gr. no no pre-post
Kolko et al.,
1987 333 3 & 4 gr. by school yes pre-post
Kraizer et al.,
1988 44 K-2 gr. by child yes pre-post
Liang & McGrath,
1991 117 3-6 yrs by child yes pre-post
Miltenberger &
Thiesse-Duffy,
1988 24 4-7 yrs by child no pre-post
Nelson,
1981/85 931 5 & 6 gr. by school yes post
Plummer,
1984 69 5 gr. no no Grp 1-pre-post
Grp 2=post only
Ratto,
1988 39 3-5 yrs by child yes pre-post
Ray & Dietzel,
1984 139 3 gr. by room no Grp l-pre,
follow-up 1,
follow-up 2
Grp 2-post,
follow-up 1,

follow-up 2

Table 1 (cont'd)

12

 

 

Study N Age/ Random Control Design
Grade Assignment Group
into Groups

Saslawsky &
Wurtele,

1986 67 K,1,5,6 by child* yes modified

Solomon 4

Sigurdson et al.,

1987 137 4-6 gr. no no pre-post
Swan et al.,

1985 63 8-11 yrs by child no Grp 1-pre-post

Grp 2-post only

Wall,

1983 97 4 & 5 gr. no no post
Wolfe et al.,

1986 290 4 & 5 gr. by room yes post
Wurtele,

1990 24 4 yrs by child yes pre-post
Wurtele et al.,

1986 71 K,l,5,6 by child yes post
Wurtele et al.,

1987 26 K by child no pre-post
Wurtele et al.,

1991 52 3-5 yrs by child yes pre-post
Wurtele & Miller-

Perrin,

1987 26 K no no pre-post

* Randomized by school, then by child.

13
employed a posttest only design. Three of these six studies used a
control group, but only one of these three randomly assigned children to
the experimental or control condition.

In addition to the problems mentioned above (non-random
assignment, no control group, and no pretest) another methodological
problem that has been virtually ignored by researchers in this field is
the effect of the pretest on the child's experience of the program
and/or on the child's posttest scores. Without any knowledge of the
effects of a pretest it is impossible to generalize the findings of
evaluation studies to populations who will receive the intervention but
will not receive a pretest. This is important because at some point
prevention programs will have to be implemented without an evaluation
and researchers need to know if they will get the same results (i.e.
learning) when children receive a program without first having had a
pretest.

The only way to assess the effects of pretest sensitization is to
use a Solomon four-group design (Solomon, 1949). The Solomon four-group
design consists of four separate steps: (1) subjects are randomly
assigned to one of four groups, (2) two of the groups receive a pretest,
(3) the treatment is administered to two of the groups, only one of
which is given a pretest, and (4) all four groups receive a posttest
(Huck & Sandler, 1973). See Figure l for the diagrammed design.

In their meta-analysis of 32 studies of pretest sensitization
effects, Willson and Putnam (1982) found an average effect size
elevation of +.22 for groups that had received a pretest. Effect sizes

larger than +.22 were found when the duration between pretest and

l4

 

Group 1 R O X 0
Group 2 R 0 0
Group 3 R X 0
Group 4 R O

R - random assignment; 0 - pre or posttest; X - treatment

 

Figure 1.

The Solomon Four-Group Design (Campbell & Stanley, 1963)

posttest was from several days (.44) to two weeks (.81). Smaller effect
sizes were found when the duration between pretest and posttest was less
than one day or greater than one month. Willson and Putnam also found
that studies using a pretest that measured cognitive changes had an
effect size of almost one-half of a standard deviation above the
non-pretested groups' means. Almost all of the sexual abuse prevention
program evaluation studies use dependent measures that assess cognitive
changes in childrens' understanding of what sexual abuse is and how they
would behave in a potentially abusive situation. Child abuse prevention
program evaluation studies could be very susceptible to the pretest
sensitization effects that have been noted due to the time interval
between pre and posttests (greater than a day, less than a month) and
the type of dependent measures used (cognitive change measures).

Four child sexual abuse prevention program evaluation studies have
examined the effects of pretest sensitization, but none used a true

Solomon four-group design (Gilbert et al., 1989; Plummer, 1984;

15

Saslawsky & Wurtele, 1986; Swan, Press & Briggs, 1985). One study that
assessed preschoolers with reactive measures found evidence of a pretest
sensitization effect (Gilbert et al., 1989). The other two studies,
using older children and less reactive measures, did not find a pretest
sensitization effect (Saslawsky & Wurtele, 1986; Swan et al., 1985).
Because the fourth study (Plummer, 1984) made no distinction in the data
analysis and reporting of the results between the posttest scores of the
pretested and unpretested groups, it is impossible to ascertain the
effect of the pretest, if any, on the posttest scores.

Gilbert et al. (1989) used a pretest/posttest condition and a
posttest only condition to evaluate the effectiveness of 7 different
prevention programs taught at 7 different preschools. They did not
state whether children were randomly assigned to the two groups.
Ninety-three children completed both the pretest and the posttest; 25
children completed only the posttest.

Gilbert et a1. (1989) used measures that are very unusual for this
type of research. On their anxiety scale, children demonstrated how
scared a bunny might feel in a variety of nonabusive situations by
moving the arms of a cardboard bunny into different positions. The
"Bunny Book" is an instrument designed to assess a child's understanding
of four concepts central to most prevention programs: the touch con-
tinuum, secrets, support systems, and self-assertion. The Bunny Book
uses colorful pictures and velcro bunny faces that the child attaches to
the figures to demonstrate what the bunnies might be feeling in the
different situations. In the final measure, a set of wooden rabbits is

used to act out a "Stranger Story." This instrument measures childrens'

16
understanding of stranger danger and their ability to be self-assertive.

There was a significant difference between the two groups'
posttest scores on various sections of the Bunny Book. Gilbert et a1.
(1989) postulate that ”the experimental group experienced a learning
effect from the pretest, thus enabling them to achieve higher scores
the second time they were tested" (p. 50) compared to the posttest only
group. They explain all of the differences between the two groups as a
learning effect from having taken the pretest because both of the groups
received the intervention.

Because these measures are more novel and motivating than the
typical paper and pencil measures administered by researchers conducting
these evaluations, they may also be more reactive and subject to greater
pretest sensitization. This may explain why Gilbert et a1. (1989) found
evidence for pretest sensitization and the other two studies did not.

Saslawsky and Wurtele (1986) used a modified Solomon four-group
design to evaluate a commercially produced film that teaches children
personal safety skills. They randomly assigned 67 children, balanced
for gender and grade, from two public schools, to either the film or the
control condition. Twenty-six of the children were from K and lst grade
and 41 were from 5th and 6th grade. Children at school A received a
pretest. When the posttest means from school A were compared to the
posttest means from school B no significant differences were found, so
the results were combined.

The evaluation consisted of two paper and pencil measures: the
Personal Safety Questionnaire (PSQ) and the 'What If' Situation Test

(WIST). As the reader may recall from an earlier section of this paper,

17
the PSQ consists of 13 closed-ended questions (possible responses are
yes, no, I don't know) about sexual abuse. It is administered in a
group format to older children. The WIST is administered individually.
Children are read 6 vignettes (2 that are benign and 4 that are poten-
tially abusive) and then asked what they might do if they were in such a
situation. Saslawsky and Wurtele found no support for a pretest
sensitization effect.

Swan et al. (1985) found no evidence for pretest sensitization in
an evaluation of the play "Bubbylonian Encounters" with a group of 63
second through fifth graders attending an urban Catholic school. They
randomly assigned children to one of two experimental conditions: a
pretest-posttest group and a posttest only group. For the pretest and
posttest assessment, children watched 5 videotaped vignettes and were
asked to circle which response best described the type of touch depicted
in the vignette. Then in a paper and pencil measure they were asked
about familial abuse, self-protective behaviors, and their overall
response to the play.

The available research on pretest sensitization in the field of
child sexual abuse prevention is very inconclusive. The Gilbert study
supports the notion of a pretest sensitization effect; however, there
are so many methodological flaws with this study that it is difficult to
generalize their findings. There is no mention of the method used to
place the children into the two conditions (pretest/posttest and
posttest only). There is no nontreatment control group to demonstrate
the internal validity of this research. Although the centers were

equally represented in the pretest/posttest condition, they were not

18
equally represented in the posttest only condition. This might mean
that the two groups being compared at posttest are not equivalent.

Saslawsky and Wurtele (1986) only pretested children from school
A, so, in effect, this was not a true Solomon four-group design. They
may have been comparing two very different groups of children. They
acknowledge that further research is needed to corroborate their pretest
sensitization results (Saslawsky & Wurtele, 1986). Their measures (the
PSQ and the WIST) are not as reactive as the series of instruments used
in the Gilbert et a1. (1989) research. Their results may not be
applicable to preschoolers, because their population was older; older
children are familiar with testing, preschoolers are not. Also, because
their intervention was very short, their results may not generalize to
programs with longer interventions.

There are also many flaws with the Swan et al. (1985) study. In
addition to the problems of a small, heterogeneous sample, and lack of a
control group, at least one dependent measure appeared to suffer from a
ceiling effect: "...posttests of both groups showed virtually the same
percentage of correct responses as were found in the pretest. There is
thus no evidence that learning occurred after seeing the play" (p. 401).

In conclusion, Saslawsky and Wurtele (1986) and Swan et a1. (1985)
may not have found a pretest sensitization effect because they evaluated
a short prevention program, studied older children who were more
familiar with testing, and employed less reactive measures. Gilbert et
a1. (1989) may have found a pretest effect not only because of the
unique type of measures employed but also because of the age of the

subjects and the length of their program.

19
Rationale

Primary prevention programs are one of the most popular methods
used to combat the sexual abuse of children. Unfortunately, programs
have been implemented without evaluations or they have been evaluated
with no regard for the effects of the pretest on the evaluation. Just
as a program without an evaluation is limited because the true
effectiveness of the program is not known, so too is an evaluation of
limited value if the effects of administering a pretest are not
understood. Not only is it necessary to evaluate child sexual abuse
prevention programs; the effects of the evaluation itself must also be
examined. A Solomon four-group design is necessary in order to assess
the effects of pretest sensitization.

To date, no researchers have used a true Solomon four-group design
when evaluating child sexual abuse prevention programs. This is
problematic because data from other research areas has indicated a
positive effect size for groups that have received a pretest. This
effect size is even larger when the duration between pre and posttests
is greater than a few days.

The present investigation employed a Solomon four-group design to
examine the effects of pretest sensitization on two types of dependent
measures that have been used in other evaluation research and discussed
previously (The Personal Safety Questionnaire (PSQ) and The 'What-If'
Situation Test (WIST)). These measures were administered individually
to preschoolers over a period of two days. The posttest was conducted
three to four weeks after the pretest. The two dependent measures were

assessed for possible pretest sensitization effects; the more reactive

20
measure, the WIST, was expected to produce a greater pretest
sensitization effect than the less reactive measure, the PSQ. Further,
component analysis of the WIST (rather than simply a composite score)
allowed for the possibility of discovering differential pretest effects
for particular self-protective skills.

It is important to acknowledge the difficulty researchers have
when undertaking primary prevention community based research.
Fortunately, not every study needs to be designed as a Solomon four, but
the literature should contain more examples of this type of design to
ensure that there are no pretest sensitization effects. This project
was designed to help fill this void and to ensure that when programs are
implemented, the same results will be found even when a pretest is not
administered.

Hypotheses
I. Children in the experimental group will have higher posttest

scores on both the PSQ and WIST than will children in the control group.

II. Children who received a pretest with the WIST and the PSQ
will have significantly higher posttest scores on these measures than

will children who did not receive a pretest.

III. Children in the experimental group who were pretested with
the PSQ and WIST will have significantly higher posttest scores than

will children in the experimental group who did not receive the pretest.

21

IV. The pretest effect will be significantly higher on the more
reactive dependent measure, the WIST, as compared to the less reactive
measure, the PSQ.

METHOD

Subjegts

The participants were 121 preschoolers, aged 37 to 80 months (M -
51.8 months). There were 63 (52%) boys and 58 (48%) girls. Ninety-nine
(82%) of the children were white, 16 (13%) were African-American, 5 (4%)
were Asian, and l (1%) was Hispanic.

The children were from six preschool centers. Center 1 was an
inner city daycare center and comprised 12% of the sample. Center 2
consisted primarily of children whose parents attended or taught at a
major mid-western university. This center included many international
students' children. Twenty-one percent of the children came from this
daycare center. Children from Center 3 attended a church affiliated
nursery school and comprised 15% of the sample. Centers 4 and 5 were
suburban daycare centers; 21.5% of the population came from each of
these centers. Center 6 was a rural preschool, and 9% of the children
came from this center.
Materials

The Personal Safety Questionnaire (PSQ) is a 13 item questionnaire
developed by Saslawsky and Wurtele (1986), designed to assess a child's
knowledge of sexual abuse. Children answer either Yes, No, or I'm Not
Sure to the 13 questions. The measure has been modified for use in this
project. Only seven of the original questions were kept and the wording

was modified slightly to make it easier for preschoolers to understand.

22
Two questions were added about who the boss of a child's body is. Four
questions were added asking about appropriate touch situations (i.e.,
cleaning a child's body). Four additional questions were added
concerning touching an adult's genitals or being touched by an adult on
the child's genitals.

The PSQ is administered individually to each child and takes
approximately 10 minutes to administer. The new version of the PSQ
begins by asking the child to identify the "private parts" on a drawing
of two children, a boy and a girl. The child is then asked to explain
what the phrase "private parts" means. If the child is unable to do
either of the tasks, the interviewer shows the child where the "private
parts" are on the drawing and explains what the phrase means. Only
after the child has demonstrated that he or she understands the concept
of "private parts" does the interview continue.

This measure was divided into two components for analysis in this
project. The two questions asking children to identify private parts
were analyzed together to form the variable "PSQ Identification."
Children receive one point for each correct answer. Scores range from 0
to 2 on this variable. An internal consistency reliability (Cronbach's
Alpha) of .87 was determined based on the pretest scores of 62 children.
The second PSQ variable "Information" was constructed from 6 of the 17
questions. Children receive one point for each correct answer and no
points for incorrect or 'I'm Not Sure' responses. Scores range from 0
to 6 on this variable. A Cronbach's Alpha of .79 was found for these

six questions based on the same 62 children. See Appendix A for a copy

23
of the measure and Appendix B for the six questions used in PSQ
Information variable.

The "What If" Situation Test (WIST) was also designed by Saslawsky
and Wurtele (1986). It has been changed for use in this study. The
WIST originally consisted of six vignettes, two benign and four
inappropriate situations but was shortened to 4 stories. Three are
situations that might lead to sexual abuse; a child is asked to expose
him/herself, a child is offered a bribe to inappropriately touch an
adult, and an adult offers a bribe to touch a child. The fourth story
is an appropriate touch or benign situation.

The WIST is also an individually administered measure. The
interviewer reads each story while the child looks at a colorful picture
that depicts the story. After each vignette is read, the child is asked
a standard list of open-ended questions to determine: 1) whether the
child knows if the situation is inappropriate, 2) what the child would
say or do in the situation, 3) if the child would tell someone about the
incident and, if the child would tell, 4) who would he/she tell and, 5)
what he/she would say. 0n the revised WIST, children are also asked to
generate as many different things that they would say or do in an
abusive situation. The WIST takes approximately 15 minutes to
administer.

The revised WIST has been separated into six conceptual variables a
to assess the above concepts, individually, rather than with one global
score. These variables are based on a recently conducted confirmatory
factor analysis of a revised version of the WIST, similar to the one

used in this project (Liang & McGrath, 1991).

24

The first variable, Benign, assessed a child's ability to
recognize benign or appropriate touch. This score was based on one
vignette and ranged from 0 to 1. The second component, Recognition,
measured a child's ability to recognize inappropriate touch. This
variable is based on three vignettes and scores range from 0 to 3. The
Say No variable assessed a participant's ability to reject verbally a
perpetrator. This variable is a mean score of all responses that a
child would say to a perpetrator. A child received no points for all
compliant answers, 1 point for trying to change the situation indirectly
(i.e. threats or excuses), and 2 points for verbalizing that the child
wants the behavior to stop. Scores on this variable range from 0 to 2.
The Go variable measured whether a child would leave the situation. A
participant received no points for all responses in which the child
would remain in the situation, 1 point for delayed escape or help
seeking responses, and 2 points for leaving the scene. This variable
ranged from 0 to 2 and was a mean score of all a child's responses.
Tell Who is the number of people to whom the child would disclose the
potential abuse. Children could give up to 5 responses for each of the
three vignettes. This variable is the sum of all responses that a child
gave and ranges from 0 (if a child would not tell anyone) to 15. Tell
What assessed how specific a child would be when disclosing the
potential abuse. No points were given to children who would not
disclose abuse. Children were given 1 point for an uninformative
disclosure, not identifying the abuser or the situation. They were
given 2 points for being able to identify correctly the perpetrator or

give accurate information about the situation or if a vague description

25
of both is given. Children receive 3 points for being able to correctly
identify the abuser and what occurred. This score is summed across all
three abusive stories and ranges from 0 to 6. An interrater reliability
of .89 was obtained for the scoring of this measure. See Appendix C for
a copy of the measure.
Design and Procedure

Parents at the six preschools were contacted through a letter
inviting them to allow their child to participate in the evaluation.
Parental consent was obtained before any child participated in this
study. Upon return of the consent forms, children were placed in one of
four conditions. See Appendix D for a copy of the invitation and
consent forms.

Children within each classroom were matched for gender and then
randomly assigned to one of four experimental conditions: (1) Control
Pre/Post, children who did not receive the curriculum but were both
pretested and posttested (3-29) (2) Control Post Only, children who did
not receive the curriculum and were only posttested (n—36) (3)
Experimental Pre/Post, children who received the curriculum and were
both pretested and posttested (3&33) and (4) Experimental Post Only,
children who received the curriculum and were only posttested (ga29).
Complete data was obtained for 121 children. This resulted in the
following cell sizes for each experimental condition: Control Pre/Post
(n—27), Control Post Only (n—35), Experimental Pre/Post (g=30) and
Experimental Post Only (3-29).

Following the pretest, the two treatment conditions received the

Grossmont College Child Sexual Abuse Prevention Program, a five-day

26
curriculum that uses a book, a puppet show, role playing, and discussion
to teach children personal safety skills designed to prevent sexual
abuse. This program also includes a teacher training component and an
educational meeting for parents prior to implementation of the
childrens' program. Children in the two pretest conditions were
administered the PSQ and WIST prior to the intervention. Then, the
children in the experimental groups received instruction in an area of
the school where the control groups could not hear what was being
taught. Following the program, all the children were given a posttest.
After the administration of a two month follow-up (a component of the
overall research project not relevant to the present research study),
the delayed treatment control groups received the program.

All children were individually interviewed by trained graduate or
undergraduate research assistants, blind to the childrens' condition.
Interviews lasted from 20 to 25 minutes at each testing period for two
days because of the time it took to administer two other measures in
addition to the PSQ and WIST. These two measures are part of the larger
research project but will not be discussed here. Children were
interviewed in a quiet area within the preschool. Verbal consent was
obtained from each child prior to starting the interview.

All interviewers had received training in both general
interviewing techniques and specific techniques in how to administer the
measures used in this research project. The interviewers had played
with the children to establish rapport a week prior to the beginning of

the interviewing process.

 

RESULTS

There were four experimental conditions in this study: (1) Control
Pre/Post, children who did not receive the curriculum but were both
pretested and posttested (n—27) (2) Control Post Only, children who did
not receive the curriculum and were only posttested (3-35) (3)
Experimental Pre/Post, children who received the curriculum and were
both pretested and posttested (3-30) and (4) Experimental Post Only,

children who received the curriculum and were only posttested (Q=29).
Pretest Comparisons

The two dependent measures, the Personal Safety Questionnaire
(PSQ) and the 'What If‘ Situation Test (WIST) were separated into
conceptual components and each of these variables was analyzed
separately. The PSQ variables were Identification and Information. The
WIST was divided into six variables: Benign, Recognition, Say No, Go,
Tell Who, and Tell What. This created a total of eight variables to be
analyzed. Pretest t-test analyses between experimental and control
groups on these eight variables indicated successful randomization for
all but one variable. On PSQ Information, the experimental group's mean
was significantly lower than the control group's mean (M - 3.73 and M =
4.66, t - 1.95, p<.05; respectively).

Analyses were conducted to determine if gender, age (a median

split was used), or testing center significantly affected any of the

27

28

Table 2

lfultivariate Analyses of Variance for Gender on the PSQ (n = 62)
and WIST (n - 61)

 

 

 

Source E B
PSQ (df - 2, 59)

.10 .903
WIST (df - 6, 54)

1.07 .391

 

eight variables at pretest. Three Multivariate Analyses of Variance
(MANOVAs) were used to analyze the two PSQ variables and three MANOVAs
were used to analyze the six WIST components. There were no pretest
gender differences for any of these variables; PSQ-[£(l, 59) - .10] and
WIST-[2(l, 54) - 1.07] (See Table 2).

There were multivariate age effects on the PSQ [£(1, 57) = 4.22,
p<.02] but not on the WIST [£(1, 52 - 1.21]. Univariate F-tests
revealed a significant effect for the variable Identification [£(1, 58)=
7.87, p<.01]. Older children were significantly better able to identify
private parts at pretest than were younger children (M = 1.52, M = .89;
respectively) (See Table 3).

On the PSQ, there was no multivariate analysis of variance effect
for testing center [£(10, 110) a 1.28] but there was on the WIST [£(30,
202) - 1.53, 2<.05]. Univariate F-tests revealed a significant
difference between testing centers on three WIST variables; Say No [£(5,
55) - 2.54, 2<.04], Tell Who [E(5, 55) - 3.90, p<.01] and Tell What

[F(5, 55) - 4.38, p<.01] (see Table 4). Examination of the means did

29
any consistent patterns. See Table 5 for WIST testing center pretest
means.
Solomon Eour Group Analyses
Because the children were successfully randomized on all but one

variable and were similar at pretest on most demographic variables, the
analyses to test the proposed hypotheses were undertaken. The four
Hypotheses were analyzed using eight 2 X 2 (treatment X testing)
analyses of variance (ANOVAs). The results of each dependent measure
will be presented separately.

PSQ Analyses, A significant main effect for treatment was found on
both of the PSQ variables: PSQ Identification [3(1, 118) - 13.83,
p<.001], and PSQ Information [F(1, 118) - 7.42, p<.01]. The children in
the two treatment groups (M - 1.92 and 5.04; respectively) scored higher
on these two variables than did children in the control groups (M - 1.50
and 4.31; respectively). A significant main effect for testing, in the
predicted direction, was found only for PSQ Identification [£(1, 118) -
7.42, p<.01]. Children who received the pretest (M - 1.86) had higher
posttest scores than the children who did not receive a pretest (M -
1.55). A significant interaction between treatment and testing was also
observed only for this variable [r(1, 118) - 5.51, 2<.o2]. A Scheffé
test [£(3, 118) - 9.39, p<.001] demonstrated that the Control Post Only
group's mean (M - 1.20) was significantly lower than the Control
Pre/Post group's mean (M - 1.79) and it was lower than the experimental
groups' means (Pre/Post M = 1.93, Post Only M - 1.90). See Tables 6 and

7.

30

Table 3

Multivariate Analyses of Variance for Age on the PSQ (n

and WIST (n - 59)

- 60)

 

 

 

 

 

Source E p
PSQ (df - 2, 57)

4.22 .020
PSQ Univariate F-tests:

(df - 1, 58) Error
MS MS

Identification 5.82 .74 7.87 .007
Information 11.30 3.55 3.18 .080
WIST (df - 6, 52)

1.21 .315

 

F‘ .‘IJTJEIth

31

Table 4
Multivariate Analyses of Variance for Center on the PSQ (n - 62)
and WIST (n - 61)

 

 

 

 

 

Source E p
PSQ (df - 10, 110)

1.28 .250
WIST (df - 30, 202)

1.53 .046
WIST Univariate F-tests:

(df - 5, 55) Error
MS MS

Benign .15 .13 1.17 .334
Recognition 1.68 1.10 1.53 .196
Say No 1.03 .41 2.54 .039
Go .61 .30 2.00 .093
Tell Who 41.78 10.71 3.90 .004
Tell What 24.51 5.60 4.38 .002

 

32

 

 

Table 5

WIST Pretest Center Means

Variable Center 3 Mean

Say No l 7 .84
2 13 1.36
3 9 1.20
4 14 1.64
5 13 0.88
6 5 1.19

Tell Who 1 7 1.14
2 13 2.00
3 9 3.00
4 14 5.93
5 13 2.23
6 5 6.20

Tell What 1 7 0.29
2 13 0.92
3 9 0.56
4 14 3.43
5 13 1.08
6 5 4.60

 

 

 

 

33

WIST Analyses. A significant main effect for treatment at
posttesting was found on five of the WIST variables: Recognition [£(1,
117) - 10.17, p<.01], Say No [£(l, 117) - 11.91, p<.001], Go [£(l, 117)
- 11.31, p<.001], Tell Who [£(1, 117) - 13.13, p<.001], and Tell What
[F(1, 117) - 19.56, p<.001]. On these five variables, the children in
the treatment groups scored significantly higher than the children in
the control groups. A significant main effect for testing was found on
two of the WIST variables: Benign [2(1, 117) - 4.13, p<.04] and
Recognition [£(1, 117) - 3.78, p<.05]. Both of these main effects were
in the opposite direction from that which had been predicted. Groups
receiving the pretest had lower posttest scores (Benign M - .75;
Recognition M - 2.34) than the groups that had not received the pretest
(Benign M - .90;Recognition M - 2.65). No significant interactions were
observed. See Tables 8 and 9.
Group X Time Repeated Measures Analyses

Although not hypothesized, a post hoc analysis was undertaken to
determine if there was a significant change from pretest to posttest for
the experimental group compared to the control group on the dependent
variables. Eight repeated measures ANOVAs were performed on the two
groups who were both pre- and posttested (see Table 10 for pretest and
posttest means).

PSQ Identification did not reveal a significant group by time
interaction [£(1, 56) - 2.09]. Both the control and the experimental
groups were significantly better able to identify 'private parts' at

posttest. For PSQ Information, a significant group by time interaction

34

Table 6
Analysis of Variance of PSQ Variables for Posttest Scores
(n - 122)

 

 

PSQ Variables SS DF MS E p
IDENTIFICATION:

Treatment 5.73 1 5.73 13.83 .001
Test 3.08 l 3.08 7.42 .007
Treatment X Test 2.28 1 2.28 5.50 .021
Residual 48.87 118 .41

INFORMATION:

Treatment 16.99 1 16.99 7.42 .007
Test .00 l .00 .00 .998
Treatment X Test 4.77 1 4.77 2.08 .151
Residual 270.02 118 2.29

 

35

Table 7
Means for the Posttest Scores on Each of the PSQ Variables
(n - 122)

 

 

Variable Group Posttest Mean
Identification
1-control/prepost 1.79
2-control/post only 1.20
3-experimental/prepost 1.93
4-experimental/post only 1.90
Information
l-control/prepost 4.50
2-control/post only 4.11
3-experimenta1/prepost 4.83
4-experimenta1/post only 5.24

 

36

Table 8
Analysis of Variance of WIST Variables for Posttest Scores
(n - 121)

 

 

Variables SS DF MS E p
BENIGN:

Treatment .08 1 .08 .55 .458
Test .59 l .59 4.13 .044
Treatment X Test .06 1 .06 .38 .536
Residual 16.66 117 .14

RECOGNITION:

Treatment 7.83 l 7.83 10.17 .002
Test 2.91 1 2.91 3.78 .054
Treatment X Test .01 1 .01 .01 .927
Residual 90.14 117 .77

SAY NO:

Treatment 3.59 l 3.59 11.91 .001
Test .14 1 .14 .48 .491
Treatment X Test .41 1 .41 1.35 .248
Residual 35.30 117 .30

GO:

Treatment 3.03 1 3.03 11.32 .001
Test .08 1 .08 .32 .576
Treatment X Test .87 l .87 3.26 .074
Residual 31.29 117 .27

TELL WHO:

Treatment 237.70 1 237.70 13.13 .001
Test 9.13 1 9.13 .50 .479
Treatment X Test 3.71 l 3.71 .21 .652
Residual 2117.93 117 18.10

TELL WHAT:

Treatment 144.10 1 144.10 19.57 .001
Test .80 1 .80 .11 .742
Treatment X Test .33 1 .33 .05 .832
Residual 861.57 117 7.36

37

Table 9
Means for the Posttest Scores on Each of the WIST Variables
(n - 121)

 

 

Variable Group Posttest Mean
Benign
l-control/prepost 0.70
2-control/post only 0.89
3-experimenta1/prepost 0.80
4-experimenta1/post only 0.90
Recognition
l-control/prepost 2.07
2-control/post only 2.40
3-experimenta1/prepost 2.60
4-experimental/post only 2.90
Say No
l-control/prepost 1.32
2-control/post only 1.27
3-experimenta1/prepost 1.54
4-experimenta1/post only 1.73
Go
l-control/prepost 0.84
2-control/post only 0.73
3-experimental/prepost 0.98
4-experimental/post only 1.20
Tell Who
l-control/prepost 4.93
2-control/post only 4.03
3-experimenta1/prepost 7.37
4—experimental/post only 7.17
Tell What
l-control/prepost 1.33
2-control/post only 1.60
3-experimental/prepost 3.63
4-experimenta1/post only 3.69

F

38

was observed [E(l, 56) - 7.28, p<.01]. The control group's mean
decreased from pretest to posttest (4.60, 4.50; respectively) while the
experimental group's mean increased (3.67, 4.83; respectively). See
Table 11.

Two of the WIST variables showed significant group by time
interactions. For Tell Who [E(l, 54) - 4.26, p<.04], the experimental
group showed a significantly greater increase from pretest to posttest [—
when compared to the control group (3.20 to 7.37 vs. 3.39 to 4.85; 5
respectively). For Tell What there was also a significant group by time
interaction [£(1, 54) - 11.55, p<.001]. The control group's pretest mean
was 1.73, it decreased at posttest to 1.23. The experimental group's
mean at pretest was 1.80, at posttest it increased to 3.63. See Tables

10 and 12.

 

39

Table 10
Means and Standard Deviations for the Pre and Posttested Groups on the
PSQ (n - 58) and WIST (n - 57)

 

 

 

 

 

 

Pretest Posttest
PSQ Group Mean Mean
(SD) (SD)
IDENTIFICATION:
Control 1.36 1.79
(.91) (.57)
Experimental 1.13 1.93
(.90) (.37)
INFORMATION:
Control 4.61 4.50
(1.66) (1.77)
Experimental 3.67 4.83
(2.09) (1.29)
WIST Group Mean Mean
(SD14, (SD)
BENIGN:
Control .85 .70
(.36) (.47)
Experimental .83 .80
(.38) (.41)
RECOGNITION:
Control 2.00 2.07
(1.07) (1.17)
Experimental 1.97 2.60
(1.16) (.77)
SAY NO:
Control 1.25 1.32
(.61) (.52)
Experimental 1.14 1.54
(.73) (.49)
GO:
Control .84 .84
(.53) (.58)
Experimental .71 .98
(.64) (.47)
TELL WHO:
Control 3.39 4.85
(3.89) (4.82)
Experimental 3.20 7.37
(3-23) (4.12)
TELL WHAT:
Control 1.73 1.23
(2.92) (2.16)
Experimental 1.80 3.63

(2.59) (2.47)

40

 

 

 

Table 11

Repeated Measures Analysis of Variance for PSQ Variables

(n - 58)

Source of

Variation SS DF MS E p
IDENTIFICATION:

Group .04 l .04 .07 .786
Time 10.93 1 10.93 22.81 .000
Group X Time 1.00 l 1.00 2.09 .154
Within Subjects

Error 26.83 56 .48

Between Subjects

Error 31.65 56 .57

INFORMATION:

Group 2.67 1 2.67 .61 .437
Time 8.13 1 8.13 5.03 .029
Group X Time 11.75 1 11.75 7.28 .009
Within Subjects

Error 90.42 56 1.61

Between Subjects

Error 244.09 56 4.36

 

41

Table 12
Repeated Measures Analysis of Variance for WIST Variables
(n - 57; for Tell Who & Tell What n - 56)

 

Source of

 

 

 

Variation SS DF MS F p
BENIGN:

Group .04 l .04 .22 .642
Time .23 1 .23 1.79 .186
Group X Time .09 1 .09 .72 .401
Within Subjects

Error 7.19 55 .13

Between Subjects

Error 10.82 55 .20

RECOGNITION:

Group 1.72 1 1.72 1.31 .258
Time 3.56 1 3.56 3.96 .052
Group X Time 2.22 1 2.22 2.47 .121
Within Subjects

Error 49.41 55 .90

Between Subjects

Error 72.61 55 1.32

SAY NO:

Group .10 1 .10 .21 .646
Time 1.56 1 1.56 6.40 .014
Group X Time .76 l .76 3.12 .083

Within Subjects
Error 13.37 55 .24

Between Subjects
Error 25.77 55 .47

 

42

Table 12 (cont'd)

 

Source of

 

 

 

 

Variation SS DF MS E p
GO:

Group .00 l .00 .00 .985
Time .50 1 .50 2.01 .162
Group X Time .52 l .52 2.08 .155
Within subjects

Error 13.72 55 .25

Between Subjects

Error 20.81 55 .38

TELL WHO:

Group 38.00 1 38.00 1.85 .179
Time 220.61 1 220.61 18.46 .000
Group X Time 50.96 1 50.96 4.26 .044
Within Subjects

Error 645.31 54 11.95

Between Subjects

Error 1107.99 54 20.52

TELL WHAT:

Group 42.55 1 42.55 4.38 .041
Time 12.38 1 12.38 3.77 .057
Group X Time 37.92 1 37.92 11.55 .001

Within Subjects
Error 177.33 54 3.28

Between Subjects
Error 524.16 54 9.71

 

DISCUSSION

This study is the first attempt to evaluate the effects of
pretesting on a sexual abuse prevention curriculum using the correct
methodological design, the Solomon four-group design. The three studies
discussed earlier (Gilbert et al., 1989; Saslawsky & Wurtele, 1986; and
Swan et al., 1985), only used variants of this design. The use of
component scores as knowledge measures and the reactivity of the two
dependent measures will be discussed in light of the pretest
sensitization results.

A Solomon four-group design allows the researcher to assess the
effect(s) of a pretest on the posttest scores. Composite scores are
generally used to analyze childrens' level of knowledge in sexual abuse
prevention programs (e.g. Ratto & Bogat, 1990; Saslawsky & Wurtele,
1986). This method of analysis does not accurately elucidate what
skills children have learned as a result of participation in a program
and what skills they have not yet learned. If composite scores are used
one also risks obscuring any subtle pretesting effect(s). A composite
score might allow one to determine whether pretesting effects exist or
not, but it would be impossible to ascertain what aspects of knowledge
the pretest was specifically effecting. Component scores allow one to
more fully understand the subtle effects of pretesting. In this study,

the pretested groups' scores at posttest were significantly different

43

 

44
from the non-pretested groups' scores on three of the eight variables:
PSQ Identification, WIST Benign, and WIST Recognition.

At posttest, the pretested children were better able to identify
'private parts' on PSQ Identification than the non-pretested children.
Recall that during administration of the PSQ, children who were unable
to identify 'private parts' were taught by the interviewer what the term
meant and where 'private parts' were. Thus, children had to demonstrate
that they understood private parts before the interviewer would
continue. The pretested children had had this experience; thus, their
proficiency at posttesting is probably the result of learning this task
the first time and retaining the knowledge for the posttest. Gilbert et
a1. (1989) also explain the difference at posttest between their two
groups as a learning effect from the pretest.

There was also a treatment by testing interaction on this
variable; the experimental groups both did well (see Table 7) whether
they received a pretest or not, whereas the control group did
significantly better if they had a pretest. This finding does not seem
to be the result of an interaction between the pretest and the program.
In other words, the pretest probably did not sensitize the children so
that they experienced the curriculum differently from the non-pretested
children. If this had been so, one would expect the Pre/Post
Experimental group's score to be significantly higher than the Posttest
Only Experimental group's score, and it is not.

The other two variables, Benign and Recognition, for which
pretest sensitization effects were found, are from the WIST. For both

of these variables, children who received a pretest performed less well

 

 

 

 

45
than did children who only received a posttest (see Table 9). This
finding is opposite from that which was predicted. It was hypothesized
that children who had received a pretest would do better, possibly
'because this was a more reactive measure and also because the children
had.been alerted to the questions and might pay closer attention to the
curriculum. Participation in the program had no effect on childrens'
ability to recognize benign touch, but after the program the
experimental children could more accurately recognize inappropriate
touch than could the control children. However, the pretested children
were less able to accurately identify either of these situations. The
pretest may have inhibited children learning this particular prevention
concept but not the other four WIST concepts. That is, the pretest did
not affect a child's ability to learn the no, go, tell message (Say No,
Go, Tell Who and Tell What variables).

A review of the curriculum shows that the no, go, and tell
components are repeated throughout and the children are encouraged to
role play or generate verbal responses for these components. Although,
the concepts of good and bad touch are emphasized, children get no
chance to practice these concepts.

Reppucci and Haugaard (1989) discuss the difficulty that adults
have in defining situations as sexually abusive. Some acts are clearly
abusive but many others require subtle judgement calls. They question
whether it is possible for preschool-aged children to make such subtle
distinctions, given that adults, with their far superior cognitive
abilities, have difficulty with such decisions. Realizing that a

behavior is sexually abusive might be the first skill that is necessary

"T In; :...:I

 

46
for a prevention program to work (Reppucci & Haugaard, 1989; Wurtele,
1990), because the other skills will not be implemented until a
situation necessitates protection.

Clearly, recognizing that a situation is or will become dangerous,
is a difficult and complex process, and one that the curriculum used in
this study may not have adequately taught to the children who were
pretested before they received the program. The children may also have

been confused by the questions asked on the pretest. In all likelihood,

 

the items in the pretest were questions the children had never heard or
thought about before. The curriculum may not have sufficiently
explained the difference between good and bad touch for these children,
and, as noted above, the curriculum offers no opportunity for children
to practice these recognition skills. Some of these children, including
the control children, remained confused at posttest. Although
significantly greater than the control groups' means, the Pre/Post
Experimental childrens' means were still lower than the Posttest Only
Experimental childrens' means. Perhaps the pretested children paid
closer attention to the curriculum, waiting for this complex and
confusing concept to be clearly explained to them, but it was not. The
Pre/Post Control children would have had time to think about good and
bad touch, but with no additional clarifying information, these children
may have been even more confused at posttest when they were asked the
exact same questions. They may have thought their answers were wrong at
pretest and therefore they were being asked the questions again. This
might explain why their posttest means were also lower than those of the

Control Post Only children.

47

Liang and McGrath (1991) propose an alternative explanation for
their, somewhat similar, results. They pre- and posttested 117 children
with the WIST and found no significant condition by time interactions
for the WIST variables Recognition and Benign. On the WIST, thirty-two
percent of their pretested children were able to say no and leave a
potentially abusive situation without being able to correctly identify a
situation as abusive, 83% of these children were from their two younger
age groups. Liang and McGrath propose that younger childrens' responses
may be more affectively based and less cognitively based in sexually
abusive situations. These children may be responding to "instinctive
emotional cues" that allow them to reject a perpetrator and leave a
situation without a cognitive understanding that the situation is
inappropriate.

It is important to realize that had these results been analyzed
with composite scores, the effect of having a pretest on learning the
recognition components would not have been evident. One should also
remember that, to date, this is the only Solomon four-group study in the
literature. Future evaluations of sexual abuse prevention programs,
using component analyses and a Solomon four-group design, are needed to
replicate these findings. We need to understand whether the increased
difficulty pretested children have learning to recognize good and bad
touch is a true problem in this research area, or if it is a problem
unique to the curriculum used in this project.

It was hypothesized that the pretest effect would be significantly
greater on the more reactive dependent measure, the WIST. Reactivity,

as previously discussed, is the effect that the measuring instrument or

 

 

 

48
the measuring process has on the participant's response(s). By virtue
of its colored pictures, open-ended questions, and vignettes, the WIST
was expected to be more novel and motivating, and therefore more
reactive, than the PSQ, which asks questions requiring only yes or no
answers. This hypothesis was not supported. Not only were the two WIST
components with significant pretesting effects in the direction opposite
from that which had been predicted, but the pretest effect on PSQ
Identification was much more significant than that observed for the WIST
variables.

The reactivity hypothesis was based, in part, on the Gilbert et
al. (1989) findings. The Gilbert project (1989), which found
significant pretest sensitization effects with preschoolers, used
unusual and probably very reactive measures: a moveable bunny to measure
anxiety, a book with velcro bunny faces to measure feelings, and small
wooden rabbits to act out a stranger story. It may be that the WIST is
not reactive enough to affect the childrens' responses to the same
extent as the Gilbert measures.

Future research should continue to look at the reactivity of
dependent measure(s). Research that uses such measures should be
sensitive to the effect that the measuring instruments have because
ultimately such instruments will not be used in the presentation of a
prevention program. If the experience of the pretest has significantly
affected the experience of the program, researchers will never be able
to ensure that they will get the same results when programs are
implemented with no pretest.

There were significant posttest main effects for treatment on

 

 

 

I II . I . ll Ilill

 

49
seven of the eight variables; only the recognition of benign touch was
not significant. Post hoc analyses were undertaken on the children who
were assessed in both pre and posttest conditions to corroborate the
Solomon four-group results. The results from the post hoc analyses
revealed significant group by time interactions for three of the eight
variables: PSQ Information, WIST Tell Who, and Tell What. Four of the
remaining five variables showed trends toward significance (see Tables

11 and 12). It is likely that the decrease in power due to the reduced

 

sample size prevented these four variables from obtaining significance,
because, as stated above, posttest analyses indicated significant
treatment effects for these four variables.

In conclusion, these results demonstrate the importance of
analyzing specific skills in the evaluation of pretesting effects in
sexual abuse prevention programs and point to the necessity of using
large sample sizes. The current study suggests that pretesting can
interfere with learning the difference between benign and abusive touch.
If pretesting affects a childs' ability to learn this concept, our
evaluations of prevention programs may be flawed. Children may learn
this concept better when not pretested. It is hoped that both creators
of prevention programs and evaluators of these programs will take this
information into consideration when designing curricula and when

evaluating programs.

APPENDIX A

Personal Safety Questionnaire

50

MIC.

 

Interviewer

 

m
Ivaluation of a Dereonal Safety Curriculu-

Directione to the Interviewer: thie interview 1.1 being contacted to deter-ine the count
of knowledge the child hee at thie tine. the on.y inetance where the interview ehould be
ueed to inetruct the child ie the explanation of private parte. For all other quotient
read than ae written. Probe by repeating queatione ae written.

 

1mm: (techild) Ei.lynaleie . What iayournane! I'n‘oin‘tobe
eating you eone queetione and writing down whet you tell no. Thie ien't a teat at all.
You can eay whatever you want to.

A. Here in a picture. Can you
point to the girl'e private
parte? Can you point to the
boy'e private parte?

 

I. Can you tell ne whet 'private parte' neane?

 

 

 

m: (I! the child cannot verbalire a definition. tell hie/her that private parte
are thoee parte of the body covered by a bathing euit and then point to picture to
chew where private parte are located. I! the child waen't able to point to private
parte correctly. eat the child to ehow you. on the picture. where the private parte
are.)

 

51

Do you know what the word 'boaa' means? A beer to eoaeona who aaye what to do.

1.
2.

10.

ll.

12.

13.

14.

15.

l6.

17.

Are you the boae or your body?

If a big pereon touchee a kid'a private
parte. hae the kid done eonething wrong?

Do kide have to let ALL big people touch
their private parte?

1! a big pereon touchee a kid'e private
parte and tella the kid to keep it a eecret,
ehould the kid tell eoaeone about it?

If the kid tella aonabody but that pereon
doean‘t believe ’en. ahould the kid
forget about it?

Are atrangera the ONLY people who try
to touch kide’ private parte?

1! a atranger wanta to look at a kid'a
private parte, ahould the kid try to get
away?

Ia it OK for doctora to look at kide'
private parte if they're hurt there?

Ia it OK for kida to touch a bigger
pereon'a private parte?

1! kide need help cleaning their bodiea.
ia it OK for Non or Dad to touch their
private parte?

1! a babyaitter wanta to touch a kid'a
private parte. what ahould the kid eay?

to it OK for kida to touch their own
private parte?

If a big pereon touchee a kid'e private
parte and pronieed never to do it again.
ahould the kid etill tell aoneona about it?

Ia it OK for a big pereon to touch your
private parte if it feele good?

Ia it OK for your Men or Dad to give you a
hug if you want one?

Ia it OK for grownupa to aaka kide touch
or look at 5h.L; private parte?

Ia your teacher the beer of your private
parte?

Yea

Yea

Yea

Yea

Yea

Yea

Yea

Yea

Yea

Yea

Yea

Yea

Yea

Yea

Yea

Yea

Yea

No
No

No

No

No

No

No
No

No

No
No

No
No
No
"9

No

I'n Not

I'm Not

I'm Not

I'n Not

I'm Not

I'n Not

I'm Not

I'm Not

1'! Not

I'm Not

I'n Not

I'm Not

I'm Not

I'm Not

I'm Not

I’m Not

I'n Not

Sure

Sure

Sure

Sure

Sure

Sure

Sure

Sure

Sure

Sure

Sure

Sure

Sure

Sure

Sure

Sure

Sure

 

 

APPENDIX B

PSQ Subscale

 

f‘m

52

APPENDIX B

PSQ Subscale

l.

2.

Do kids have to let ALL big people touch their private parts?
Is it OK for kids to touch a bigger person's private parts?

If a babysitter wants to touch a kid's private parts, what should the
kid say?

Is it OK for a big person to touch your private parts if it feels
good?

Is it OK for grownups to make kids touch or look at their private
parts?

 

Is your teacher the boss of your private parts?

 

APPENDIX C

"What If" Situation Test

 

53

Code#
Interviewer Initials

 

What If Situation Test (WIST)
M.S.U. Version

Interviewer;

0.K. (child's name), I am going to show you some pictures and read you
some stories about the people in the pictures.

STO N
Let's pretend you were riding your bike and you fell off and hurt
your private parts. When you went home, you told your parents what
happened and they wanted to look at your private parts.
1. Would it be 0.K. for your parents to look at your private parts?
(Please CIRCLE one response)
YES...go to question 2a.
N0....go to question 2b.

DK....go to next vignette.

2a. Why would that be 0.K.?

2b. Why wouldn't that be 0.K.?

PRAISE CHILD FOR HIS/HER EFFORT AND/OR COOPERATION, MQI_£QR_31§1M§R
Efififlflnfiflflip For example, "You're thinking so hard, (child's name);
that's great!" Or "It's really great the way you're listening so
carefully to my questions, (child's name)!”

54

TO 0

Let's pretend a neighbor who is a big person, said to you, "Hey
(childfs name), let's play a real fun game! You take off all your
clothes and I'll take pictures of your private parts with my camera!"

1. Would it be 0.K. for your neighbor to take pictures of your
private parts?

(Please CIRCLE one response)
YES...go to question 2a.
NO....go to question 2b.
DK....go to question 3.

2a. Why would that be 0.K.?

2b. Why wouldn't that be 0.K.?

 

55

If your neighbor asked you to do that, what would you say to them
or what would you 19?

[NOTE #1; After every solution encourage the child to generate more
solutions by saying, "0.K. That‘s something you could try. Anything

else?" NOTE BENE: Take care to respond with the same enthusiasm and the
same words to every solution the child offers.]

3.

[NOTE #2; If a child begins to perseverate, after three enumerations,
stop child and say, "0.K. (name of child), that's kind of like what you
(Explain how three previous responses are

thought of doing before.
The idea here is to think of all the different things you could A

alike).
say or do. What's something different you would say or do?]

 

[N915 #3; If a child says "IDK," prompt by asking, "How would you talk
to your neighbor, or how would you act if they wanted to take pictures

of your private parts?"]

[NOTE #4; Number every response.]

SAY Responses DO Responses

 

 

 

 

4a. If child has m m W:

"So far you said that you would say to your neighbor (repeat all
SAY responses). What else would you tell your neighbor?" OR "How

would you talk to your neighbor?"

If child has MQ SAX mm:

"Try to think really hard now. What would you say to your
neighbor?" OR "What would you tell your neighbor?" OR "How would

you talk to your neighbor?"

Prompted SAY Responses

1.

 

2.

 

 

 

5a.

56

4b. If child has m D_Q RESPONSES:

"Now before you told me that if your neighbor tried to take
picUnms of your private parts, you would (repeat all DO
responses). What else would you do?" OR "You know, how would you
§c_t if your neighbor asked to take pictures of your private
parts?"

If child has Mg 129 RESPONSES:
"Try to think really hard. What would you do if your neighbor
tried to take pictures of your private parts?" OR "You know, how
would you act if your neighbor asked to take pictures of your
private parts?"

Prompted DO Responses

1.

 

2.

 

3.

 

4.

 

Now I'm going to ask you a different kind of question. Let's
pretend your neighbor wanted to take pictures of your private
parts, AND THEY ASKED YOU TO KEEP IT A SECRET. Some kids would
keep that a secret, but other kids would tell somebody. Which kid
is more like you?

(Please CIRCLE one response)

TELL SOMEBODY ........ go to question 5b, then to question 6.
KEEP IT A SECRET ..... go to question So, then to next story.
DK .............. go to next story.

5b. Why would you tell somebody?

So. Why would you keep it a secret?

 

 

57
IF CHILD WOULD TELL SOMEBODY, ASK QUESTIONS 6 AND 7.
6. Who would you tell?
[NOTE #1; Prompt each time child stops answering by asking "Anybody
else?"]

[NOTE #2; Number responses.]

ii}-

7. What would you say to (name of first person listed above)?

[NOIEz If child does not respond, rephrase the question by asking, "How
would you let (persons(s) named above) know what happened to you?" If

child still does not respond, and/or seems uncomfortable, offer to let

child whisper what he/she would say in your ear.]

PRAISE CHILD FOR HIS/HER EFFORT AND/OR COOPERATION, MQI_£QR_31§[HER
mm For example, "You're thinking so hard, (child's name);
'that's great!" 0r "It's really great the way you're listening so
carefully to my questions, (child's name)!"

I!

 

58

STOEX IMEEE

Let's pretend you had a babysitter and they said to you, "(name of
child), I' 11 let you stay up late to watch your favorite TV show if YOU
touch MY private parts for a little while. It'll be fun!"

1. Would it be 0.K. for your babysitter to ask you to touch her private
parts?

(Please CIRCLE one response)
YES...go to question 2a.
NO....go to question 2b.
DK....go to question 3.

2a. Why would that be 0.K.?

2b. Why wouldn't that be 0.K.?

———

a

 

59

3. If your babysitter asked you to do that, what would you say to her
or what would you 99?

[lflTIE #1; After every solution encourage the child to generate more
.solutions by saying, "0.K. That's something you could try. Anything
[else?" NOTE BENE: Take care to respond with the same enthusiasm and the
same words to every solution the child offers.]

[NOTE #2; If a child begins to perseverate, after three enumerations,
stop child and say, "0.K. (name of child), that's kind of like what you
thought of doing before. (Explain how three previous responses are
.alike). The idea here is to think of all the different things you could
say or do. What's something different you would say or do?]

[NOTE #3; If a child says "IDK," prompt by asking, "How would you talk
to your babysitter, or how would you act if she wanted ypp to touch
Mgr private parts? "]

[MQIE_£4; Number every response.]

SAY Responses DO Responses

 

 

 

 

 

 

4a. If child has AME £51 RESEONSES:

"So far you said that you would say to your baybsitter (repeat all
SAY responses). What else would you tell your babysitter?" OR
"How would you talk to your babysitter?"

If child has HQ 5A1 REEBQMfiEfi:
"Try to think really hard now. What would you say to your
babysitter?" OR "What would you tell your babysitter?" OR "How
would you talk to your babysitter?"

Prompted SAY Responses

1.

 

2.

 

 

 

 

60

id). If child has AMI QQ RESPONSES:

"Now before you told me that if your babysitter asked you to touch
Mgr private parts, you would (repeat all DO responses). What else

would you do?” OR "You know, how would you act if your babysitter
wanted to touch private parts?"

If child has Mg 29 RESPONSES:

"Try to think really hard. What would you pp if your babysitter
asked ypu to touch her private parts?" OR "You know, how would you
acr if your babysitter asked to touch private parts?"

Prompted DO Responses

1.

 

2.

 

3.

 

4.

 

5a. Now I'm going to ask you a different kind of question. Let's
pretend your babysiter asked you to touch her private parts, AND
THEY ASKED YOU TO KEEP IT A SECRET. Some kids would keep that a

Secret, but other kids would tell somebody. Which kid is more
like you?

(Please CIRCLE one response)

TELL SOMEBODY ........ go to question 5b, then to question 6.
KEEP IT A SECRET ..... go to question 5c, then to next story.
DK .............. go to next story.

5b. Why would you tell somebody?

So. Why would you keep it a secret?

IF CHILD WOULD TELL SOMEBODY, ASK QUESTIONS 6 AND 7.

 

 

61
IF CHILD WOULD TELL SOMEBODY, ASK QUESTIONS 6 AND 7.

6. Who would you tell?

 

[NOTE #1' Prompt each time child stops answering by asking "Anybody
else?"]

[NOTE #2' Number responses.]

 

7. What would you say to (name of first person listed above)?

[NOIE: If child does not respond, rephrase the question by asking, "How
‘would you let (persons(s) named above) know what happened to you?" If

child still does not respond, and/or seems uncomfortable, offer to let

child whisper what he/she would say in your ear.]

PRAISE CHILD FOR HIS/HER EFFORT AND/OR COOPERATION, W
mm For example, "You're thinking so hard, (child's name);
that's great!" Or "It's really great the way you're listening so
carefully to my questions, (child's name)!"

 

 

62
STORY EQLJR
Let's pretend you were playing at the park and a man you like said

to you, "Hey, (name of child), I'll buy you an ice cream cone if you
‘take off your pants and let me touch your private parts."

1. Would it be 0.K. for that man to ask you to take off your pants so
he could touch your private parts?

(Please CIRCLE one response)
YES...go to question 2a.
NO....go to question 2b.
DK....go to question 3.

2a. Why would that be 0.K.?

2b. Why wouldn't that be 0.K.?

 

 

 

63

3. If that man in the park asked you to do that, what would you say
to him or what would you g2?

[NOTE #1; After every solutiop encourage the child to generate more
solutions by saying, "0.K. That's something you could try. Anything
else?” NOTE BENE: Take care to respond with the same enthusiasm and the
same words to every solution the child offers.]

[NOTE #2; If a child begins to perseverate, after three enumerations,
stop child and say, "0.K. (name of child), that's kind of like what you
thought of doing before. (Explain how three previous responses are
alike). The idea here is to think of all the giffierent things you could

say or do. What's something different you would say or do?] F
[NOTE #3; If a child says "IDK," prompt by asking, "How would you talk ;
to the man, or how would you act if he wanted to touch your private

parts?"]
[NOTE #4; Number every response.]

SAY Responses DO Responses

 

 

 

 

 

4a. If child has AM} 551 BEEEONSES:

"So far you said that you would say to the man in the park (repeat
all SAY responses). What else would you tell him?" OR "How would
you talk to the man in the park?"

If child has M9 EA} REEEQNSEE:

"Try to think really hard now. What would you say to that man in
the park?" OR "What would you tell him?" OR "How would you talk to
the man in the park?"

Prompted SAY Responses

1.

 

2.

 

 

 

64

4b. If child has AM! 29 RESPONSES:

5a.

"Now before you told me that if that man in the park asked to
touch your private parts, you would (repeat all DO responses).
What else would you do?" OR "You know, how would you pp; if that
man in the park asked to touch your private parts?"

If child has Mg QQ RESPONSES:

"Try to think really hard. What would you g9 if that man in the
park asked to touch your private parts?" OR "You know, how would
you acr if that man in the park asked to touch your private

parts?"

Prompted DO Responses

1.

 

2.

 

3.

 

4.

 

Now I'm going to ask you a different kind of question. Let's
pretend the man in the park wanted to touch your private parts,
AND THEY ASKED YOU TO KEEP IT A SECRET. Some kids would keep that
a secret, but other kids would tell somebody. Which kid is more
like you?

(Please CIRCLE one response)

TELL SOMEBODY ........ go to question 5b, then to question 6.
KEEP IT A SECRET ..... go to question 5c, then to next story.
DK .............. end of measure.

5b. Why would you tell somebody?

So. Why would you keep it a secret?

 

65
IF CHILD WOULD TELL SOMEBODY, ASK QUESTIONS 6 AND 7.
6. Who would you tell?
[NOTE #1; Prompt each time child stops answering by asking "Anybody
else?"]

[NOTE #2: Number responses.]

7. What would you say to (name of first person listed above)?

[MQIEz If child does not respond, rephrase the question by asking, "How
would you let (persons(s) named above) know what happened to you?" If

child still does not respond, and/or seems uncomfortable, offer to let

child whisper what he/she would say in your ear.]

PRAISE CHILD FOR HIS/HER EFFORT AND/0R COOPERATION, MOI EQR MIEZHEE
RESIQMEE;b For example, "You're thinking so hard, (child's name);

that's great!" Or "It's really great the way you're listening so
carefully to my questions, (child's name)!"

APPENDIX D

Consent Form

_-—.- .n. 4-‘-4-

66

Dear Parents,

Our preschool is concerned about all types of personal safety that
involve children. We routinely teach children safety rules about fire
and crossing the street. This year we are pleased to offer a sexual
abuse prevention program for children. We believe it is important to
teach children to recognize threatening situations, should they arise,
and to seek help. This is a personal safety program that teaches
children a "No, Go, Tell" message; it is not a sex education program.

It does not provide any explicit information about human sexuality.

The program starts with an optional parent meeting. Parents get
to learn more about the problem of sexual abuse and how to talk about it
with their children. There is also a parent-child workbook that can be
used to reinforce the message of the program at home. The program for
the children takes place in the classroom and involves five, 20 minutes
presentations that include a puppet show, picture stories, and
activities.

We are presenting this program in cooperation with Dr. Marianne
McGrath and Dr. Anne Bogat from Michigan State University. During the
last several years, they have offered the program to several preschools
in this area. Parents and teachers were enthusiastic about the program
and felt it provided a valuable experience for the children. Dr.
McGrath and Dr. Bogat conduct an ongoing evaluation of the program in
order to make continual improvements in it. This is done by
interviewing children before the beginning of the program, shortly after
the program is completed, and 2 months later. The interviewing will be
conducted by Dr. McGrath, Dr. Bogat, or one of their trained assistants.
Children will be asked if they want to answer some questions about
personal safety. They'll be told that there are no right or wrong
answers; we'd just like to know what they think. The questions ask
children to identify appropriate and inappropriate touch and to state
what they would say and do in various pretend situations. Half of the
children who have permission to participate will receive the program
first, the other half will receive it in 2 months. The procedure makes
it possible to determine what children already know about the topic, how
much of what they know is correct, and how much they learn from the
program.

u w n w ha 0 wa to i ate
1 a ete th t f n sen i b o ou h d‘s
Leaphpr. If you have questions about the program, please feel free to
discuss these with your child's teacher before completing the form, or
leave a message for Dr. McGrath or Dr. Bogat at 353-8690 and they'll
return your call and answer your questions.

Remember. WWW
And, if you decide to let your child participate, there will be a parent
meeting before the sexual abuse prevention program begins.

67

Parent Permission Form
Sexual Abuse Prevention Program

1. I understand the evaluation process of the Grossmont College Child
Sexual Abuse Prevention Program. I have read the description of
the project and I understand what my child's participation will
involve.

2. I understand my child's participation is optional and that my
child must also give his/her verbal permission.

3. I understand that either myself or my child may discontinue my
child's participation in the evaluation at any time without
penalty and that my child will still have the opportunity to
participate in the Grossmont Child SeXual Abuse Prevention
Program.

 

4. I understand that all results of the study will be kept in strict
confidence and all responses of my son or daughter will remain
anonymous except if my child reports possible sexual abuse. If
such a report occurs, Dr. McGrath or Dr. Bogat will notify both
myself and the preschool teacher immediately. Within these
restrictions, the results of the study will be made available to
me at my request.

5. I understand that participation in the study does not guarantee
additional benefits to my child or to me.

I hereby give my approval to allow my child to be interviewed about
his/her knowledge of personal safety. I understand that this
information will be used as part of a research project being conducted
by Dr. Marianne McGrath and Dr. Anne Bogat at Michigan State University.
NAME OF CHILD

BIRTHDATE

SIGNATURE OF PARENT/GUARDIAN
DATE

--------- > Please complete a separate form for each child in your
family who you want to participate *****

--------- > RETURN THIS FORM TO YOUR CHILD'S TEACHER AS SOON AS
POSSIBLE. THANKS.

REFERENCES

REFERENCES

Anderson-Varney, T. J. (1988). The evaluation of a personal safety
curriculum for preschoolers. Unpublished master's thesis, Michigan
State University, East Lansing.

Barnard, G. W., Fuller, A. K., Robbins, L., & Shaw, T. (1989). The child
mo ester: n nte rated a roach to evaluatio and treatment. New
York: Brunner/Mazel, Inc.

Becker, J. V., & Kaplan, M. S. (1988). Assessment and treatment of the
male sex offender. In D. H. Schetky & A. H. Green (Eds.), Child

sexua abuse“ A handbook for healt care and 1e a1 rofessionals
(pp.136-152). New York: Brunner/Mazel.

 

Bierker, S. B. (1989). About sexual abuse. Springfield, IL.:Charles C.
Thomas.

Binder, R. L., & McNiel, D. E. (1987). Evaluation of a school based
sexual abuse prevention program: Cognitive and emotional effects.

Child Abuse and Neglect, ii, 497-506.

Blumberg, E. J., Chadwick, M. W., Fogarty, L. A., Speth, T. W., &
Chadwick, D. L. (1991). The good touch/bad touch component of
sexual abuse prevention training: Unanticipated positive conse-
quences. o nal o Inte erso a Violence, §(1), 12-28.

Borkin, J., & Frank, L. (1986). Sexual abuse prevention for
preschoolers: A pilot program. Child Welfgrg, LXV, 75-82.

Browne, A., & Finkelhor, D. (1986). Impact of child sexual abuse: A
review of the research. Psychological Bulletin 99 66—77.

’ _’

Campbell, D. (1957). Factors relevant to the validity of experiments in
social settings. Psychological Bulletin, §g(4), 297-312.

Campbell, D. T., & Stanley, J. C. (1963). Experimental and
quasi-experimental designs for research. Chicago: Rand McNally.

Christian, R., Dwyer, S., Schumm, W. R., & Coulson, L. A. (1988).
Prevention of sexual abuse for preschoolers: Evaluation of a pilot

program. Psychological Reports, pg, 387-396.

68

69

Cole P. M., & Woolger, C. (1989, April). The role of emotion in the
parepting difificuities pf incest victims. Symposia presented at

the meeting of the Society for Research in Child Development,
Kansas City, Missouri.

Conte, J. R. (1986). Evaluating prevention education programs. In M.
Nelson & K. Clark (Eds.), e educato 's uide to reventin child
sexual abusg (pp. 126-129). Santa Cruz, CA: Network.

Conte, J. R., Rosen, C., & Saperstein, L. (1986). An analysis of
programs to prevent the sexual victimization of children. Journal

of Erimary Erevention, g, 141-155.

Conte, J. R., Rosen, C., Saperstein, L., & Shermack, R. (1985). An
evaluation of a program to prevent the sexual victimization of

young children. thlg Abuse gnd Neglect, 2, 319-328.

Crawford, D. A. (1981). Treatment approaches with pedophiles. In M. Cook
& K. Howells (Eds.), d exu nteres ch’ldren (PP.
181-217). New York: The Academic Press.

Downer, A. (1984). Evgiuariop of talking about touching. Unpublished

manuscript.

Finkelhor, D. (1984). Child sexual abuse; New theory and research. New
York: The Free Press.

Finkelhor, D. (1986). Abusers: Special topics. In D. Finkelhor (Ed.), A
sourcebook on child sexual abuse (pp. 119-142). Beverly Hills, CA:
Sage.

Finkelhor, D., & Baron, L. (1986). High-risk children. In D. Finkelhor

(Ed.), A §ourcebook on child sexual abuse (pp. 60-88). Beverly
Hills, CA: Sage.

Finkelhor, D., & Hotaling, G. (1984). Sexual abuse in the National
Incidence Study of Child Abuse and Neglect: An appraisal. Child

Abuse and Neglect, E, 23-33.

Finkelhor, D., Hotaling, C., Lewis, I. A., & Smith, C. (1990). Sexual
abuse in a national survey of adult men and women: Prevalence,
characteristics, and risk factors. Child Abuse and Neglect, lg,

19-28.

Finkelhor, D., & Strapko, N. (in press). Sexual abuse prevention
education: A review of evaluation studies. In D. J. Willis, E. W.
Holder, & M. Rosenberg (Eds.), Child abuse prevention. New York:
Wiley.

Garbarino, J. (1987). Children's response to a sexual abuse prevention
program: A study of the Spiderman comic. Child Abuse and Neglect,
ii, 143-148.

70

Gilbert, N., Berrick, J. D., Le Prohn, N., & Nyman, N. (1989).

Erotspripg young childrep from sexual abuse; Does preschool
training work? Lexington, MA: Lexington Books.

Green, A. H. (1988). Overview of the literature on child sexual abuse.
In D. H. Schetky & A. H. Green (Eds.), Child sexual abuse (pp.30-
56). New York: Brunner/Mazel, Inc.

Harvey, P., Forehand, R., Brown, C., & Holmes, T. (1988). The prevention
of sexual abuse: Examination of the effectiveness of a program
with kindergarten-age children. Eehavior Therapy, i2, 429-435.

Haugaard, J. J., & Reppucci, N. D. (1988). The sexuai abuse of children:

A pomprehensive gpide to current knowledge and intervention
strategies. San Francisco: Jossey-Bass Pub.

Hill, J. L., & Jason, L. A. (1987). An evaluation of a school-based
child sexual abuse primary prevention program. Psychotherapy
Bulietip, 2;, 36-38.

Huck, S. W., & Sandler, H. M. (1973). A note on the Solomon 4-group
design: Appropriate statistical analyses. The Journal of
Experimentai Educatiop, 52(2), 54-55.

Jacobson, A., & Herald, C. (1990). The relevance of childhood sexual
abuse to adult psychiatric inpatient care. Mospital snd Community
Psychiatry, Al, 154-158.

KarHinHanpeE.&(R890). Research Design in Clinical Psychology. New

Kenning, M., Gallmeier, T., Jackson, T. L., & Plemons, S. (1987, July).
Evsiuation of child sexusl abuse prevention programs; A summary of
ryp studies. Paper presented at the National Conference on Family
Violence, Durham, NH.

Koblinsky, S., & Behana, N. (1984). Child sexual abuse: The educator's
role in prevention, detection, and intervention. Young Children,
;2(6), 3-15.

Kolko, D. J., Moser, J. T., Litz, J., & Hughes, J. (1987). Promoting
awareness and prevention of child sexual victimization using the
Red Flag/Green Flag program: An evaluation with follow-up. Journal

of Family Violence, g(1), 11-35.

Kraizer, S. K., Fryer, G., & Miller, M. (1988). Programming for
preventing sexual abuse and abduction: What does it mean when it
works? Child Welfare, LXVII(l), 69-78.

 

71

Liang, B., & McGrath, M. (1991, August). How well do children iearn
sexual abuse prevention concepts: Paper presented in M. P. McGrath

(Chair), Mew Eerspectives on Sexual Abuse Education for
Ersspnoolers. Symposium conducted at the 1991 annual convention of
the American Psychological Association, San Francisco, CA.

Miltenberger, R. G., & Thiesse-Duffy, E. (1988). Evaluation of
home-based programs for teaching personal safety skills to

children. Journal pf Applied Eehavior Analysis, 2i(l), 81-87.
Nelson, D. E. (1981/85). An evaluation of the student outcomes and
ra te c the " u' e in Char e" ro ram.

Unpublished manuscript.

Nelson, M. and Clark, K. (1986). Prevention programs at work. In M.
Nelson & K. Clark (Eds.), educator's u de to eventin child
sexuai abuse (pp. 149-176). Santa Cruz, CA: Network Publications.

Peters, 8. D., Wyatt, G. E., & Finkelhor, D. (1986). Prevalence. In D.

Finkelhor (Ed.), A sourcebook on child seruai abuse (pp. 15-59).
Beverly Hills, CA: Sage.

Plummer, C. (1984). Preventing sexual abuse; Whar in-school programs
teach children. Unpublished manuscript.

Ratto, R. (1988). An evaluation of a preschopi curriculum to prevent the
sexual abuss of children. Unpublished master's thesis, Michigan
State University, East Lansing.

Ratto, R., & Bogat, G. A. (1990). An evaluation of a preschool
curriculum to prevent the sexual abuse of children. Journal of

Qommuniry Esychplogy, ifi, 289-297.
Ray, J., & Dietzel, M. (1984). Teaching child sexuai abuse prevention.

Unpublished manuscript.

 

Reppucci, N. D., & Haugaard, J. J. (1989). Prevention of child sexual
abuse: Myth or reality. Ansrican Esychologist, AA, 1266-1275.

Russell, D. E. (1984). Sexual exploitation. Beverly Hills: Sage.

Saslawsky, D. A., & Wurtele, S. K. (1986). Educating children about
sexual abuse: Implications for pediatric intervention and possible

prevention. Journal of Eediatric Psychology, ii, 235-245.

Salter, A. C. (1988). Treating child sex offendersrsnd victims: A
practical guide. Newbury Park: Sage.

 

Sedlak, A. J. (1989, April). National incidence of child sbuse and
neglect. Paper presented at the meeting of the Society for
Research in Child Development, Kansas City, Missouri.

72

Sigurdson, E., Strang, M., Doig, T.'(1987). What do children know about
preventing sexual assault? How can their awareness be increased?

Cansgisn Journal pf Esychiatry, 1;, 551-557.

Solomon, R. L. (1949). An extension of the control group design.

Esypnoiogicai Buiietin, 39, 137-150.

Stone, M. E. (1986). New myths about child sexual abuse. In M. Nelson &
K. Clark (Eds.), The educator's gnide to preventing child sexual

spnss (pp. 130-132). Santa Cruz, CA: Network.
Stud ' ud he Nat o a Incidence and Prevalence of Child
Abuse sng Msgiesr; 12§§. (1988). (DHHS Publication Nol ADM 88-

1257). Washington, DC: U.S. Government Printing Office.

Swan, H. L., Press, A. N., & Briggs, S. L. (1985). Child sexual abuse
prevention: Does it work? Chiid Weifiars, EMIV, 395-405.

Wall, H. R. (1983). Ch' d a t abu e eventio o'ect‘ Pilot
progrsn evaluation. Unpublished manuscript.

Willson, V. L., & Putnam, R. R. (1982). A meta-analysis of pretest
sensitization effects in experimental design. American Educational
Research Journal, i2(2), 249-258.

Wolfe, D. A., MacPherson, T., Blount, R., & Wolfe, V. V. (1986).
Evaluation of a brief intervention for educating school children
in awareness of physical and sexual abuse. Gniig Abuse and
Neglect, iQ, 85-92.

Wurtele, S. K. (1990a, September). v uatin o rams desi ed to
ve c d se ua abu e' Meas rin t e s'x R's. Paper
presented at the 8th International Congress on Child Abuse and
Neglect, Hamburg, Germany.

Wurtele, S. K. (1990b). Teaching personal safety skills to four-year-
old children: A behavioral approach. Behavior Therapy, El, 25-
32.

Wurtele, S. K., Currier, L. L., Gillispie, E. I., & Franklin, C. F.
(1991). The efficacy of a parent-implemented program for teaching
preschoolers personal safety skills. Behavior Therapy 22 69-83.

Wurtele, S. K., Kast, L. C., & Kondrick, P. A. (1988, August).
Qeveiopment of an instrument to evaluate sexuai abuse prsyention

programs. Paper presented at the American Psychological Associa-
tion convention, Atlanta, GA.

 

Wurtele, S. K., Marrs, S. R., & Miller-Perrin, C. L. (1987). Practice
makes perfect? The role of participant modeling in sexual abuse
prevention programs. Journal of Consulting and Clinical
Psychology, §§(4), 599-602.

73

Wurtele, S. K., & Miller-Perrin, C. L. (1987). An evaluation of side
effects associated with participation in a child sexual abuse

prevention program. Journal of School Mealtn, £1, 228-231.

Wurtele, S. K., Saslawsky, D. A., Miller, C. L., Marrs, S. R., &
Britcher, J. C. (1986). Teaching personal safety skills for
potential prevention of sexual abuse: A comparison of treatments.

gpurnsi of Consulting and Clinicai Esychology, EA, 688-692.

Wyatt, G. E., & Powell, G. J. (1988). Identifying the lasting effects of
child sexual abuse: An overview. In G. E. Wyatt & G. J. Powell

(Eds.), Lasting effects of child sexual abuse (pp. 11-18). Newbury
Park, CA: Sage.

Zigler, E., & Hall, N. W. (1989). Physical child abuse in America: Past,
present, and future. In D. Cicchetti & V. Carlson (Eds.), Child
maitreatment (pp. 38-75). New York: Cambridge University Press.

AN STATE UNIVERSITY LIB

IIIIIIIIIIIIIIIII I III

I II IIIIIIIIIIIIIIIIIIIIIII

31293 02723 8199