Journal of Abnormal ChiMPsyehology, Vol. 7, Vol. 7, No. 4, 1979, pp. 397-404
Help-Agents' Views About Clinical Interactions with Acting-Out Children 1 Ellis L. Gesten, Emory L. Cowen, 2 and Arthur R. Orgel
University of Rochester Elyse Schwartz
Vanderbi# University
A scale to measure nonprofessional child-aides' views about, and practices in, working primarily with acting-out children is described. The scale was used as part o f an overall evaluation o f a program to teach the aides Ginottian limitsetting approaches for work with such children. A prior study had shown that the training was followed by significantly more favorable treatment outcomes. Compared to 44 nontrained child-aMes, the 19 trained aides had significantly higher postprogram scale scores on opinions and beliefs about acting-out children, changes in feeling about working with them, and actual observed changes in their playroom behaviors. Item analysis indicated that, following training, ames felt more comfortable with and had a richer repertoire o f techniques for dealing with acting-out children. Specifically, they found it easier to set limits and to deal with overtly aggressive behaviors. Those changes may be key factors in explaining the significantly greater effectiveness o f the trained aides working with acting-out children. Cowen, Oregel, Gesten, & Wilson (1979, this volume) recently reported a program to train nonprofessional child-aides in a specific intervention modality for work with young acting-out, school-maladapted children. Their work was set in the larger context of the Primary Mental Health Project (PMtlP), a program Manuscript received in final form February 9, 1979. 1This study was supported by a grant (MH-11427-02) from the NIMH Experimental and Special Training Branch, for which the authors express sincere appreciation. Thanks also to Mary Boike, Michael A. DeStefano, and Alice B. Wilson for their contributions to the data analyses. 2Address all correspondence to Dr. Emory L. Cowen, Department of Psychology, University of Rochester, Rochester, New York 14627. 397 0091-0627/79/1200-0397503.00/0 9 1 9 7 9 P l e n u m P u b l i s h i n g Corporation
Gesten, Cowen,Orgel, and Schwartz
398
for early detection and prevention of school adjustment problems (Cowen, Trost, Lotion, Dorr, Izzo, & Isaacson, 1975). Two considerations prompted it: (1) consistent findings from the extended child-therapy literature (e.g., Levitt, 1971) that aggressive acting-out children had poor therapy outcomes and (2) recent PMHP studies showing that acting-out children in PMHP (a) did significantly less well than shy-anxious ones (Lorion, Cowen, & Caldwell, 1974) and (b) were seen as the most difficult and least enjoyable youngsters to work with, and as having the poorest program prognoses (Cowen, Gesten, & DeStefano, 1977; DeStefano, Gesten, & Cowen, 1977). The new training program, as described by Cowen et al. (1979, this volume), was based on Ginott's (1961) constructive limit-setting approach. Its main emphases were on the needs (1) for therapy to proceed in a physically and psychologically "safe" environment; (2) to distinguish between destructive, hostile, antisocial feelings (always acceptable) and direct behavioral expressions of such feelings (neve r acceptable); and (3) to provide appropriate symbolic (sublimatory) outlets for the latter. That orientation led to a four-step technique of therapeutic intervention: reflection of feelings, intervention, sublimatory teaching, and reflection of feelings evoked by intervening (Orgel, 1980). Aides trained in the new approach had better child-outcomes, both generally and with acting-out children, than equally skilled, experienced aides who did not have the training (Cowen et al., 19791 this volume). That study, however, did not consider the questions of (a) what aides actually learned in the program, (b) changes in their sense of comfort and feelings about working with acting-out children, and (c) changes in the playroom behaviors observed with such children. The present study develops an instrument to assess those variables, on the assumption that the prior positive program findings for acting-out children were due to the specifics of the training program. If so, a comparison of trained versus untrained aides should reflect differential (1) beliefs about how to work with referred, particularly acting-out, children, (2) changes in feelings about working with them, and (3) changes in children's actual playroom behaviors that parallel the program's concepts and thrusts.
METHOD
Subjects Subjects were 63 PMHP child-aides, of whom 19 went through training in consecutive-year groups of 10 and 9, and 44 did not have the additional training. Although the two groups were comparable in age, background, and selection criteria, a higher proportion of Es came from PMHP CORE (laboratory-demonstration) shoools and thus tended to be somewhat more experienced.
Help-Agents' Views
399
Scale The training program's main emphases helped to frame specific scale items. The program's conceptual teachings should theoretically have changed aides' beliefs ("opinions," on the scale) about good therapy with children, particularly A children. The program's precepts plus opportunities to practice the new approach might lead to changes in the aides' feelings ("views," on the scale) of comfort and confidence in treating A children. And putting program principles into practice should have changed children's behavior in the playroom. Those, at least, were the three areas that the scale sought to tap. Three of the authors, each with background and experience in child-clinical work and knowledge of both PMHP and the new training approach, generated a sizable initial pool of items. After dropping ambiguous and overlapping items, 55 items (13 opinion, 26 change in view, and 16 change in behavior) remained. 3 Opinions. These 13 items were stated as positive or negative assertions about aspects of the play interaction, both generally (e.g., "Children are not likely to express deep personal feelings to an aide"; "An aide's specific techniques are unimportant; the relationship she has with a child is what counts") and specifically with respect to acting-out (e.g., "It's not desirable for a child to express strong anger in any form in the playroom"; " I f you have to set limits with child in the playroom, there's probably something wrong with the relationship"). Aides responded on Likert scales (5 = strongly agree and 1 = strongly disagree). Items were keyed so that high scores reflected agreement with the new approach. Change in View. This group of 26 items included two related clusters: (a) aides' specific handling techniques and practices (e.g., "Know how to handle a child who 'acts u p ' " ; "Know what words to use. when I respond to a child"), and (b) aides' feelings or reactions about interactions with children (e.g,, "Feel threatened when a child deliberately provokes me"; "Feel comfortable when a child gets really angry"). These items also included both general child-interaction issues and ones pertaining specifically to acting-out children. The common element linking them was the aides' feelings of comfort and mastery in the playroom. Section 2 items were answered in terms of an 18-month time frame, beginning just before the first group started its training and ending after the second group finished training. The task for each item was to judge, on 7-point scales, whether a practice or feeling better described the respondent before or after the program. Items were keyed so that high scores reflected stronger feelings of comfort and mastery, with the system's teachings, over time. Change in Behavior. All 16 items, preceded by the hypothetical stem "The children I s e e . . . ," described behaviors or reactions of children in the 3Actually the scales, as distributed, included 60 items. Five items (1 opinion, 3 change in view, 1 change in behavior) were later dropped because of disagreements in keying.
400
Gesten, Cowen, Orgel, and Schwartz
playroom. The section included general items (e.g., "Get into expressing feelings fairly rapidly"; "learn the 'rules' of the playroom") plus ones that dealt with acting-out related issues (e.g., "Say negative things about parents and family members"; "Express anger through words or symbolic play"). Section 3 items were rated on the same 7-point, change-over-time scale used for Section 2. Items were keyed so that high scores reflected the occurrence of behaviors and responses consistent with the training program's goals. The questionnaire was distributed by mail to all Ss in June 1976. Only those returned by August 1976, after one follow-up request, were included in the study.
RESULTS Table I presents scale means and sigmas for both groups, mean differences and t ratios testing their significance. Trained aides had significantly higher scores on all measures. Structurally comparable analyses were also done for individual items; Table II lists items on which group differences exceeded p ~< .10. Seventeen items, 15 of which were changes in view or behavior, discriminated significantly (p ~< .05) between groups. Differentiators included both general items and those pertaining specifically to acting-out children.
DISCUSSION The study's main purpose was to develop a measure of (nonprofessional) clinicians' beliefs about how best to interact with children, particularly actingout children, in the playroom. Scale items were based on Ginott's (1961) conceptual framework for viewing and dealing with such youngsters. His approach
Table I. Comparison of Program and Nonprogram Aides' Opinions, Changes in View, and Changes in Behavior Program aides "(N = 19) Scale section
X
Opinions Changes in view Changes in behavior Total
55.58 149.58 90.22 295.37
cr 2.93 12.76 9.32 18.90
Nonprogram adies (iV = 44) X
a
Dm
t
p
52.36 136.18 84.30 272.85
6.28 24.26 9.36 26.67
3.22 13.40 5.92 22.52
2.77 2.86 2.31 3.83
.01 .01 .05 .001
Help-Agents' Views
401
Table II. Scale Items Differentiating Between Program and Nonprogram Aides
Program aides (N = 19) Opinions No. 5. The aide's maternal quality is her most important asset. 6. Children are not likely to express deep personal feelings to an aide. 7. No matter what children's specific problems are, aides should be using pretty much the same techniques. 11. One of the best ways to establish rapport with a child is to engage in an interesting or " f u n " game with him. Changes in view 16. Feel confident in what I do, as an aide, with children, 17. Know how to handle a child who "acts up." 19. Have an adequate repertoire of responses to cover the situations I get into with children. 21. Know a variety of techniques to use with children. 23. Know how to handle personally hostile actions by a child. 26. Feel comfortable when a child gets really angry. 28. Have an adequate "language of communication" for children in the playroom. 33.
Feel motherly toward the children I see.
34.
Feel comfortable in handling such "unpleasant" behaviors as tantrums or strong expressions of hostility. 37. Feel comfortable working with children who have predominant acting-out and aggression problems. 42, Know what words to use when I respond to a child. Changes in behavior The children I see 47, say negative things about parents and family members.
Nonprogram aides (N = 44)
P
3.84
3.30
1.81
.10
4.89
4.48
2.24
.02
4.00
4.55
-1.66
.10
4.32
2.61
5.79
.01
6.63
6.30
1.94
.10
6.58
5.72
4.36
.001
6.37
5.42
4.02
.001
6.32
5.93
2.00
.05
6.37
5.67
2.20
.05
6.47
5.77
3.90
.001
6.16
5.79
1.65
.10
5.21
4.43
2.25
.02
6.37
5.74
2.7l
.01
6.58
5.74
4.42
.001
6.11
5.61
2.42
.02
5.74
5.21
2.29
.02
Gesten, Cowen, Orgel, and Schwartz
402 Table II. Continued
Program aides (N = 19) 48. say negative things about teachers, classmates, and class happenings. 49. require that I set limits. 51. accept their own anger. 52. show assertive behaviors in the playroom. 53. learn the "rules" of the playroom. 56. accept limits, constructively.
Nonprogram aides (N = 44)
t
p
5.66
5.07
2.39
.02
5.74 5.63 5.46
4.05 5.09 4.86
5.99 2.00 1,61
.001 .05 .10
6.05 6.26
4.86 5.32
4.04 4.38
.001 .001
emphasizes distinctions between feelings and acts, conditional and unconditional empathy, and includes the active teaching of sublimatory skills. Nineteen nonprofessional child-aides were trained in the approach, which they used under supervision, in clinical interactions with young acting-out children in schools. Trained aides had significantly higher scores than nontrained control aides on the scale's three main sections: opinions and beliefs about interventions with children, especially acting-outers; feelings of comfort in working with them; and observed behavior changes in playroom interactions with them. Those findings suggest that trained aides absorbed the program's precepts, put them into practice, and, as a result, came to feel more comfortable with a group of children seen as the most difficult to work with of all child problem groups (Cowen et al., 1977; DeStefano et al., 1977), and as having poor therapy outcomes (Levitt, 1971). Although conceptually "desirable" changes in attitude and practice suggest that the teaching program "took," the acid test of program effectiveness is whether ultimate target persons (in this case young acting-out school children) have benefitted from it. Thus the present findings must be linked with the prior demonstration that the trained aides had significantly more favorable therapeutic outcomes with acting-out children, compared either to control aides or to themselves before training (Cowen et al., 1979, this volume). Item analysis data provide cues about program elements that might have contributed to those earlier positive findings. In general, items reflecting greater aide comfort and mastery and positive change (as defined by the "system") in children's playroom behaviors, over time, better differentiated trained from nontrained aides than intellectualized opinions about treatment. Aides' greater confidence manifested itself in two (related) ways: (1) feeling that they had a richer, more appropriate repertoire of intervention tools (e.g., "Know how to handle a child who 'acts u p ' " ; "Have an adequate language of communication";
Help-Agents' Views
403
"Know how to handle personally hostile reactions by a child") and (2) feeling more comfortable in their actual playroom contacts with children (e.g., "Feel confident in what I do as an aide"; "Feel comfortable in handling tantrums or strong expressions of anger"). Change-in-behavior item-differentiators also reflected two dimensions: (1) limit-setting (e.g., "Accept limits constructively"; "Learn the rules of the playroom") and (b) the occurrence and handling of hostility (e.g., "Accept their own anger"; "Show assertive behaviors in the playroom"). Knowing better, and feeling more comfortable about, how to respond to the provocative, oft-threatening behaviors of acting-out children, whether in direct response or through the therapeutic use of limit-setting, may well be key factors that contributed to the positive outcomes found for acting-out children seen by trained aides in the earlier study (Cowen et al., 1979, this volume). One other aspect of the data bears mention. Although most of the specific item-differentiators concerned acting-out children, several were general. Illustratively, although trained aides did not devalue the therapeutic importance of maternal qualities, they relied less than nontrained aides on those qualities alone. Similarly, they depended less on interesting and "fun" games as rapport-builders. Several items cited above (e.g., "acquiring a richer language of communication" and "repertoire of techniques") also pertained to children in general, not just acting-outers. Such differentiators help to explain some of the prior study's (Cowen et al., 1979, this volume) "generalization" effects. Thus, although the approach was targeted to acting-out children, the trained aides found it applicable to other children as well; the latter also had more favorable therapeutic outcomes. The present study thus makes two main contributions: (1) developing a measure, based on a cohesive conceptual framework, of clinicians' reactions to, and practices with, acting-out children; and (2) pinpointing specific changes in views and playroom happenings that may be important preconditions for positive therapeutic outcomes with that heretofore difficult clinical group.
REFERENCES
Cowen, E. L. Some problems in community program evaluation research. Journal of Consulting and Clinical Psychology, 1978, 46, 792-805. Cowen, E. L., Gesten, E. L., & DeStefano, M. A. Nonprofessional and professional help agents' views of interventions with young maladapting school children. American Journal of Community Psychology, 1977, 5, 469-479. Cowen, E. L., Orgel, A. R., Gesten, E. L., & Wilson, A. B. The evaluation of an intervention program for young schoolchildren with acting-out problems. Journal o f Abnorreal Child Psychology, 1979, 7, 381-396. Cowen, E. L., Trost, M. A., Lorion, R. P., Dorr, D., Izzo, L. D., & Isaacson, R. V. New ways in school mental health: Early detection and prevention o f school maladaptation.
New York: Human SciencesPress, 1975.
404
Gesten, Cowen, Orgel, and Schwartz
DeStefano, M. A., Gesten, E. L., & Cowen, E. L. Teachers' views of the treatability of children's school adjustment problems. Journal o f Special Education, 1977, 11, 275-280. Ginott, H. G. Group psychotherapy with children. New York: McGraw-Hill, 1961. Levitt, E. E. Research on psychotherapy with children. In A. E. Bergin & S. L. Garfield (Eds.), Handbook o f psychotherapy and behavior change: An empirical analysis. New York: Wiley, 1971. Lotion, R. P., Cowen, E. L., & Caldwell, R. A. Problem types of children referred to a school based mental health program. Journal of Consulting and Clinical Psychology, 1974, 42, 491-496. Oregel, A. R. Haim Ginott's approach to discipline. In D. Dorr & M. Zax (Eds.), Comparative approaches to discipline for children and youth. New York: Springer, 1980.
Help-Agents' Views About Clinical Interactions with Acting-Out Children 1 Ellis L. Gesten, Emory L. Cowen, 2 and Arthur R. Orgel
University of Rochester Elyse Schwartz
Vanderbi# University
A scale to measure nonprofessional child-aides' views about, and practices in, working primarily with acting-out children is described. The scale was used as part o f an overall evaluation o f a program to teach the aides Ginottian limitsetting approaches for work with such children. A prior study had shown that the training was followed by significantly more favorable treatment outcomes. Compared to 44 nontrained child-aMes, the 19 trained aides had significantly higher postprogram scale scores on opinions and beliefs about acting-out children, changes in feeling about working with them, and actual observed changes in their playroom behaviors. Item analysis indicated that, following training, ames felt more comfortable with and had a richer repertoire o f techniques for dealing with acting-out children. Specifically, they found it easier to set limits and to deal with overtly aggressive behaviors. Those changes may be key factors in explaining the significantly greater effectiveness o f the trained aides working with acting-out children. Cowen, Oregel, Gesten, & Wilson (1979, this volume) recently reported a program to train nonprofessional child-aides in a specific intervention modality for work with young acting-out, school-maladapted children. Their work was set in the larger context of the Primary Mental Health Project (PMtlP), a program Manuscript received in final form February 9, 1979. 1This study was supported by a grant (MH-11427-02) from the NIMH Experimental and Special Training Branch, for which the authors express sincere appreciation. Thanks also to Mary Boike, Michael A. DeStefano, and Alice B. Wilson for their contributions to the data analyses. 2Address all correspondence to Dr. Emory L. Cowen, Department of Psychology, University of Rochester, Rochester, New York 14627. 397 0091-0627/79/1200-0397503.00/0 9 1 9 7 9 P l e n u m P u b l i s h i n g Corporation
Gesten, Cowen,Orgel, and Schwartz
398
for early detection and prevention of school adjustment problems (Cowen, Trost, Lotion, Dorr, Izzo, & Isaacson, 1975). Two considerations prompted it: (1) consistent findings from the extended child-therapy literature (e.g., Levitt, 1971) that aggressive acting-out children had poor therapy outcomes and (2) recent PMHP studies showing that acting-out children in PMHP (a) did significantly less well than shy-anxious ones (Lorion, Cowen, & Caldwell, 1974) and (b) were seen as the most difficult and least enjoyable youngsters to work with, and as having the poorest program prognoses (Cowen, Gesten, & DeStefano, 1977; DeStefano, Gesten, & Cowen, 1977). The new training program, as described by Cowen et al. (1979, this volume), was based on Ginott's (1961) constructive limit-setting approach. Its main emphases were on the needs (1) for therapy to proceed in a physically and psychologically "safe" environment; (2) to distinguish between destructive, hostile, antisocial feelings (always acceptable) and direct behavioral expressions of such feelings (neve r acceptable); and (3) to provide appropriate symbolic (sublimatory) outlets for the latter. That orientation led to a four-step technique of therapeutic intervention: reflection of feelings, intervention, sublimatory teaching, and reflection of feelings evoked by intervening (Orgel, 1980). Aides trained in the new approach had better child-outcomes, both generally and with acting-out children, than equally skilled, experienced aides who did not have the training (Cowen et al., 19791 this volume). That study, however, did not consider the questions of (a) what aides actually learned in the program, (b) changes in their sense of comfort and feelings about working with acting-out children, and (c) changes in the playroom behaviors observed with such children. The present study develops an instrument to assess those variables, on the assumption that the prior positive program findings for acting-out children were due to the specifics of the training program. If so, a comparison of trained versus untrained aides should reflect differential (1) beliefs about how to work with referred, particularly acting-out, children, (2) changes in feelings about working with them, and (3) changes in children's actual playroom behaviors that parallel the program's concepts and thrusts.
METHOD
Subjects Subjects were 63 PMHP child-aides, of whom 19 went through training in consecutive-year groups of 10 and 9, and 44 did not have the additional training. Although the two groups were comparable in age, background, and selection criteria, a higher proportion of Es came from PMHP CORE (laboratory-demonstration) shoools and thus tended to be somewhat more experienced.
Help-Agents' Views
399
Scale The training program's main emphases helped to frame specific scale items. The program's conceptual teachings should theoretically have changed aides' beliefs ("opinions," on the scale) about good therapy with children, particularly A children. The program's precepts plus opportunities to practice the new approach might lead to changes in the aides' feelings ("views," on the scale) of comfort and confidence in treating A children. And putting program principles into practice should have changed children's behavior in the playroom. Those, at least, were the three areas that the scale sought to tap. Three of the authors, each with background and experience in child-clinical work and knowledge of both PMHP and the new training approach, generated a sizable initial pool of items. After dropping ambiguous and overlapping items, 55 items (13 opinion, 26 change in view, and 16 change in behavior) remained. 3 Opinions. These 13 items were stated as positive or negative assertions about aspects of the play interaction, both generally (e.g., "Children are not likely to express deep personal feelings to an aide"; "An aide's specific techniques are unimportant; the relationship she has with a child is what counts") and specifically with respect to acting-out (e.g., "It's not desirable for a child to express strong anger in any form in the playroom"; " I f you have to set limits with child in the playroom, there's probably something wrong with the relationship"). Aides responded on Likert scales (5 = strongly agree and 1 = strongly disagree). Items were keyed so that high scores reflected agreement with the new approach. Change in View. This group of 26 items included two related clusters: (a) aides' specific handling techniques and practices (e.g., "Know how to handle a child who 'acts u p ' " ; "Know what words to use. when I respond to a child"), and (b) aides' feelings or reactions about interactions with children (e.g,, "Feel threatened when a child deliberately provokes me"; "Feel comfortable when a child gets really angry"). These items also included both general child-interaction issues and ones pertaining specifically to acting-out children. The common element linking them was the aides' feelings of comfort and mastery in the playroom. Section 2 items were answered in terms of an 18-month time frame, beginning just before the first group started its training and ending after the second group finished training. The task for each item was to judge, on 7-point scales, whether a practice or feeling better described the respondent before or after the program. Items were keyed so that high scores reflected stronger feelings of comfort and mastery, with the system's teachings, over time. Change in Behavior. All 16 items, preceded by the hypothetical stem "The children I s e e . . . ," described behaviors or reactions of children in the 3Actually the scales, as distributed, included 60 items. Five items (1 opinion, 3 change in view, 1 change in behavior) were later dropped because of disagreements in keying.
400
Gesten, Cowen, Orgel, and Schwartz
playroom. The section included general items (e.g., "Get into expressing feelings fairly rapidly"; "learn the 'rules' of the playroom") plus ones that dealt with acting-out related issues (e.g., "Say negative things about parents and family members"; "Express anger through words or symbolic play"). Section 3 items were rated on the same 7-point, change-over-time scale used for Section 2. Items were keyed so that high scores reflected the occurrence of behaviors and responses consistent with the training program's goals. The questionnaire was distributed by mail to all Ss in June 1976. Only those returned by August 1976, after one follow-up request, were included in the study.
RESULTS Table I presents scale means and sigmas for both groups, mean differences and t ratios testing their significance. Trained aides had significantly higher scores on all measures. Structurally comparable analyses were also done for individual items; Table II lists items on which group differences exceeded p ~< .10. Seventeen items, 15 of which were changes in view or behavior, discriminated significantly (p ~< .05) between groups. Differentiators included both general items and those pertaining specifically to acting-out children.
DISCUSSION The study's main purpose was to develop a measure of (nonprofessional) clinicians' beliefs about how best to interact with children, particularly actingout children, in the playroom. Scale items were based on Ginott's (1961) conceptual framework for viewing and dealing with such youngsters. His approach
Table I. Comparison of Program and Nonprogram Aides' Opinions, Changes in View, and Changes in Behavior Program aides "(N = 19) Scale section
X
Opinions Changes in view Changes in behavior Total
55.58 149.58 90.22 295.37
cr 2.93 12.76 9.32 18.90
Nonprogram adies (iV = 44) X
a
Dm
t
p
52.36 136.18 84.30 272.85
6.28 24.26 9.36 26.67
3.22 13.40 5.92 22.52
2.77 2.86 2.31 3.83
.01 .01 .05 .001
Help-Agents' Views
401
Table II. Scale Items Differentiating Between Program and Nonprogram Aides
Program aides (N = 19) Opinions No. 5. The aide's maternal quality is her most important asset. 6. Children are not likely to express deep personal feelings to an aide. 7. No matter what children's specific problems are, aides should be using pretty much the same techniques. 11. One of the best ways to establish rapport with a child is to engage in an interesting or " f u n " game with him. Changes in view 16. Feel confident in what I do, as an aide, with children, 17. Know how to handle a child who "acts up." 19. Have an adequate repertoire of responses to cover the situations I get into with children. 21. Know a variety of techniques to use with children. 23. Know how to handle personally hostile actions by a child. 26. Feel comfortable when a child gets really angry. 28. Have an adequate "language of communication" for children in the playroom. 33.
Feel motherly toward the children I see.
34.
Feel comfortable in handling such "unpleasant" behaviors as tantrums or strong expressions of hostility. 37. Feel comfortable working with children who have predominant acting-out and aggression problems. 42, Know what words to use when I respond to a child. Changes in behavior The children I see 47, say negative things about parents and family members.
Nonprogram aides (N = 44)
P
3.84
3.30
1.81
.10
4.89
4.48
2.24
.02
4.00
4.55
-1.66
.10
4.32
2.61
5.79
.01
6.63
6.30
1.94
.10
6.58
5.72
4.36
.001
6.37
5.42
4.02
.001
6.32
5.93
2.00
.05
6.37
5.67
2.20
.05
6.47
5.77
3.90
.001
6.16
5.79
1.65
.10
5.21
4.43
2.25
.02
6.37
5.74
2.7l
.01
6.58
5.74
4.42
.001
6.11
5.61
2.42
.02
5.74
5.21
2.29
.02
Gesten, Cowen, Orgel, and Schwartz
402 Table II. Continued
Program aides (N = 19) 48. say negative things about teachers, classmates, and class happenings. 49. require that I set limits. 51. accept their own anger. 52. show assertive behaviors in the playroom. 53. learn the "rules" of the playroom. 56. accept limits, constructively.
Nonprogram aides (N = 44)
t
p
5.66
5.07
2.39
.02
5.74 5.63 5.46
4.05 5.09 4.86
5.99 2.00 1,61
.001 .05 .10
6.05 6.26
4.86 5.32
4.04 4.38
.001 .001
emphasizes distinctions between feelings and acts, conditional and unconditional empathy, and includes the active teaching of sublimatory skills. Nineteen nonprofessional child-aides were trained in the approach, which they used under supervision, in clinical interactions with young acting-out children in schools. Trained aides had significantly higher scores than nontrained control aides on the scale's three main sections: opinions and beliefs about interventions with children, especially acting-outers; feelings of comfort in working with them; and observed behavior changes in playroom interactions with them. Those findings suggest that trained aides absorbed the program's precepts, put them into practice, and, as a result, came to feel more comfortable with a group of children seen as the most difficult to work with of all child problem groups (Cowen et al., 1977; DeStefano et al., 1977), and as having poor therapy outcomes (Levitt, 1971). Although conceptually "desirable" changes in attitude and practice suggest that the teaching program "took," the acid test of program effectiveness is whether ultimate target persons (in this case young acting-out school children) have benefitted from it. Thus the present findings must be linked with the prior demonstration that the trained aides had significantly more favorable therapeutic outcomes with acting-out children, compared either to control aides or to themselves before training (Cowen et al., 1979, this volume). Item analysis data provide cues about program elements that might have contributed to those earlier positive findings. In general, items reflecting greater aide comfort and mastery and positive change (as defined by the "system") in children's playroom behaviors, over time, better differentiated trained from nontrained aides than intellectualized opinions about treatment. Aides' greater confidence manifested itself in two (related) ways: (1) feeling that they had a richer, more appropriate repertoire of intervention tools (e.g., "Know how to handle a child who 'acts u p ' " ; "Have an adequate language of communication";
Help-Agents' Views
403
"Know how to handle personally hostile reactions by a child") and (2) feeling more comfortable in their actual playroom contacts with children (e.g., "Feel confident in what I do as an aide"; "Feel comfortable in handling tantrums or strong expressions of anger"). Change-in-behavior item-differentiators also reflected two dimensions: (1) limit-setting (e.g., "Accept limits constructively"; "Learn the rules of the playroom") and (b) the occurrence and handling of hostility (e.g., "Accept their own anger"; "Show assertive behaviors in the playroom"). Knowing better, and feeling more comfortable about, how to respond to the provocative, oft-threatening behaviors of acting-out children, whether in direct response or through the therapeutic use of limit-setting, may well be key factors that contributed to the positive outcomes found for acting-out children seen by trained aides in the earlier study (Cowen et al., 1979, this volume). One other aspect of the data bears mention. Although most of the specific item-differentiators concerned acting-out children, several were general. Illustratively, although trained aides did not devalue the therapeutic importance of maternal qualities, they relied less than nontrained aides on those qualities alone. Similarly, they depended less on interesting and "fun" games as rapport-builders. Several items cited above (e.g., "acquiring a richer language of communication" and "repertoire of techniques") also pertained to children in general, not just acting-outers. Such differentiators help to explain some of the prior study's (Cowen et al., 1979, this volume) "generalization" effects. Thus, although the approach was targeted to acting-out children, the trained aides found it applicable to other children as well; the latter also had more favorable therapeutic outcomes. The present study thus makes two main contributions: (1) developing a measure, based on a cohesive conceptual framework, of clinicians' reactions to, and practices with, acting-out children; and (2) pinpointing specific changes in views and playroom happenings that may be important preconditions for positive therapeutic outcomes with that heretofore difficult clinical group.
REFERENCES
Cowen, E. L. Some problems in community program evaluation research. Journal of Consulting and Clinical Psychology, 1978, 46, 792-805. Cowen, E. L., Gesten, E. L., & DeStefano, M. A. Nonprofessional and professional help agents' views of interventions with young maladapting school children. American Journal of Community Psychology, 1977, 5, 469-479. Cowen, E. L., Orgel, A. R., Gesten, E. L., & Wilson, A. B. The evaluation of an intervention program for young schoolchildren with acting-out problems. Journal o f Abnorreal Child Psychology, 1979, 7, 381-396. Cowen, E. L., Trost, M. A., Lorion, R. P., Dorr, D., Izzo, L. D., & Isaacson, R. V. New ways in school mental health: Early detection and prevention o f school maladaptation.
New York: Human SciencesPress, 1975.
404
Gesten, Cowen, Orgel, and Schwartz
DeStefano, M. A., Gesten, E. L., & Cowen, E. L. Teachers' views of the treatability of children's school adjustment problems. Journal o f Special Education, 1977, 11, 275-280. Ginott, H. G. Group psychotherapy with children. New York: McGraw-Hill, 1961. Levitt, E. E. Research on psychotherapy with children. In A. E. Bergin & S. L. Garfield (Eds.), Handbook o f psychotherapy and behavior change: An empirical analysis. New York: Wiley, 1971. Lotion, R. P., Cowen, E. L., & Caldwell, R. A. Problem types of children referred to a school based mental health program. Journal of Consulting and Clinical Psychology, 1974, 42, 491-496. Oregel, A. R. Haim Ginott's approach to discipline. In D. Dorr & M. Zax (Eds.), Comparative approaches to discipline for children and youth. New York: Springer, 1980.
