OUP user menu

Impact of a Computer-assisted Education Program on Factors Related to Asthma Self-management Behavior

Ross Shegog PhD, L. Kay Bartholomew EdD, MPH, Guy S. Parcel PhD, Marianna M. Sockrider MD, DrPH, Louise Mâsse PhD, Stuart L. Abramson MD, PhD
DOI: http://dx.doi.org/10.1136/jamia.2001.0080049 49-61 First published online: 1 January 2001


Objective: To evaluate Watch, Discover, Think and Act (WDTA), a theory-based application of CD-ROM educational technology for pediatric asthma self-management education.

Design: A prospective pretest posttest randomized intervention trial was used to assess the motivational appeal of the computer-assisted instructional program and evaluate the impact of the program in eliciting change in knowledge, self-efficacy, and attributions of children with asthma. Subjects were recruited from large urban asthma clinics, community clinics, and schools. Seventy-six children 9 to 13 years old were recruited for the evaluation.

Results: Repeated-measures analysis of covariance showed that knowledge scores increased significantly for both groups, but no between-group differences were found (P = 0.55); children using the program scored significantly higher (P < 0.01) on questions about steps of self-regulation, prevention strategies, and treatment strategies. These children also demonstrated greater selfefficacy (P < 0.05) and more efficacy building attribution classification of asthma self-management behaviors (P < 0.05) than those children who did not use the program.

Conclusion: The WDTA is an intrinsically motivating educational program that has the ability to effect determinants of asthma self-management behavior in 9- to 13-year-old children with asthma. This, coupled with its reported effectiveness in enhancing patient outcomes in clinical settings, indicates that this program has application in pediatric asthma education.


This paper describes an evaluation of the impact of Watch, Discover, Think and Act (WDTA), a computer-based educational program designed to teach asthma self-management skills to urban, minority children.13 The impact evaluation assesses the program's immediate effect on knowledge, self-efficacy, and attributions, which are variables related to the target behavior of self-management. It also examines the motivational aspects of the program on children who use it.

Asthma Self-management

Asthma is among the most common chronic diseases in the United States, and its medical and social consequences are more severe in inner-city populations.4 Asthma is characterized by chronic inflammation of the airways that leads to episodes of bronchospasm and excess mucus production. It has a high prevalence (about 5 percent of children under 18 years old) and contributes significantly to school absenteeism.57 Asthma morbidity and mortality rates have increased substantially among children since 1980, and the asthma death rate among children 5 to 14 years old rose from 1.7 to 3.2 per million between 1980 and 1993.8 Asthma morbidity, mortality, and hospitalization rates have been disproportionately high among the poor and medically underserved. African American and Hispanic children in the inner cities have been reported to be more likely to have asthma than white children. The annual prevalence of asthma among children living in inner cities is 1.5 to 2 times higher than that of the U.S. population as a whole.912

Asthma requires constant self-management by the patient to maintain control of symptoms, prevent exacerbations, attain normal lung function, and maintain normal activity levels. Self-management refers to the behaviors that people with asthma and their family members perform to lessen the impact of this chronic illness (Table 1). Self-management includes adherence to medical regimens as well as the complex cognitive-behavioral tasks of self-monitoring, decision making, and communicating about both symptoms and treatment regimens.1315 Determinants of asthma self-management behavior include a person's behavioral capability for management behaviors1618 and degree of self-efficacy in performing those behaviors.13,1719 Attribution has also been implicated in children's self-management behaviors and in control of asthma and other diseases.2022

View this table:
Table 1

Asthma Self-management

▪   Monitor symptoms of asthma “directly” and compare with personal standard. Use object measures (e.g, peak flow meter) to monitor and compare symptoms with personal standard.
▪   Monitor for personal environmental triggers
▪   Monitor asthma self-management efforts and compare with personal standard
Identify problems:
▪   Using monitoring (as described above), identify when a problem exists.
Implement solutions:
▪   Keep regular appointments with health care providers.
▪   Refer to asthma action plan.
▪   Maintain medication for chronic condition as prescribed.
▪   Maintain “normal” exercise levels.
▪   On the basis of symptoms or the environment, make medication adjustments, including administration of rescue medication, as prescribed.
▪   Avoid or remove asthma triggers.
▪   Call health care professional in an acute situation.
▪   Communicate with family members and with health care providers.
▪Evaluate success of actions and return to monitoring.

Computer-based Patient Education

Computers have diverse uses in health promotion and disease prevention—for example, as tools for education on AIDS and responsible sexuality and as adjuncts to medical therapy for alcohol rehabilitation and rheumatoid arthritis.2326 The use of virtual fantasy worlds, such as the Starbright Pediatric Network for pediatric inpatients, has been shown to reduce dependence on medications and to enhance pain management.27 The Health Touch database system has been shown to facilitate physician–patient interaction in primary care practice.28 Furthermore, empirical research has identified the value of using computers to provide behavior-change messages tailored to client belief characteristics, such as stage of change and health beliefs, and to demographic characteristics, such as gender and ethnicity.29

Several computer-based applications have been developed to assist in asthma management and education. For adult asthma patients, computer-assisted instruction (CAI) has been used to help them monitor and avoid house dust-mite allergen.30 Several CAI programs have been developed for children with asthma. These include Clubhouse Asthma,31 Bronkie the Bronchiasaurus,32 Wee Willie Wheezie,33 Air Academy: The Quest for Airtopia,34 and Asthma Command.35 Bronkie the Bronchiasaurus has been shown to positively affect knowledge, self-efficacy, and communication about asthma in children who use it.36 Airtopia has been shown to positively affect asthma knowledge in children when used in the context of a general health curriculum.34 Asthma Command has been evaluated in clinic settings where children who used the program showed increases in knowledge and in self-reported asthma management, compared with children in the control group.35 There were no demonstrated differences in visits to physicians, emergency rooms, or hospitals.

Although these trends are encouraging, there remains a need for programs that can be tailored to the asthma characteristics of the individual child and that can help patient and families learn self-management skills. We describe here such a program, WDTA, a second-generation asthma CAI program that has been rigorously evaluated.3

Watch, Discover, Think and Act Computer-assisted Instruction

The WDTA computer-based education program has taken a motivational approach to teaching asthma self-management skills to urban, minority children.13 The program is a multimedia application that uses three types of computer-based instructional strategies—a simulation of real-world activities in which the child can learn and practice self-regulatory processes; tutorials with which the child can learn and practice asthma-specific skills; and a game treatment to enhance motivation. The broad objectives of the program are to provide asthma self-management skills training as an adjunct to medical care and enhance clinical care provided for the asthma patient by supplying information to health care providers and parents regarding the child's asthma self-management capabilities and progress.

The program specifically targets change in factors that may be determinants of this self-management. The program was developed using social cognitive theory change methods to improve the child's knowledge, self-efficacy, and attributions.37 These methods include verbal reinforcement, guided practice with feedback, persuasion, goal setting, incentives, and symbolic modeling. The program addresses the need to individualize asthma education and to teach self-regulatory skills, components of what Creer et al.38 refer to as a second generation of asthma self-management programs.38

Overview of the Game Procedure and Graphical User Interface

Watch, Discover, Think and Act is an interactive multimedia computer program for providing intensive, tailored self-management education to inner-city children, of upper elementary and middle school ages, who have asthma. Designed for use in primary care clinics and physician's offices, it gives these children the opportunity—in a safe, non-threatening, even fun, computer environment—to learn how to manage their asthma .

Originally developed on CD-ROM for the Apple platform, using MacroMind Director 5.0 authoring software,39 WDTA makes use of text, graphics, animation, sound, and video clips.

The program comprises four stages—data input, introduction, game scenarios, and data output. The first use of the program occurs after the encounter with the physician. A clinic staff member helps the child with data input, using the keyboard and mouse to provide information about the child's physician and about the child, such as name, age, and duration of asthma. Information on personal asthma symptoms and environmental triggers, medications, and peak flow make up the child's asthma profile.

Once this information has been entered, the child self-navigates through the program using the mouse. During the introduction stage, the child chooses a character and a coach. The character represents a child approximately 12 years old who has asthma. The child can choose whether the character and coach are male or female and can choose between African American and Hispanic ethnicity. The coach is a teenager, slightly older than the target population for the game, who has learned to manage asthma.

The child is given an overview of the game, which involves a mission to rescue plans for anti-pollution technology from the castle of Dr. Foulair. The goal of the game is to move the character through three real-life scenarios with multiple scenes (home, school, and neighborhood) and then to Dr. Foulair's castle. To progress from one scenario to the next, the child must successfully manage the character's asthma by following the four self-regulation steps of watching (monitoring symptoms and environmental triggers, taking maintenance or preventive medication, and keeping appointments), discovering (deciding whether an asthma problem exists and what its probable cause is), thinking (deciding on a list of possible actions), and acting (choosing an action such as taking rescue or symptom relief medicine, removing or avoiding triggers, getting help). The child must also collect mission handbook pages to complete tutorials on asthma and click and drag tools to be used in the castle.

The game scenarios include 18 real-world and 4 castle situations that present the problems that inner-city children with asthma must deal with successfully to manage their disease. The game screen shown in Figure 1 shows one of the scenario scenes containing environmental asthma triggers (dust, fur, and feathers), tools (the peak flow meter in the table drawer), and mission handbook tutorial pages (on the bookshelf). Tutorial components contain information developed de novo as well as components drawn from National Institutes of Health source material.40,41 In the tutorials, video segments depicting medicine-taking procedures were obtained by permission to use pre-existing educational materials.42

Figure 1

Watch, Discover, Think and Act computer game screen.

In Figure 1, four self-regulatory icons for WDTA are vertically aligned to the right of the scene, near the center of the screen. Below these, a fifth icon represents a cellular phone with which to contact the coach. The health status boxes on the far right of the screen show the character's symptoms, peak flow, medication taking, appointment scheduling, and personal triggers; they appear during the “watch” step of the asthma self-management process. The WDTA program tracks each child's progress so that children can resume the game where they left off on a previous visit.

The data output stage provides information for the child, the parents, and the health care provider. Players receive constant feedback from the game as to their progress, and at the end of each session children receive a certificate that congratulates them on the progress made and reiterates the four steps of self-regulation. To guide further asthma education, the parents and physician receive a progress report that indicates the scenario the child reached, the child's use of the self-regulation steps, and tutorial scores.


Design and Sample

A prospective pretest posttest trial with randomly assigned intervention and comparison subjects was used for the study. The WDTA program was evaluated in terms of its effect on cognitive impact variables of knowledge, self-efficacy, and attribution; factors associated with successful asthma management. The hypothesis was that children using the intervention would experience improvement in these cognitive variables. A sample of 76 children was recruited from six clinics and seven schools in a large urban area.

Five subjects were lost to posttesting because of lack of availability for follow-up (n= 4) and refusal to participate further (n= 1). The resulting study sample comprised 71 children who were 8 to 13 years old and had been diagnosed as having asthma by their health care providers. The mean age of subjects in the study was 10.7 years. Forty-six boys and 25 girls completed the posttest. Based on subject self-report, the study sample was primarily white non-Hispanic (47.9 percent) and African American (40.8 percent). The remaining five children were Hispanic (11.3 percent).

Intervention and Data Collection Protocol

Children were recruited over ten months and assigned to intervention or comparison groups after parents gave written consent and children signed assent forms. The children in the intervention group received intervention in the form of playing WDTA, and those in the comparison group received no intervention. Data for each subject were collected in three sessions over a three-week period. Sessions were held on a university campus affiliated with a large medical center. Baseline data were collected from all children in session one. Measures included paper-and-pencil assessments of knowledge, self-efficacy, and attributions regarding asthma self-management.

Session two for each subject occurred one week after session one. In session two, children in the intervention group independently used WDTA. Following children's completion of the program instructions, the investigator assessed each learner's comprehension and provided further orientation to ensure understanding of how to use the program correctly. The children progressed through the first three levels of the program. At the end of the third level, the investigator asked the children whether they wanted to continue to the fourth level or to stop using the program. Process data of achievement and time on task was collected in this session. Children in the comparison group received no intervention in the second session and were not required to spend time at the study site. Session three occurred one week later. In this session, posttest data were collected from each child. In addition to the measures previously listed, these data included attitudes toward computer-assisted instruction.


For data collection, a pencil-and-paper format was used and questionnaires were given to each subject or parent (depending on the instrument) to complete without assistance. However, the investigator observed children completing example items of each questionnaire to ensure that they understood the instructions before they proceeded.

Child Knowledge of Asthma Management

Knowledge was assessed in two ways—by a questionnaire and by three open-ended questions. The Child Knowledge of Asthma Management Questionnaire was developed to assess change in knowledge in children between 7 and 12 years old and was based on the educational objectives of the design document for WDTA. Each item provided the child with a statement about asthma self-management and possible responses of “yes,” “no,” and “don't know.” The instrument was pilot-tested on a study sample of 101 children who had diagnoses of asthma. In the pilot sample, the Cronbach coefficient alpha for the 30-item questionnaire was 0.86 for children 7 to 13 years of age, and scores ranged from 5 to 30. The mean was 20.4 (±6.1), and skewness was –0.99 (SE skew, 0.26). In a clinical study involving 171 urban children with asthma who were 7 to 16 years of age, the alpha was 0.73.2

At posttest, children were also asked three open-ended questions about strategies for asthma self-management, and their answers were matched to a set of possible correct responses. The questions (and ranges of accurate responses) were as follows: What are the four steps that you should follow to manage your asthma? (0–4) What can you do to stop a problem before it starts? (0–10) What can you do to stop an asthma problem after it starts? (0–16)

Child Self-efficacy for Asthma Self-management

Self-efficacy is the belief that one has the skill and ability necessary to perform a behavior in a variety of circumstances and in the face of various obstacles.37 Self-efficacy was assessed using a 23-item questionnaire developed to determine change in confidence in performing self-regulatory and asthma-specific behaviors in children 7 to 12 years old. The questionnaire included confidence in monitoring symptoms, environment, medicine taking, and health care use; confidence in deciding whether there is an asthma problem; and confidence in determining appropriate solutions and acting on them. Pilot testing on 101 children with asthma revealed an internal consistency using a Cronbach coefficient alpha of 0.88. In the intervention trial of 171 urban-dwelling children with asthma who were 7 to 16 years old, the Cronbach coefficient alpha was 0.77, indicating acceptable internal consistence.2

Causal Attributions

Attributions relate to the belief that self-management behavior is controllable and is subject to personal effort.43 Attributions were assessed using a 22-item questionnaire developed for this study to determine change in causal attribution along the dimensions of locus and stability in children 7 to 12 years old. The item format was modeled on that used in the Children's Attributional Style Questionnaire, also known as the KASTAN-R44 (N. Kaslow, personal communication, 1994). For each item in the questionnaire, the child is provided with a positive or negative self-management outcome. The subject is asked to imagine experiencing the outcome presented. An example of a positive outcome is “you go camping and have no asthma problems the whole time.” An example of a negative self-management outcome is “you always have breathing problems when you swim.” Two stems accompany each outcome. The child is asked to choose the stem that gives the best reason why the event would happen to him or her. For each item, one attribution dimension (i.e., locus or stability) is varied while the other dimension is kept constant. Pilot-testing of this questionnaire determined coefficient alphas for the locus subscales, which were 0.54 (success), 0.53 (failure), and 0.64 (combined). Coefficient alphas for the stability subscales were 0.70 (success), 0.67 (failure), and 0.74 (combined). The locus and stability subscales were not significantly correlated (r = 0.14, P = 0.31).

Attribution Classification

Children were asked to classify the four principal self-management behaviors (taking care of asthma, avoiding triggers, watching for symptoms, and taking medicine) in terms of the attributional dimensions of locus, stability, and controllability. Children were provided with a self-management behavior (e.g., taking medicine) and asked to check boxes on a semantic differential scale. This instrument uses a format based on Russell's causal dimension scale, which was developed to assess causal attribution dimensions in adults.45 Scale items assessing perceptions of locus were “something other people do for you or get you to do” vs. “something you do yourself.” Scale items assessing perception of control were “something you cannot control” vs. “something you can control.” Scale items assessing perception of stability were “something you sometimes do” vs. “something you always do.” The coefficient alpha for the 12-item scale is 0.75, based on baseline data from the 71 children in the current study.

Demographic and Health Information

Demographic information was obtained from primary caregivers, who were asked about their child's medicines, personal best peak flow, asthma triggers, and symptoms. Information on the use of emergency and hospital services, school grades, and absenteeism was also obtained. Parents also completed a six-item severity scale, developed by Rosier et al.,46 which classified severity as low, moderate, mild, or high. The scale assesses number of episodes, coughing or wheezing, and curtailment of activities. An item reliability of 0.89 has been reported by the developers of the scale.46 In the intervention trial of 171 inner-city children with asthma, the coefficient alpha for the scale was 0.77.2

Computer Experience

Children were also asked three questions about their computer experience. Responses to a question about how often they used computers ranged from “not at all” to “a few times a day.” To find out what they used computers for, the children were asked to check all that applied from a list of eight common applications, such as word processing, games, and graphics or art and to list other uses. The children indicated where they used computers by checking all that applied from home, school, and friend's home and by listing other locations.


The motivational impact of the WDTA (referred to as intrinsic motivation) was assessed using methodology similar to that described by Parker and Lepper.47 In that study, children were given a choice of computer treatments that they could use. In the current study, motivational value of the program was assessed by determining the number of children continuing to the end of the program (level four) after being given the option to stop using the program at level three. Children were also asked to indicate on a four-point Likert scale how much they liked the computer game, liked to work with it, and liked the story that went with it. Possible responses ranged from “I didn't like it” to “I liked it very much.” In addition, the children were asked to compare WDTA with their favorite board game, computer game, and school subject and with other asthma education they had had, using a three-point Likert scale with categories of “less fun,” “as much fun,” and “more fun.” The children were interviewed about the game and asked what they would tell their best friend and other children with asthma about the game.

Attitude toward Computer-assisted Learning

Attitude toward computer-assisted instruction was assessed with the Attitude toward Computer-assisted Learning scale by Askar et al.48 The scale consists of ten items, scored on a three-point response scale— yes (3), sometimes (2), and no (1); the negatively worded items are reversed to a positive direction for scoring purposes. Askar et al. report the alpha reliability estimate of the total score to be 0.81.48

Process of Computer Use

Four characteristics of the child's computer use were collected—fidelity, time on task, tutorial score, and levels completed. The fidelity of program use was observed by the investigator, who rated children on seven categories using a five-point scale (never, rarely, sometimes, usually, always). The categories were as follows: needing assistance; understanding the program directions; following the WDTA self-regulatory sequence; making appropriate decisions in the game scenarios; engaging with the game scenarios; and attending to the tutorial segments.

Data Analysis

Sample Size Estimates

Data from the pilot test of the Child Knowledge of Asthma Management and the Child Self-efficacy for Asthma Self-management questionnaires with 101 children 7 to 13 years of age with asthma were used to calculate sample size requirements. On the basis of these data and predetermined estimates of alpha at 0.05 and power at 0.80, a sample of 80 subjects was calculated to be adequate to determine a difference 0.4 SD between groups using an analysis of covariance (ANCOVA) procedure.49

Statistical Analysis

The study sought to answer two questions: Is the computer program intrinsically motivating for children who use it? Do those children exposed to the computer program experience significantly greater change in knowledge, self-efficacy, and attributions than children who are not exposed?

To answer the first question, an 80 percent threshold for wanting to continue playing the game beyond the third level was set to indicate that the program was intrinsically motivating. To answer the second question, separate 2 × 2 repeated-measures analyses of variance were conducted to determine whether a significant change in knowledge, self-efficacy, and attribution had occurred in the intervention group compared with the comparison group. Analysis of covariance was conducted on the open-ended knowledge questions, which were collected only at posttest. Pretest scores from the Child Self-management Knowledge Questionnaire were used as covariates for this analysis.


Asthma Severity, Computer Use, and Demographic Variables

Forty-six percent of the study population was classified as having moderate-to-severe asthma. Most of the children in the study were reported by their parents as receiving above-average grades at school, with 73 percent reporting typical grades of As and Bs. Another 23.9 percent of children's parents reported their child's typical grades to be Bs and Cs. The children in the study were also experienced with computers. One third of the children in the study sample reported using computers a few times a day, and 90 percent of them reported using computers a few times a month. Computers were used principally in the school (81.7 percent of the children) and the home (61.9 percent). Computers were used for a variety of purposes, including school projects (70.4 percent), games (63.4 percent), and word processing (56.3 percent). In the study sample, 97.2 percent of caregivers had completed high school or had some college or a higher degree, and most study subjects came from two-parent households (68.6 percent). Most primary caregivers reported being employed full-time (67.1 percent), and 4.3 percent were employed half-time. Only 9.9 percent of the study sample were Medicaid recipients.

Differences between Groups at Baseline

There were no statistically significant differences between the two groups in any of the variables that might have affected performance on the CAI program (Table 2). The groups were also compared on mean impact variables at baseline. The mean score for the knowledge pretest was found to be significantly greater in the intervention group (P = 0.027). No difference was found for the other variables.

View this table:
Table 2

Demographic and Severity Variables, by Number (%), for the Total Sample, Intervention, and Control Groups

VariableTotal Study Population, N = 71Intervention Subjects, n = 38Comparison Subjects n = 33Significance
   Male46 (64.8)25 (65.8)21 (63.9)
   Female25 (35.2)13 (34.2)12 (36.4)
   Hispanic5 (7.0)2 (5.3)3 (9.1)
   Asian1 (2.6)1 (2.6)
   Black29 (40.8)16 (42.1)13 (39.4)0.53b
   White34 (47.9)17 (44.7)17 (51.5)
   Other2 (2.8)2 (5.3)
Asthma severity:
   Low11 (15.5)5 (13.2)6 (18.2)
   Mild27 (38.0)13 (34.2)14 (42.4)
   Moderate25 (35.2)17 (44.7)8 (24.2)
   Severe8 (11.3)3 (7.9)5 (15.2)
School grades:
   As & Bs52 (73.2)25 (65.8)27 (81.8)
   Bs & Cs17 (23.9)12 (31.6)5 (15.2)0.27b
   Cs & Ds2 (2.8)1 (2.6)1 (3.0)
Computer experience:
   None at all
   Few times/yr7 (9.9)3 (7.9)4 (12.1)
   Few times/mo17 (23.9)11 (28.9)6 (18.2)0.73b
   Few times/wk24 (33.8)12 (31.6)12 (36.4)
   Few times/day23 (32.4)12 (31.6)11 (33.3)
  • Note: Mean ages, in years (SD), were as follows: total study population, 10.69 (±1.14); intervention subjects, 10.92 (±1.17); comparison subjects, 10.42 (±1.06); signficance, 0.067a.

  • a Two-tailed t-test

  • b Chi-square test.

Motivational Value of Watch, Discover, Think and Act

With respect to the intrinsic motivation of the program, all the children in the intervention group who had time to do so (32 of 38) continued to use the program when told they could stop. This represents 84.2 percent of the sample and exceeds the 80 percent a priori criteria set for the study. The six children who did not proceed to the fourth level did not complete the entire game because of time constraints but indicated that they would have continued to play if time had permitted.

No significant pretest-posttest difference was found between the groups regarding attitude about using computers as a learning medium. The children in the intervention group showed increased positive attitudes toward computers following the intervention, but this was not significant (P = 0.327). The children indicated that they liked the game (91.1 percent) and liked working with the computer (86.3 percent). They also liked the story that went with the program (72.3 percent). The children indicated that the WDTA program was as much or more fun than their favorite board game (88.9 percent), computer game (72.2 percent), subject in school (91.6 percent), and other asthma education (91.7 percent).

Program Effectiveness

With respect to the effectiveness of the program, Table 3 presents pretest and posttest mean scores for the impact variables. Knowledge posttest scores were significantly higher than pretest scores for the study sample (F1 = 37.87, P = 0.00). However, repeated-measures analysis of variance revealed no between-group differences in knowledge (F1 = 0.35, P = 0.55). A posttest-only analysis was found to be significant between the groups for the open-ended questions about the four-step self-regulation process (F1 = 189.18, P = 0.00), prevention strategies (F1 = 12.33, P = 0.00), and treatment strategies (F1 = 17.48, P = 0.00). Because of the difference in baseline knowledge between the groups, an ANCOVA analysis was conducted using pretest knowledge scores as the covariate. Children in the intervention group had significantly greater scores on the open-ended knowledge questions than children in the comparison group when we controlled for knowledge pretest scores (all P < 0.01).

View this table:
Table 3

Pretest and Posttest Means for Impact Variables of Knowledge, Self-Efficacy, and Attribution

VariableGroupnPretest Mean (SD)Posttest Mean (SD)
Knowledge questionnaireIntervention3818.6 (±5.1)21.1 (±5.4)
Control3315.7 (±5.8)17.8 (±6.3)
Control320.06 (±0.3)
   PreventionIntervention382.7 (±1.1)
Control331.8 (±1.0)
   TreatmentIntervention382.7 (±1.4)
Control331.5 (±0.8)
Self-efficacy* Intervention3853.4 (±9.7)56.5 (±9.8)
Control3251.6 (±9.7)51.5 (±10.7)
   Locus (+)Intervention352.6 (±1.0)2.9 (±0.9)
Control322.9 (±0.9)3.0 (±0.9)
   Locus (–)Intervention373.2 (±1.1)3.5 (±0.8)
Control323.3 (±0.7)3.6 (±0.7)
   Locus (Total)Intervention355.6 (±1.8)6.4 (±1.6)
Control326.3 (±1.3)6.6 (±1.2)
   Stability (+)Intervention374.3 (±1.8)4.8 (±1.8)
Control334.2 (±1.8)4.5 (±1.7)
   Stability (-)Intervention364.8 (±1.7)5.0 (±1.7)
Control324.6 (±1.8)5.3 (±1.7)
   Stability (Total)Intervention369.0 (±2.8)9.9 (±2.7)
Control328.8 (±2.3)5.3 (±1.7)
Attribution classificationIntervention3736.1 (±7.1)39.5 (±6.4)
Control3335.3 (±7.9)36.5 (±6.7)
  • * Outlier removed.

Assessment of the self-efficacy change scores revealed the existence of an influential data point greater than three SD from the mean. Using the criteria provided by Stevens,50 this case represents an outlier. Subsequent analysis of the self-efficacy data was conducted with and without the outlier in the study sample. When the outlier was included (n= 71), repeated-measures analysis revealed no difference between the groups (F1 = 2.33, P = 0.13) with an observed power of 0.325. When the outlier was omitted, a significant improvement in self-efficacy was apparent (F1 = 4.45, P = 0.04), with an observed power of 0.547. The influential case did not differ significantly from other cases on the other variables tested.

Analysis of the total score and subscale scores of the Asthma Self-management Attribution Questionnaire revealed no difference between the intervention and comparison groups. However, the total score for attribution classification revealed a significant difference between the intervention and comparison groups; children in the intervention group exhibited significantly more positive attributions with respect to asthma self-management (F1 = 4.45, P = 0.04). Observed power for this analysis was 0.548.

Process Variables

Observations of the children's progress on the computer program revealed that most did not require assistance with the program (69.4 percent). This is probably because standard help was provided to all children at the beginning of the first scenario. Most children (80.6 percent) followed the program directions. However, they tended to stray from the watch-discover-think-act self-regulatory sequence. Sixty-six percent of the children were observed to use the process only sometimes, rarely, or never. Most children (64.6 percent) usually or always made appropriate game decisions in the scenarios, such as choosing appropriate triggers and solutions to an asthma problem or taking appropriate actions for an asthma problem. Furthermore, most (91.7 percent) usually or always understood and attended to the tutorials. The children were observed to be always or usually engaged in the game activities (97.2 percent). Most (94.4 percent) were observed to usually answer the tutorial questions correctly. This is reflected by the children's high achievement scores on the tutorial questions; the mean scores for all tutorials were greater than 87 percent.


Attention, Motivation and Appeal

A fundamental tenet of social cognitive theory is that if learning is to occur, the learner must attend to what is learned.37 This program's ability to attract children's attention and to keep them engaged in the learning activity is encouraging. All the children in the study wanted to continue to the end of the game, and they remained fully engaged in using the program, even after periods as long as 2.5 hours. This is even more compelling when we consider that the children in the study were sophisticated computer users. They compared the WDTA computer application favorably with other games and education programs they had used. Krendl and Lieberman's research in computer-based education51 suggested that effects on motivation might be simply effects of the novelty of the medium. Given that the computer was not new to the children in this study, we could not attribute the program's success to novelty alone. Children in both groups had very positive views of using computers to learn.

Snyder and Palmer52 noted the potential for instructional designers to create intricate, involving, and illustrated contexts into which educational activities can be embedded. Proponents have noted the potential motivational advantages of such procedures,47,5254 whereas opponents have viewed such developments as “sugarcoated” instruction that is likely to produce less lasting and less efficient learning.54 While the debate over “edutainment” continues, it must be acknowledged that although an education program may maintain the attention of the learner, it is only a first step in engendering change in the educational variables the program is designed to affect.

Enhancement of Knowledge, Self-Efficacy and Attribution

Children who used WDTA were able to provide a more extensive array of behavioral strategies for asthma management. They were able to reiterate the components of a four-step problem-solving framework, describe behavioral strategies to prevent asthma episodes, and describe behavioral strategies to treat asthma symptoms to a significantly greater extent than those in the comparison group. The potential of the WDTA computer program to positively influence asthma self-management behavioral capability is further demonstrated by the between-group difference in self-efficacy. Enhanced self-efficacy theoretically leads to greater likelihood of the child attempting asthma self-management at home and to greater persistence in the endeavor.37 Confidence to carry out self-management behaviors was not altered in the comparison group but improved significantly in the intervention group. The simulated experience of successfully negotiating a series of asthma self-management situations and performing simulated self-regulatory actions could change a child's self-perceived ability to perform these behaviors. The application provides a number of methods to elicit self-efficacy change in learners, including persuasion related to self-management actions, guided practice with corrective feedback, and social and symptom reinforcement for self-management success.3

The application is also designed to elicit change in attributions regarding asthma self-management behavior. To engender a move to more internal attributions, the learner is provided with direct control over the asthma self-management of a chosen character and is accompanied by an older peer coach who persuades the character that asthma self-management is largely in his or her own hands. The coach also demonstrates that self-management is based largely on effort and is not subject to external authority figures or to circumstances that cannot be changed. Repeated self-management experiences and situations reiterate the message that self-management failure is due to unstable (changeable) causes, such as lack of effort or modifiable environmental circumstances that can be managed in the future.

A significant group difference was apparent in how children classified asthma self-management behaviors. Children who had used the computer program classified the behaviors as more internal (“something I do myself”), more controllable (“something I can control”), and more stable (“something I always do”). This movement in perceived attributions indicates a move to feelings of greater autonomy in the children using the computer program. This change in attribution is theoretically related to enhanced self-efficacy and is important if these children are to be active asthma self-managers.55,56

Taken together, these results suggest that the program can influence factors related to asthma self-management. The strength of these findings should be considered in light of the relatively small sample size of 71 children.

Study Limitations

The study was limited because of the small sample size and some measurement problems. The significant change in knowledge scores in both groups points to the measurement instrument being a potential intervention in itself. This was a group of children with above-average school grades and, perhaps, above-average motivation. We might, therefore, predict some learning from noticing what they didn't know on pretest and seeking to find the answers. Parallel forms of the Asthma Knowledge Questionnaire might have provided a more stringent test. Although WDTA was effective in affecting behavioral determinants in a laboratory setting with children of moderate socioeconomic status, the effectiveness of the program in changing outcomes in a clinical setting of low-socioeconomic-status, predominantly minority children was unknown. The efficacy of the program in a laboratory setting indicated that a broader clinical-based study was warranted.

Effects on Self-management and Behavioral Outcomes: Use in Clinical Settings

A subsequent prospective pretest-posttest clinical trial with randomly assigned intervention and comparison subjects in four inner-city pediatric asthma clinics provided information on how the program might affect asthma self-management behavior.2 The clinic sites served an inner city, primarily Medicaid-recipient (government-funded) population. Two sites were affiliated with large teaching hospitals headed by respiratory specialists with staffs of rotating fellows. The other two clinics were inner-city community sites staffed by pediatricians. Inclusion in the study was voluntary. No sites refused participation.

The computers were mounted on trolleys for ease of transportation. Clinic rooms or offices provided for the children to play the game were isolated from the distractions of clinic activity. Patients who met the inclusion criteria—age (6 to 17 years), moderate-to-severe asthma (as defined by their physicians), English speaking (parents could be Spanish speaking), and no chronic disease other than asthma—were invited by the research coordinator to participate. Patient participation was voluntary.

On the first visit the patient would see the physician and then be introduced to the computer by research staff for data entry and orientation. The first session with the computer comprised data entry, orientation, and play through at least one scenario game screen. This session lasted approximately 40 minutes. The intention was for the child to use the program unassisted; however, research assistants were available if the child needed help. On subsequent visits to the clinic, the children played the computer game for approximately 30 minutes before seeing the physicians. The computer's record keeping allowed the children to continue the program from where they left off. After each play session, the child and the physician received printouts with reinforcing messages, a reminder of the self-regulatory process taught, and the scores in the game. Research assistants completed the self-management sections of the plan, and physicians completed the sections on medications and treatment for episodes. Children in the comparison group continued to have regularly scheduled clinic appointments.

Results of the outcome study in the clinical sites indicated that children who were older and those who scored higher at pretest improved their knowledge of how to manage asthma. Self-management improved for intervention children who had a more conservative estimate of their self-efficacy at pretest and for children with higher pretest scores. Children in the intervention group had a lower rate of hospitalization, and there were differences in functional status that suggest improvement for children in the intervention group. Use of the self-management program was associated with a decrease in symptoms for those children whose symptoms were milder.

Feasibility for Clinic Use

Support for asthma self-management must include the participation of health care providers so that education can be individualized to the child's treatment regimen; getting this participation can prove challenging.57 Health care providers work under growing time constraints, and in the two evaluation studies, enlisting their participation was difficult. The WDTA program is by no means a substitute for an National Asthma Education Program–recommended asthma action plan. Rather, the two work in concert. The information from the child's asthma action plan is incorporated into the game parameters of WDTA, and the program gives the child simulated experience in managing asthma according to his or her plan. Optimally, the health care provider collaborates with the family in creating an action plan and helps them identify the child's triggers. In our experience, physicians usually fail to provide the action plan. Wide dissemination of the program will require support, such as an implementation manual and training to enable the health care team to incorporate the use of the action plan and help the child use the computer game. An additional challenge is managing the system. Although self-contained and easily maintained, the computer still requires staff to troubleshoot, enter data, and introduce the patient to the program. The WDTA program was subsequently withdrawn from the clinics pending ongoing evaluation of the system based on program modifications.

Program Modifications

The results of these two evaluation studies allowed us to plan a second iteration of WDTA. Three major modifications were planned and executed. They were development of a version for younger children, increase in the amount of guidance and feedback for the older children regarding their use of the self-regulatory processes, and development of Spanish-language versions.

Development of the Program for Younger Children

We found that the instructions were not adequate to allow the learners to go through the program on their own. Another problem with the program was the amount of detailed information offered in the tutorial segments. Although the 9- to 13-year-olds in the impact study remained engaged in these tutorials, younger learners in the clinic-based study found the information presented overwhelming. Furthermore, the indicators of health status were too complex for the younger children (ages 7 to 8 years). Therefore, a new version was developed that incorporated a smaller amount of tutorial information into the simulations, simplified the health status feedback, and provided more coaching throughout.

More Guidance Regarding the Self-regulatory Process

The observations that many children could not explain the self-regulatory buttons and that some of the children in the two evaluation studies tended to not follow the four-step self-regulatory sequence in order prompted the development of increased program guidance, reinforcement, and control. In the new version the children are guided to use watch, discover, think, then act in that order, and they receive more obvious score accrual for doing so.

Spanish Version

The WDTA program was developed for urban, low-income children and was evaluated in the southwestern United States, where many children have Spanish as a first language. Therefore, the second iteration included a Spanish translation that was made using both focus group methods to test possible translations of asthma-related words as well as the back-translation method that has become the accepted standard in health education.58 Currently, all versions of the program are being used and evaluated as a part of a multi-component health-education and health services program and evaluation study in 60 elementary schools in grades one through five.

Future Research

The primary recommendation, and next logical step, for future research related to this and other technologically based programs is to conduct a diffusion study to determine the best way to achieve dissemination, adoption, implementation, and maintenance in clinics. Additional research issues involve examining the effect of parental asthma education on child behavior, conducting an evaluation of the program outside clinics, examining the effects of providing further tailoring in the program, evaluating the effectiveness of versions designed for younger children and for Spanish speakers, examining the effect of the program on patient–physician interaction, and adapting and evaluating the program for other diseases.


  • This work was supported in part by National Institutes of Health contract N01 H039220 from the National Heart, Lung, and Blood Institute and by the Texas Children's Hospital, Children's Asthma Center.


View Abstract