SINGLE SYSTEM RESEARCH DESIGN

Decreasing Excessive Media Usage While Increasing Physical Activity A Single-Subject Research Study Karen H. Larwin Gannon University David A. Larwin Kent State University Salem

The Kaiser Family Foundation released a report entitled Kids and Media Use in the United States that concluded that children’s use of media—including television, computers, Internet, video games, and phones—may be one of the primary contributor’s to the poor fitness and obesity of many of today’s adolescents. The present study examines the potential of increasing physical activity and decreasing media usage in a 14-year-old adolescent female by making time spent on the Internet and/or cell phone contingent on physical activity. Results of this investigation indicate that requiring the participant to earn her media-usage time did correspond with an increase in physical activity and a decrease in media-usage time relative to baseline measures. Five weeks after cessation of the intervention, the participant’s new level of physical activity was still being maintained. One year after the study, the participant’s level of physical activity continued to increase.

Keywords: adolescence; media usage; physical exercise; reinforcement

Childhood obesity and the corresponding lack of physical fitness in ado-lescent children have become a national epidemic. According to a 2002 Centers for Disease Control report, the number of obese and physically unfit adolescents in the United States has tripled since 1980. A number of researchers have suggested that one of the primary contributors to adoles- cent obesity is the increasing use of media, such as television, Internet,

Behavior Modification Volume 32 Number 6

November 2008 938-956 © 2008 Sage Publications

10.1177/0145445508319668 http://bmo.sagepub.com

hosted at http://online.sagepub.com

938

Authors’ Note: Please address correspondence to Karen H. Larwin, Gannon University, Organizational Learning & Leadership, 1225 Palumbo Academic Center, Erie, PA 16541; phone: 814-871-5792; fax: 814-871-7652; e-mail: larwin001@gannon.edu.

video games, and cell phones. Although a number of organizations, including the American Academy of Pediatrics (2003), theorize that there is a direct link between television viewing and rates of obesity, the research does not unilaterally support that supposition. For example, a number of studies conducted in the years between 1980 and 2000 (e.g., Anderson, Crespo, Bartlett, Cheskin, & Pratt, 1998; Crespo et al., 1996; Dietz & Gortmaker, 1993; Gortmaker et al., 1996; Lowry, Wechsler, Galuska, Fulton, & Kann, 2002) found an inverse significant relationship between television viewing and the child’s level of fitness. However, some studies, including some more recent studies, have failed to find this same significant relationship between television viewing and obesity (Durant & Baranowski, 1994; Proctor et al., 2003; Robinson et al., 1993; Robinson & Killen, 1995).

Kunkel (2001) suggests that strength in the relationship between obesity and television usage is more appropriately explained by “what” is being watched. According to Kunkel, the more commercials viewed, especially food-related advertising, the stronger the relationship becomes. Kunkel reports that the average child will see about 40,000 television commercials per year, with the majority of the advertisements during child-directed pro- gramming being food-related advertisements. As a result of the information available on the relationship between childhood obesity and television viewing, the American Academy of Pediatrics (2003) has suggested that children’s television use be limited to 1 to 2 hours of quality programming per day. What is unfortunate about these recommendations is that this report, and much of the research, has ignored the potential contribution of Internet and computer usage to adolescent obesity and has instead simply focused on television viewing.

According to the Kaiser Family Foundation (1999) report on Kids and Media Use, today’s young people are no longer just sitting in front of the television. This study reports that children and adolescents (3rd grade to 12th grade) are spending in excess of 6 ½ hours a day with various forms of media. Meyering (2005) refers to today’s adolescents as Generation-M, because today’s media-saturated youth are on the Internet, while the televi- sion is playing in the background, and they are talking on cell phones. A report by the U.S. Entertainment and Media Consumer Survey (Parr, 2005) indicates that American children are spending as much time on the Internet on a weekly basis as they are in front of the television. Because children can communicate with a number of friends at one time through instant-messaging services, they are spending increasingly more time on the Internet, whereas the television is merely background noise. This change in media usage by

Larwin, Larwin / Adolescent Fitness 939

adolescents might help to explain why the once reliable relationship between television and obesity has been found to be waning or nonexistent as the presence of the computer/Internet in the home has steadily grown in the last few years. In light of this research that children are constantly exposed to the television but not necessarily watching it, because they are simultaneously engaged in other forms of media use, these multiple forms of media (telephones, cell phones, computers and Internet, television, and video games) should all be considered when studying the relationship between media usage and adolescent fitness.

Several studies have been conducted which have targeted reducing the amount of television viewing while increasing physical activity (e.g., Jason, 1987; Jason & Brackshaw, 1999; Jason & Johnson, 1995; Jason, Johnson, & Jurs, 1993; Jason & Klich, 1982; Johnson & Jason, 1996). These studies have shown that requiring children to earn their television usage, by activ- ities such as working out on a stationary exercise bicycle, have been effec- tive in reducing each child’s television viewing, while increasing the child’s overall activity level. In one such single-subject study, Jason and Brackshaw (1999) set up a contingency plan in which an 11-year-old female participant was required to earn her television viewing time with an equivalent amount of time on a stationary bicycle. After 5 weeks, the young female’s televi- sion viewing time dropped from the baseline measure of 4 plus hours a day to about 1 hour a day. Her parents also reported an overall increase in phys- ical activity, but the researchers themselves did not attempt to track the child’s level of activity at any stage in the project. Increased activity and reduced television viewing levels were maintained after the termination of the formal intervention (17 to 18 weeks after baseline). As a result of the project, the young female lost 20 pounds in 72 days.

However, these studies do not focus on physical exercise as the primary target behavior. They did make television viewing contingent on perform- ing physical exercise, but they did not specifically track and record exer- cise. Although it may seem peculiar to set up such contingencies and not track the target behavior, the focus of the studies was actually on reducing television-viewing time rather than on increasing the target behavior. Jason and Brackshaw (1999) suggest that the pattern of results they observed in their study may be the result of a response competition. Increase in physi- cal activity competes with available time for media usage; thus, the latter decreases. However, Jason and Brackshaw were not able to assess this spec- ulation more precisely because they did not specifically track and record physical activity. The authors called for more research to investigate this possibility more fully.

940 Behavior Modification

Larwin, Larwin / Adolescent Fitness 941

The present investigation incorporates a contingency between physical exercise and media use as a potential reinforcing consequence. Although in this case focus of the investigation is on the target behavior of physical exercise, it also incorporates several features of the Jason and Brackshaw (1999) study. Jason and Brackshaw focused on reducing television view- ing. Similarly, an additional goal of the present study was an attempt to reduce the media usage of the participant, specifically Internet and cell phone usage. Although the Jason and Brackshaw study did not attempt to increase physical activity specifically, results of their study suggest that attempts to increase physical activity as a primary goal by making media usage contingent on such activity may also have the effect of reducing media use through response competition. The present study was designed to test this possibility.

The focus of the current investigation was the development of a plan to increase the physical activity and decrease media usage in a 14-year-old female by making her access to media contingent on performing physical exercise. Thus, the primary goal of the present study was to encourage a lifestyle change involving more physical exercise by the participant while simultaneously getting the participant to take more personal responsibility for being physically fit. It was predicted that getting the child to take per- sonal responsibility for her physical fitness would be necessary to assist and support the acquisition and maintenance of the primary goal. The secondary goal of the present study aimed to reduce the amount of time the child spent on the Internet and using the cell phone. To accomplish these goals, the par- ticipant was required to perform physical activity, in the form of treadmill use, in exchange for nonschool-related time on the Internet during Treatment Phase 1, and both Internet and phone time during Treatment Phase 2.

Method

Participant

For the present study, the assenting participant was a 14-year-old adoles- cent female who had been prescribed by both her physician and counselor to engage in increased levels of physical activity, to deal with boredom and complaints about unhappiness regarding her level of physical fitness. The child’s counselor referred the parents to the researcher, who then assisted the parents and child in designing and implementing the behavior change program.

The participant was 5’2” and weighed approximately 135 pounds at the onset of the study. In the previous 9 months, the participant followed a prescribed eating plan in an effort to lose excess weight, during which she lost a total of 30 pounds. However, parents report that the child’s weight loss was the result of dietary changes and not because of any increase in exercise. The participant indicated that she was still concerned about her weight because she had regained about 10 pounds after the initial weight loss during the 3 months immediately before the onset of the study. The participant reported frustration that the attempts to lose the weight she regained had not been successful with further self-directed efforts in the form of dietary adjustments.

The participant admitted that she finds exercise and physical activity in general to be aversive. She also admitted that she enjoys time on the Internet, chatting with friends, and cell phone communication. However, the participant reported that she did not watch an excessive amount of tele- vision. The participant indicated that homework and her participation in organized after-school activities competed with her ability to begin the walking regimen that had been recommended by her physician. Other than her brief interactions with friends at school and during the bus-ride home, the participant reported that using the Internet and/or phone were her pri- mary forms of socializing. She agreed to the challenge of participating in the present study because she was unhappy with what she described as a flabby belly, body fat, and the recent weight gain.

Apparatus

The Bowflex TreadClimber (Model # TC5000) was chosen for use in the present investigation because it was already available in the home of the participant, and suitable to provide a form of exercise that was not lim- ited by weather conditions. The TreadClimber is a hybrid of a treadmill and a stair climber. TreadClimber has two separate revolving belts that the user walks on, which also move up and down in a stepping motion, if that option is selected by the user. Used only as a treadmill, the TreadClimber is on an incline, which can be increased or decreased by changing the resistance level of the machine. The TreadClimber has the ability to record linear distance, vertical height climbed, speed, resistance, heart rate, and calorie burn during each use. These outputs were used to track specifically the linear distance covered and speed of “travel” for each session. For the present study, the TreadClimber was used as an inclined treadmill, and not as a stair climber.

942 Behavior Modification

Larwin, Larwin / Adolescent Fitness 943

In addition, the participant was instructed to record a body mass compo- sition (BMC) measurement every 3 days. This was measured using the TANITA BC533 InnerScan Body Composition Monitor. This device mea- sures body weight (lbs.), BMC (lbs/in2), muscle mass, and water mass. The participant was directed to record only the BMC measures to avoid unnec- essary ideation about weight fluctuations, which the parents reported the adolescent participant was prone to do. Specifically, the InnerScan was placed in an accessible location in the home, and the participant was instructed on the proper use of the instrument to obtain appropriate BMC measurement.

Procedure

In addition to the researcher and the participant, two other individuals were involved in conducting the present study. First, one of the child’s parents served as “coach” and monitored the child’s adherence to the program requirements, the child’s data collection in regard to the target behavior, and awarded consequences for successful completion of the tar- get behavior. A research assistant, working at the instruction of the researcher, assisted the parent in implementing the program and monitoring the child’s data collection activities. The research assistant maintained reg- ular communication with the parent and the child independently; commu- nicating with each at least three times weekly. The research assistant verified that activities were completed as specified and reinforcements were received or denied. This strategy was employed based on the reasoning that increasing accountability for project activities would increase the likeli- hood that the participant would follow the project activities through faith- fully. Lees and Dygdon’s (1988) research found that self-monitoring/ management projects were more successful when a coach or friend was pre- sent to help and encourage the participant in identifying improvements and to keep the participant focused on long-term goals.

A changing-criterion design (Kazdin, 1989) was used in the study. The target behavior was walking on the TreadClimber, measured in miles trav- eled. The specific primary goal for the present project was to increase the number of miles that the participant walked on the TreadClimber. Duration of the project was 7 weeks: the baseline stage lasted for 2 weeks, Treatment Phase 1 lasted for 2 weeks, and Treatment Phase 2 lasted for 3 weeks. Follow-ups were conducted at 5 weeks after and again at 1 year after ces- sation of the formal intervention.

Before the baseline stage, the research assistant provided data collection forms to the parent and the child. These forms allowed the child to record daily treadmill activity, daily Internet usage, daily phone usage, and BMC measurements. In addition, the parent was required to monitor the child’s daily treadmill activity, daily Internet usage, daily phone usage, and BMC readings and confirm the accuracy of the child’s reporting by initialing the data collection form. Each form provided room for data recording for a 2- week period. The research assistant provided the child and the parent with instruction on how to use the form and monitored their data recording. The parent returned the hard copies of the data collection forms to the research assistant. The child was also asked to report her recordings to the research assistant in a weekly e-mail communication, along with any personal com- ments about her weekly activity on the treadmill that she chose to share. This method of data collection was used to check the consistency of the child’s report with the parent’s report.

For the baseline stage of the present investigation, the participant’s activ- ity on the TreadClimber was monitored and recorded on the data sheets for a 2-week period. The participant was also asked to record Internet usage and phone usage (in minutes) during the baseline stage. The target behavior was examined in terms of frequency, speed, and distance traveled. Frequency was recorded in terms of number of sessions per week that the TreadClimber was used. Speed was based on the participant’s choice of miles/hour setting on the TreadClimber. Distance was recorded in terms of how many miles were traveled on the TreadClimber during each session of use. In addition, baseline measures of body mass, using the BMC, were taken every 3 days during the baseline phase for use as a reference to evaluate the degree of change in body mass observed during the subsequent treatment phases. The participant continued to record BMC measures every 3 days during the course of the two treatment phases.

To ensure that injury did not occur, the participant was instructed to con- duct a number of stretches, to be held for 15 to 30 seconds each, before TreadClimber use. The participant began each session on the TreadClimber by entering her current weight, setting the speed to begin at 1.2 miles per hour, and setting the resistance level to seven (which remained constant throughout the baseline and treatment sessions) on the machine’s control panel before beginning her walking session. Speed was increased at the par- ticipant’s discretion, as she became comfortable with the speed, but never exceeded a speed at which she could stand and talk without strain. After the participant walked as far as she was interested/able to walk for that session, she stopped the TreadClimber and recorded the final reading of distance

944 Behavior Modification

and minutes. Immediately after the session, the parent verified the record- ings. Once the target behavior was recorded, it was considered to have been completed for that session.

Reinforcement for performing the target behavior was provided on three levels during the course of the two treatment phases. For the first treatment phase, the participant received 1 hour of daily nonschool-related Internet time for every mile walked on the TreadClimber. For every addi- tional mile walked above and beyond the required 1 mile, the participant could earn an additional 1 hour of Internet time per night. For the second treatment phase, the participant received 1.5 hours of daily phone and/or Internet time in exchange for 1.5 miles walked per day. In addition, during the second treatment phase, for every ½ mile walked above and beyond the required 1.5 miles, the participant could earn an additional ½ hour of phone (but not Internet) time per night. Additional phone-usage time could only be earned with ½-mile increments, and additional distance walked more than ½ mile could not be rolled over for additional time on the fol- lowing day. For each treatment phase, the participant was encouraged to walk on the TreadClimber each day, at least 5 days of 7 days of the week. If the participant chose not to exercise on a particular day, one of the five she was encouraged to or for either of the two which she was not, the par- ticipant was not allowed any Internet time during Treatment Phase 1, or phone or Internet time during Treatment Phase 2 for those days. Finally, for successfully reducing her BMC by three units during the course of the two treatment phases, the participant was allotted up to 100 dollars by the parents to make a purchase of her choice.

During the treatment phases of the present project, if the participant did not achieve at least the 1-mile distance for a particular session in Treatment Phase 1, or the 1.5 mile distance in Treatment Phase 2, she did not receive any reinforcement in terms of Internet usage or Internet and/or phone usage, for that session. In the first treatment phase, fractions of a mile were not considered and did not accumulate for additional per-session Internet usage rewards. During Treatment Phase 1, the participant was asked to record her phone-usage time. Because phone-usage time was not the focus of the first treatment phase, there were no restrictions on the child’s phone access and usage during the first treatment phase. However, for both Treatment Phase 1 and Treatment Phase 2, Internet usage was contingent on exercise and therefore limited to the time earned by per- forming the target behavior. Likewise, for Treatment Phase 2, phone usage was contingent on exercise, and therefore limited to the time earned by performing the target behavior.

Larwin, Larwin / Adolescent Fitness 945

In addition to making the present study a changing-criterion design, the goal in increasing the difficulty of the mileage-walked requirement was to allow the participant to gradually increase the target behavior to a desirable level, while maintaining motivation and ultimately success. The more mod- est requirement for the target behavior in the first treatment phase was to make sure the participant had a chance of earning reinforcements and avoid- ing discouragement associated with initial goals that are set too high, while developing and adapting her body physically to the demands of this exercise. The goal for the end of this project was for the participant to be walking 1.5 miles per day/at least 5 days of the week; though her ultimate personal long- term goal is to walk 2 miles per day at least 6 days of the week.

The third level of reinforcement used in the present study addressed the participant’s desire to lose some inches around her waistline. This was addressed in the present study with the requirement that the participant demonstrates a reduction of at least 3 BMC units to earn 100 dollars to spend on something of her choice. The ultimate goal was that the participant would experience some natural reinforcement by the improvement in BMC and lost inches. It was hoped that this naturally occurring reinforcement would sup- port the continuation of the target behavior after the program was complete and its formal response-consequence contingencies were terminated.

Results

To demonstrate effectiveness of the exercise modification plan, data are presented on number of miles walked, speed of walking (mph), and BMC during the baseline, Treatment Phase 1, and Treatment Phase 2 peri- ods of the project. As can be seen in Figure 1, the participant was consis- tently not using the TreadClimber at all before Treatment Phase 1 of this investigation. During the 14 days of the baseline phase, the participant did not use the TreadClimber even once. During Treatment Phase 1 of the project, the participant began to use the TreadClimber more often. She used the treadmill every day—7 days a week—during the 14-day period and averaged of approximately .98 miles of walking per day. During the course of Treatment Phase 1, the participant’s behavior was stable at the criterion; her exercise level was precisely at 1 mile on 11 of the 14 days, with 2 days above the criterion and 1 day below.

During Phase 2, the participant again increased her use of the TreadClimber. During Treatment Phase 2, the participant exercised every day during the 21-day period and averaged 1.81 miles per day. This average

946 Behavior Modification

Larwin, Larwin / Adolescent Fitness 947

surpassed the treatment goal of 1.5 miles per day walked; her exercise level was at criterion on 9 days, above criterion on 10 days, and below criterion on 2 days. It is also interesting to note the change in exercise behavior during the course of Treatment Phase 2. The 2 days below criterion occurred toward the beginning of Treatment Phase 2, on Day 1 and Day 3. The days where she was well above the criterion occurred with greater frequency during the last half of Treatment Phase 2. This pattern indicates a substantial increase in exercise behavior from Treatment Phase 1 to Treatment Phase 2; and a sub- stantial increase during the course of Treatment Phase 2 in particular.

The present study did not include a contingency involving speed of walking. However, these data were recorded by the participant as part of the plan to monitor and record walking behavior. The participant demonstrated a steady increase in walking speed during Treatment Phase 1. The partici- pant walked at a speed of 3.0 miles per hour during the beginning of Treatment Phase 1 (Day 1 and Day 2), and this speed consistently increased to an average speed of 3.5 miles per hour during the course of Treatment Phase 1. In fact, for 7 of the 14 days of Treatment Phase 1, the participant exercised at the maximum possible setting available on the TreadClimber (4.0 miles/hour). These days were concentrated toward the end of Treatment Phase 1; 7 of the last 9 days of Phase 1, she exercised at the maximum speed setting. However, as indicated in Figure 2, although the participant

Figure 1 Overall Miles Walked During Baseline and Treatment Phases

0

0.5

1

1.5

2

2.5

3

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51

Days

D is

ta n

c e i

n M

il e s

BASELINE PHASE ONE PHASE TWO

948 Behavior Modification

did maintain this increase in average speed above her phase one starting point during the course of phase two, she did not maintain her speed dur- ing Treatment Phase 2 at the level it was at toward the end of Treatment Phase 1. The number of days at which the participant exercised at the max- imum speed of 4.0 miles/hour was only 3 days during the entire 21 days of Treatment Phase 2. Although her average speed during Treatment Phase 2 was 3.67 miles per hour, indicating a slight increase in average speed for Treatment Phase 2 relative to Treatment Phase 1, this increase is a result of the average for Treatment Phase 1 being negatively reduced by a few days of much lower speed concentrated at the beginning of Treatment Phase 1. Figure 2 presents data on the speed at which the participant walked during the baseline phase, Treatment Phase 1, and Treatment Phase 2.

Figure 3 presents data on the BMC measurements taken during the base- line phase, treatment phase 1, and treatment phase 2. As indicated, the par- ticipant experienced a steady decrease in BMC measures across the two treatment phases. The total BMC decrease during the project was 3.2 lbs/in2.

Finally, an examination of the Internet and phone usage of the partici- pant indicates a steady decline in Internet and phone usage across all three stages of this project. The reported Internet time during Treatment Phase 1 reflects the Internet time earned through exercise; the phone time reported

Figure 2 Overall Speed During Baseline, Treatment Phase 1,

and Treatment Phase 2

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49

Days

M il e s P

e r

H o

u r

BASELINE PHASE ONE PHASE TWO

Larwin, Larwin / Adolescent Fitness 949

during Treatment Phase 1 was not included in the contingency for this phase. As indicated, in Table 1, phone usage did not show a dramatic decrease from Treatment Phase 1 to Treatment Phase 2. However, as indi- cated earlier, the participant was required to exercise more to have phone time during Treatment Phase 2.

Discussion

The present investigation attempted to incorporate several features of the Jason and Brackshaw (1999) study in which an adolescent was required to

Figure 3 Body Mass Composition Measurements During Baseline,

Treatment Phase 1, and Treatment Phase 2 (A Break in the Data Point Within a Phase Indicates a Phase Interrupt)

Table 1 Mean Daily Usage of Nonhomework-related

Internet and Phone (In Minutes)

Period Baseline Phase 2 Phase 2

Internet 38.64 25.07 10.95 Phone 63.21 41.71 31.56

950 Behavior Modification

earn television time by exercising. Specifically, the present study aimed to reduce Internet usage and cell phone usage of the adolescent-aged partici- pant, rather than television viewing, in addition to the primary goal of encour- aging a lifestyle change involving more physical exercise by the participant.

For the present investigation, the participant was required to perform physical activity, in the form of treadmill use, in exchange for nonschool- related time on the Internet during Treatment Phase 1 and Internet and phone time during Treatment Phase 2. During Treatment Phase 1 of this project, the participant walked 1 mile on the TreadClimber to earn 1 hour of Internet time. However, during Treatment Phase 2, the participant was required to walk at least 1.5 miles to have up to 1.5 hours of Internet and/or phone time. Additional exercise in ½-mile increments beyond the 1.5 miles required earned the participant additional phone-usage time.

Results of this study suggest that the participant was successful in meet- ing her present project goals. During Treatment Phase 1, the participant walked an average of .98 miles per day, at an average rate of approximately 3.5 miles per hour. During Treatment Phase 2, the participant walked an average 1.81 miles per day, at an average rate of approximately 3.67 miles per hour. During the two treatment phases of the project, the participant was successful in dropping her BMC approximately 3.2 BMC units. In addition, follow-up data indicated that the participant, 5 weeks after the cessation of this formal investigation, continued to average approximately 1.5 miles of walking on the treadmill per day. A 1-year follow-up indicated that the par- ticipant continued to exercise approximately 5 days a week—compared with little to no exercise before participation in the study—and averaged less than 1 hour-a-day of nonschool-related Internet and cell phone usage, compared with more than 1.6 hour-a-day before participation in the study. The partic- ipant has continued to lose excess weight, and happily reports that she has reached her goal weight of 115 pounds (however, she remains 5’2” tall). These results suggest indeed that a lifestyle change has been made.

One issue that deserves some additional comment is the changes in both the design of the study and the pattern of results demonstrated by the partic- ipant from Treatment Phase 1 to Treatment Phase 2. Although Treatment Phase 1 of the study focused on making Internet use only contingent on physical exercise, Treatment Phase 2 made Internet use and phone time con- tingent on exercise. This change was implemented as a result of the pattern of results for exercise on the treadmill demonstrated in Treatment Phase 1 and the monitoring of the participant’s phone usage during baseline and in Treatment Phase 1. The participant walked on the treadmill to the criterion

number of miles required for most of Treatment Phase 1, but did not very often increase her miles walked above the criterion even though she could have earned additional Internet time for doing so. In fact, she walked for mileage above the criterion only 2 of 14 days, and neither of those 2 days did she walk far enough more than the criterion to earn additional reinforce- ment. Combined with the observation that the participant was spending a great deal of time on the phone during baseline and Treatment Phase 1, it seemed that Internet time was simply not as powerful a reinforcer.

Therefore, phone time was added to the exercise–Internet time contin- gency for Treatment Phase 2, and was the sole contingent consequence available for additional exercise time over and above the criterion, to encourage the participant to continue to increase her daily exercise activity. In addition, the requirement for the miles walked on the treadmill over and above the criterion to earn additional reinforcement was reduced also to encourage the participant to increase her daily exercise activity. In Treatment Phase 1, the participant was required to walk on the treadmill an additional full mile to earn an extra hour of time on the Internet. However, for Treatment Phase 2, the participant was only required to exercise an addi- tional half mile to earn an additional ½ hour of phone time.

The pattern of results for Treatment Phase 2 suggests that modification of the program criterion and pattern of contingencies was successful. The par- ticipant walked above the criterion a total of 10 of 21 days; with 3 of those 10 being a ½ mile above criterion and 2 of them being a full mile above cri- terion, for a total of 5 days where the participant earned extra reinforcement. This compares with only 2 of 14 days above criterion in Treatment Phase 1, and not a single day where additional reinforcement was earned.

Another finding of interest in the pattern of results obtained is that the pattern of results for the speed at which the participant walked during the course of the two treatment phases did not correspond to the aforemen- tioned changes in walking distance during the two treatment phases. In Treatment Phase 1, the miles walked were quite stable, whereas this mea- sure increased considerably during the course of Treatment Phase 2. However, a different pattern was observed for the walking speed data. During the course of Treatment Phase 1, walking speed increased from its starting point and then stabilized at the highest level possible on the tread- mill for much of this treatment phase. In 7 of the 14 days of Treatment Phase 1, the participant walked at the highest speed possible. But for Treatment Phase 2, walking speed began to fluctuate, with far fewer days with walking speed recorded at the highest level (only 3 out of 21). And toward the end of Treatment Phase 2, when miles walked was increasing to

Larwin, Larwin / Adolescent Fitness 951

its highest levels observed during the study, walking speed stabilized at an even lower rate. Not a single day in the last 10 days of Treatment Phase 2 evidenced the highest possible walking speed.

One possible explanation for this discrepant pattern of results for walk- ing distance and walking speed is that it reflects the participant’s changing goals; a change from a simple desire to get her walking requirement com- pleted in Treatment Phase 1 and a desire to walk more miles to earn cell phone reinforcement in particular in Treatment Phase 2. In the personal report comments the participant shared with the parent and research assis- tant, she indicated that she disliked walking on the treadmill, and in Treatment Phase 1, simply wanted to get her minimum walking require- ment done as quickly as possible. Walking at the highest speed possible would get her to that goal the fastest. However, when her more valued phone time was made contingent on walking a greater distance in Treatment Phase 2, the goal became reaching the then higher mileage requirement, and walking even farther over and above that to earn additional phone time. At this point, it likely became an endurance issue. To have the stamina and endurance to reach these higher mileage goals, the participant needed to walk at a slower rate.

Another question that could be raised in reference to the pattern of results evidenced in the present study is what causal mechanisms are responsible for the changes in exercise activity. The design of the study included contingent consequences that were intended to increase miles walked on the treadmill and reduce BMC measures recorded for the partic- ipant. Although there were no such contingencies in place to affect walking speed in particular, it seems reasonable that the factors that might have affected walking mileage and BMC reduction might have also affected walking speed during the course of the study, as indicated in the preceding discussion. However, was it truly the established contingencies intended to influence walking distance and BMC measures that were responsible for the observed changes in those variables?

One possible counterargument is that the participant was simply highly motivated to change her exercise behavior and this motivation was respon- sible for the changes observed independent of the program design and implementation. In support of this line of reasoning, it could be argued that the participant has already demonstrated such motivation in her recent history of weight loss with self-imposed and self-regulated dietary changes. Perhaps such effects are responsible for the results of the present study.

There is evidence that argues against this interpretation. Although the participant did in fact lose significant weight with her own self-initiated diet

952 Behavior Modification

program, she was not able to maintain this weight loss on her own as she gained some of it back, and was not able to resume a pattern of weight loss using the self-motivational efforts she employed with some success origi- nally. Had simple personal motivation been sufficient to bring about the kinds of changes evidenced in the present study, there would have been no need for the present investigation and the participant would have already accomplished the additional weight loss and increased exercise goals she desired independently. Yet there was noticeable change in the participant’s frequency, duration, and intensity of exercise demonstrated during the course of the present study. This suggests that the program itself is more likely a causal factor behind the changes observed than is simple personal motivation and preexisting self-discipline skills of the participant.

However, even if the self-motivation hypothesis can logically be dis- missed, it is still a valid question as to whether it was the characteristics of the program itself that was the mechanism behind the changes observed in the participant’s exercise behavior. One characteristic of the changing-criterion design employed in the present investigation is that it is “less satisfactory than reversal and multiple-baseline designs in ruling out the influence of extraneous events that might account for behavior change” (Kazdin, 1989, pp. 91-92). However, Kazdin (1989) argues that this is less of a concern in changing-criterion designs “when performance matches the criterion very closely” (p. 92). The performance in the current study does often closely match the criterion. In Treatment Phase 1, for example, the participant per- formed walking behavior on the treadmill, precisely at criterion on 11 out of 14 days.

Another feature of the present study is that it confirmed a pattern of results, and a potential mechanism responsible for those results, that has been observed in other research. Jason and Brackshaw (1999) demonstrated a sig- nificant reduction in an adolescent’s television viewing time by making such viewing contingent on the performance of physical activity. The present study found the same pattern of results with respect to Internet and phone usage; both of these forms of media use were reduced when made contingent on physical exercise. Although it may seem odd that in Jason and Brackshaw’s study and in the current investigation that the participants demonstrated a reduction in the very activities they were earning as reinforcers—they appear to not be using the reinforcements they are earning—there is some evidence for why this pattern of results was observed. First, by making the activities contingent on performance of a target behavior, and not allowing them to occur at all in the absence of performance of the target behavior, and making the contingent consequence of time allowed (on the phone, for

Larwin, Larwin / Adolescent Fitness 953

example) less than how much it was naturally occurring during base line conditions, the overall performance time of these activities must logically decrease.

A greater mystery, perhaps, is why the Internet and phone usage in the present study, and television viewing in the Jason and Brackshaw (1999) study, continued to decrease after the termination of the respective investi- gations even though they were no longer serving as contingent conse- quences in a formal program. The likely answer is what was suggested by Jason and Brackshaw at the conclusion of their study and more directly observed in the present study. The behaviors used as contingent conse- quences likely decreased as a result of response competition. As exercise time increased and was maintained—perhaps by naturally occurring rein- forcers at the end of the program—there is less time for the performance of other behaviors (such as Internet and phone usage and television viewing). Although the Jason and Brackshaw study did not monitor, record, and col- lect data on their target behavior of physical activity, the present study did indeed focus on monitoring and recording the target behavior of physical exercise; both during the formal study itself and at two follow-up periods. Physical activity in the form of treadmill use during the course of the inter- vention did indeed increase, leaving less time available to the participant for Internet and phone usage, and the follow-up data indicated that the increased level of physical activity had been maintained after the termina- tion of the formal behavior change program (although not specifically in the form of treadmill usage), suggesting that response competition may have continued to remain in effect and contribute to the decreased phone and Internet use also observed at follow-up.

As indicated earlier, the participant remarked that she did not enjoy the time she spent on the TreadClimber. This is a valid concern. According to a Canadian study on factors affecting exercise involvement, Wankel (1985) found a significant connection between participants who reported enjoyment of an exercise program and their consistent maintenance of that program. Wankel maintained that even if social support was present, and goals were set, participants who did not enjoy their exercise program were less likely to follow through to meeting their goals. Future research should investigate ways to increase participants’ use and potential enjoyment of indoor exercise equipment, if long-term use of exercise equipment is to be likely.

In the present study, the project came to an end as spring weather began to return, thus enabling the participant to continue her exercise program with outdoor activities. Perhaps it was this admittedly serendipitous feature that allowed the participant to continue her exercise program in the absence

954 Behavior Modification

of the program’s formal response-consequence contingencies. One year after the study, the participant reported that she was riding her bike daily to the gym, where she participates in an exercise regimen recommended by a physical fitness trainer. The participant reported that she enjoys the variety of exercise equipment available for her use at the gym. It would appear that several naturally occurring reinforcers may have replaced the formal program’s contingent consequences to maintain the participant’s target behavior after the program terminated.

Results of the present investigation offer some hope in the face of a growing societal concern. The emergent crisis of obesity and lack of phys- ical fitness in an increasing number of adolescents in the United State today indicates that more research is needed. Finding ways in which young people might take responsibility for their own level of physical fitness, reduce excessive media use, and make healthy lifestyle changes is pertinent to the long-term health of the present and future generations. As this study suggests, exercise contingency plans that incorporate adolescents’ access to and usage of various types of media are promising in terms of their poten- tial to achieve these ends, and most certainly merit further investigation.

References

American Academy of Pediatrics. (2003). Prevention of pediatric overweight and obesity. Pediatrics, 112, 424-430.

Anderson, R., Crespo, C., Bartlett, S., Cheskin, L., & Pratt, M. (1998). Relationship of phys- ical activity and television watching with body weight and level of fatness among children. Journal of the American Medical Association, 279, 938-942.

Centers for Disease Control and Prevention. (2002). Overweight among U.S. children and adoles- cents. National Health and Nutrition Examination Survey. Retrieved on March 19, 2006, from http://www.cdc.gov/nchs/data/nhanes/databriefs/overwght.pdf#search=’cdc.gov

Crespo, C., Smit, E., Troiano, R., Bartlett, S., Macera, C., & Anderson, R. (2001). Television watching, energy intake, and obesity in U.S. children. Archives of Pediatrics & Adolescent Medicine, 155, 360-365.

Dietz, W., & Gortmaker, S. (1993). TV or not TV: Fat is the question. Pediatrics, 91, 499-501. Durant, R., & Brananowski, T. (1994). The relationship among television watching, physical

activity, and body composition of young children. Pediatrics, 94, 449-455. Gortmaker, S., Must, A., Sobol, A., Peterson, K., Colditz, S., & Dietz, W. (1996). Television

viewing as a cause of increasing obesity among children in the United States, 1986-1990. Archives of Pediatrics & Adolescent Medicine, 150, 356-362.

Jason, L. A. (1987). Reducing children’s television viewing and assess secondary changes. Journal of Clinical Child Psychology, 16, 245-250.

Jason, L. A., & Brackshaw, E. (1999). Access to TV contingent on physical activity: Effects on reducing TV-viewing and body-weight. Journal of Behavior Therapy and Experimental Psychiatry, 30, 145-151.

Larwin, Larwin / Adolescent Fitness 955

956 Behavior Modification

Jason, L. A., & Johnson, S. (1995). Reducing excessive television viewing while increasing physical activity. Child & Family Behavior Therapy, 17, 35-45.

Jason, L. A., Johnson, S., & Jurs, A. (1993). Reducing children’s television viewing with an inexpensive lock. Child & Family Behavior Therapy, 15, 45-54.

Jason, L. A., & Klich, M. M. (1982). Use of feedback in reducing television watching. Psychological Reports, 51, 812-814.

Johnson, S. Z., & Jason, L. A. (1996). Evaluation of a device aimed at reducing children’s tele- vision viewing. Child & Family Behavior Therapy, 18, 59-61.

Kaiser Family Foundation. (1999). Kids and media use. Retrieved May 14, 2008, from http://www.kaisernetwork.org/gsaresults/search?site=KFForgnopdfs&filter=0&output= xml_no_dtd&client=kff&sp=kff&getfields=*&q=Kids%20and%20media&no_pdf=1

Kazdin, A. E. (1989). Behavior modification in applied settings (4th ed.). Belmont, CA: Wadsworth. Kunkel, D. (2001). Children and television advertising. In D. Singer & J. Singer (Eds.),

Handbook of children and the media (pp. 375-393). Thousand Oaks, CA: Sage. Lees, L. A., & Dygdon, J. A. (1988). The initiation and maintenance of exercise behavior: A

learning theory conceptualization. Clinical Psychology Review, 8, 345-353. Lowry, R., Wechsler, H., Galuska, D., Fulton, J., & Kann, L. (2002). Television viewing and

its association with overweight, sedentary lifestyle, and insufficient consumption of fruits and vegetables among U.S. high school students: Differences in race, ethnicity, and gender. Journal of School Health, 72, 413-421.

Meyering, K. (2005). Meet generation M. News & resources from the Kaiser Youth Media Study released March 9, 2005. Retrieved May, 14, 2007, from http://www.frankwbaker.com/ kids_media_report.htm

Parr, J. (2005). U.S. entertainment & media consumer survey. Retrieved on March 6, 2006, from http://www.jupitermedia.com/corporate/releases/06.01.30-newjupresearch.html

Proctor, M., Moore, L., Gao, D., Cupples, L., Bradlee, M., Hood, M., et al. (2003). Television viewing and change in body fat from preschool to early adolescence: The Framinghan children’s study. International Journal of Obesity, 27, 827-833.

Robinson, T., Hammer, L., Killen, J., Kraemer, H., Wilson, D., Hayward, C., et al. (1993). Does television viewing increase obesity and reduce physical activity: Cross-sectional and longitudinal analyses among adolescent girls. Pediatrics, 81, 273-280.

Robinson, T., & Killen, J. (1995). Ethics and gender differences in the relationships between television viewing and obesity, physical activity, and dietary fat intake. Journal of Health Education, 26, 91-98.

Wankel, L. M. (1985). Personal and situational factors affecting exercise involvement: The importance of enjoyment. Research Quarterly for Exercise and Sport, 56, 275-282.

Karen H. Larwin, PhD, a graduate of Kent State University’s Measurement and Evaluation Program. Her interests include developing better methods for assessing reliability and using meta-analytic applications with single-subject designs. Presently, she teaches research methods and graduate-level statistics sections for the Organizational Learning and Leadership Program at Gannon University.

David A. Larwin is a lecturer in personality and social psychology at Kent State University Salem in Salem, Ohio. His professional interests include the psychology of the self, human aggression, and program evaluation. He has received a number of honors and awards, including a Dissertation Research Award from the American Psychological Association.